| Modifier and Type | Method and Description |
|---|---|
Map<String,INDArray> |
TFGraphRunnerService.run(Map<String,INDArray> inputs) |
| Modifier and Type | Method and Description |
|---|---|
TFGraphRunnerService |
TFGraphRunnerService.init(List<String> inputNames,
List<String> outputNames,
byte[] graphBytes,
Map<String,INDArray> constants,
Map<String,String> inputDataTypes) |
Map<String,INDArray> |
TFGraphRunnerService.run(Map<String,INDArray> inputs) |
| Modifier and Type | Method and Description |
|---|---|
T |
OutputAdapter.apply(INDArray... outputs)
This method provides conversion from multiple INDArrays to T
|
| Modifier and Type | Method and Description |
|---|---|
INDArray[] |
Operands.asArray()
This method return operands as array, in order of addition
|
INDArray |
Operands.getById(int id)
This method returns array identified its numeric id
|
INDArray |
Operands.getById(int id,
int index)
This method returns array identified its numeric id and index
|
INDArray |
Operands.getById(@NonNull String name)
This method returns array identified its name
|
| Modifier and Type | Method and Description |
|---|---|
Collection<Pair<Operands.NodeDescriptor,INDArray>> |
Operands.asCollection()
This method returns contents of this entity as collection of key->value pairs
|
| Modifier and Type | Method and Description |
|---|---|
Operands |
Operands.addArgument(int id,
@NonNull INDArray array)
This method allows to pass array to the node identified by numeric id
|
Operands |
Operands.addArgument(int id,
int index,
@NonNull INDArray array)
This method allows to pass array to multi-output node in the graph
|
Operands |
Operands.addArgument(@NonNull String id,
@NonNull INDArray array)
This method allows to pass array to the node identified by its name
|
Operands |
Operands.addArgument(String name,
int id,
int index,
@NonNull INDArray array)
This method allows to pass array to multi-output node in the graph
|
| Modifier and Type | Field and Description |
|---|---|
protected INDArray |
DifferentialFunction.scalarValue |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
DifferentialFunction.getInputArgument(int index) |
| Modifier and Type | Method and Description |
|---|---|
void |
BaseEvaluationListener.activationAvailable(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation) |
void |
Listener.activationAvailable(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation)
Called when any activation becomes available.
|
void |
BaseListener.activationAvailable(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation) |
void |
BaseEvaluationListener.activationAvailableEvaluations(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation)
|
void |
Listener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs)
Called at the end of each operation execution
|
void |
BaseListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs) |
void |
Listener.preUpdate(SameDiff sd,
At at,
Variable v,
INDArray update)
Called just before each parameter is to be updated - i.e., just before each parameter is modified.
|
void |
BaseListener.preUpdate(SameDiff sd,
At at,
Variable v,
INDArray update) |
| Modifier and Type | Method and Description |
|---|---|
void |
ArraySavingListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs) |
void |
OpBenchmarkListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs) |
| Modifier and Type | Method and Description |
|---|---|
void |
UIListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs) |
void |
UIListener.preUpdate(SameDiff sd,
At at,
Variable v,
INDArray update) |
| Modifier and Type | Method and Description |
|---|---|
void |
ProfilingListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs) |
| Constructor and Description |
|---|
LossCurve(INDArray lossValues,
List<String> lossNames) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
SDVariable.eval()
Evaluate the result of this variable
|
INDArray |
SDVariable.eval(Map<String,INDArray> placeholders)
Evaluate the result of this variable
|
INDArray |
SDVariable.getArr()
A getter for the allocated ndarray with this
SDVariable. |
INDArray |
SDVariable.getArr(boolean enforceExistence)
A getter for the allocated ndarray with this
SDVariable. |
INDArray |
ArrayHolder.getArray(String name) |
INDArray |
SameDiff.getArrForVarName(@NonNull String varName)
Get an
INDArray for a given vertex id, or null if none exists |
INDArray |
SameDiff.getEagerArrForVarName(@NonNull String varName)
Note this is a special getter for the eager holder.
|
INDArray |
SameDiff.outputSingle(Map<String,INDArray> placeholders,
String output)
Do inference for a single variable for a single batch.
|
INDArray |
ArrayHolder.removeArray(String name)
Remove the array from the ArrayHolder, returning it (if it exists)
|
| Modifier and Type | Method and Description |
|---|---|
protected Map<String,INDArray> |
SameDiff.batchOutputHelper(Map<String,INDArray> placeholders,
List<Listener> listeners,
Operation operation,
String... outputs) |
Map<String,INDArray> |
SameDiff.calculateGradients(Map<String,INDArray> placeholderVals,
@NonNull Collection<String> variables)
Calculate and return the gradients for the specified variables
|
Map<String,INDArray> |
SameDiff.calculateGradients(Map<String,INDArray> placeholderVals,
String... variables)
|
protected Map<String,INDArray> |
SameDiff.directExecHelper(Map<String,INDArray> placeholders,
At at,
MultiDataSet batch,
Collection<String> requiredActivations,
List<Listener> activeListeners,
String... outputs)
Do inference for the given variables for a single batch, with training information
|
Map<String,INDArray> |
SameDiff.output(@NonNull DataSetIterator iterator,
@NonNull List<Listener> listeners,
String... outputs)
Do inference on a network with a single input.
For example, if the variable to infer was called "softmax" you would use: |
Map<String,INDArray> |
SameDiff.output(@NonNull DataSetIterator dataSet,
String... outputs)
|
Map<String,INDArray> |
SameDiff.output(@NonNull DataSet dataSet,
String... outputs)
Do a single batch inference on a network with a single input.
For example, if the variable to infer was called "softmax" you would use: |
Map<String,INDArray> |
SameDiff.output(Map<String,INDArray> placeholders,
List<Listener> listeners,
String... outputs)
Do inference for the given variables for a single batch.
|
Map<String,INDArray> |
SameDiff.output(Map<String,INDArray> placeholders,
@NonNull List<String> outputs)
Do inference for the given variables for a single batch.
|
Map<String,INDArray> |
SameDiff.output(Map<String,INDArray> placeholders,
String... outputs)
Do inference for the given variables for a single batch.
|
Map<String,INDArray> |
SameDiff.output(@NonNull MultiDataSetIterator iterator,
@NonNull List<Listener> listeners,
String... outputs)
Perform inference.
Example: classifier inference Predictions variable name: "softmaxOutput" Evaluations to perform: EvaluationData: single output MultiDataSets Code: |
Map<String,INDArray> |
SameDiff.output(@NonNull MultiDataSetIterator dataSet,
String... outputs)
|
Map<String,INDArray> |
SameDiff.output(@NonNull MultiDataSet dataSet,
String... outputs)
Do a single batch inference on a network.
For example, if the variable to infer was called "softmax" you would use: |
Map<String,INDArray> |
SameDiff.outputAll(Map<String,INDArray> placeholders)
Do inference for all variables for a single batch.
|
List<Map<String,INDArray>> |
SameDiff.outputBatches(DataSetIterator iterator,
List<Listener> listeners,
String... outputs)
See
SameDiff.output(DataSetIterator, List, String...), but without the concatenation of batches. |
List<Map<String,INDArray>> |
SameDiff.outputBatches(DataSetIterator iterator,
String... outputs)
See
SameDiff.output(DataSetIterator, String...), but without the concatenation of batches. |
List<Map<String,INDArray>> |
SameDiff.outputBatches(MultiDataSetIterator iterator,
List<Listener> listeners,
String... outputs)
Perform inference.
Example: classifier inference Predictions variable name: "softmaxOutput" Evaluations to perform: EvaluationData: single output MultiDataSets Code: |
List<Map<String,INDArray>> |
SameDiff.outputBatches(MultiDataSetIterator iterator,
String... outputs)
|
| Modifier and Type | Method and Description |
|---|---|
void |
SameDiff.assignArray(@NonNull INDArray arr,
@NonNull SDVariable variable)
Update the constant or variable type SDVariable with the values from the specified
array.
|
void |
SameDiff.associateArrayWithVariable(INDArray arr,
SDVariable variable)
Associate the array with the given variable.
|
void |
SameDiff.associateArrayWithVariable(INDArray arr,
@NonNull String variable)
Associate the array with the given variable.
|
SDVariable |
SameDiff.constant(@NonNull INDArray constant)
Create an SDVariable with a fixed/constant value, with a generated name
Constants are not modified by training/backprop. |
SDVariable |
SameDiff.constant(String name,
@NonNull INDArray constant)
Create an SDVariable with a fixed/constant value
Constants are not modified by training/backprop. |
SDVariable |
SDVariable.setArray(INDArray array)
Associate the specified array with this variable
|
void |
ArrayHolder.setArray(String name,
INDArray array)
Set the array for the specified name (new array, or replace if it already exists)
|
void |
SameDiff.setArrayForVariable(@NonNull String varName,
@NonNull INDArray arr)
Set the stored
INDArray for a variable. |
void |
SameDiff.setEagerArrForVarName(@NonNull String varName,
INDArray arr)
Sets an array for the given variable name in the
eager session.
|
SDVariable |
SameDiff.var(INDArray arr)
Create an
SDVariable with a generated name, and assocate the specified array with it.This is a VARIABLE type SDVariable - i.e., must be floating point, and is a trainable parameter. |
SDVariable |
SameDiff.var(String name,
@NonNull INDArray arr)
Create an
SDVariable with the specified name, and associate the specified array with itThis is a VARIABLE type SDVariable - i.e., must be floating point, and is a trainable parameter. |
| Modifier and Type | Method and Description |
|---|---|
protected Map<String,INDArray> |
SameDiff.batchOutputHelper(Map<String,INDArray> placeholders,
List<Listener> listeners,
Operation operation,
String... outputs) |
Map<String,INDArray> |
SameDiff.calculateGradients(Map<String,INDArray> placeholderVals,
@NonNull Collection<String> variables)
Calculate and return the gradients for the specified variables
|
Map<String,INDArray> |
SameDiff.calculateGradients(Map<String,INDArray> placeholderVals,
String... variables)
|
OutAndGrad |
SameDiff.calculateGradientsAndOutputs(Map<String,INDArray> placeholderVals,
Collection<String> outputVars,
Collection<String> gradientVars)
Calculate the activations and the gradients for the specified variables, in one execution call.
|
SDVariable[] |
SameDiffFunctionDefinition.define(SameDiff sameDiff,
Map<String,INDArray> inputs,
SDVariable[] variableInputs) |
void |
SameDiff.defineFunction(String function,
SameDiffFunctionDefinition functionDefinition,
Map<String,INDArray> inputs) |
protected Map<String,INDArray> |
SameDiff.directExecHelper(Map<String,INDArray> placeholders,
At at,
MultiDataSet batch,
Collection<String> requiredActivations,
List<Listener> activeListeners,
String... outputs)
Do inference for the given variables for a single batch, with training information
|
INDArray |
SDVariable.eval(Map<String,INDArray> placeholders)
Evaluate the result of this variable
|
Map<String,INDArray> |
SameDiff.output(Map<String,INDArray> placeholders,
List<Listener> listeners,
String... outputs)
Do inference for the given variables for a single batch.
|
Map<String,INDArray> |
SameDiff.output(Map<String,INDArray> placeholders,
@NonNull List<String> outputs)
Do inference for the given variables for a single batch.
|
Map<String,INDArray> |
SameDiff.output(Map<String,INDArray> placeholders,
String... outputs)
Do inference for the given variables for a single batch.
|
Map<String,INDArray> |
SameDiff.outputAll(Map<String,INDArray> placeholders)
Do inference for all variables for a single batch.
|
INDArray |
SameDiff.outputSingle(Map<String,INDArray> placeholders,
String output)
Do inference for a single variable for a single batch.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
ThreadSafeArrayHolder.getArray(@NonNull String name) |
INDArray |
SingleThreadArrayHolder.getArray(@NonNull String name) |
INDArray |
OptimizedGraphArrayHolder.getArray(String name) |
INDArray |
ThreadSafeArrayHolder.removeArray(@NonNull String name) |
INDArray |
SingleThreadArrayHolder.removeArray(@NonNull String name) |
INDArray |
OptimizedGraphArrayHolder.removeArray(String name) |
| Modifier and Type | Method and Description |
|---|---|
void |
ThreadSafeArrayHolder.setArray(@NonNull String name,
@NonNull INDArray array) |
void |
SingleThreadArrayHolder.setArray(@NonNull String name,
@NonNull INDArray array) |
void |
OptimizedGraphArrayHolder.setArray(String name,
INDArray array) |
| Modifier and Type | Method and Description |
|---|---|
void |
OptimizedGraphArrayHolder.setFunction(String name,
Supplier<INDArray> fn) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BatchOutputConfig.execSingle()
Deprecated.
|
INDArray |
OutputConfig.execSingle()
Do inference and return the results for the single output variable specified.
|
INDArray |
BatchOutputConfig.outputSingle()
Do inference and return the results for the single output
Only works if exactly one output is specified
|
| Modifier and Type | Method and Description |
|---|---|
Map<String,INDArray> |
BatchOutputConfig.exec()
Deprecated.
|
Map<String,INDArray> |
OutputConfig.exec()
Do inference and return the results.
|
List<Map<String,INDArray>> |
OutputConfig.execBatches()
Do inference and return the results in batches.
|
List<INDArray> |
OutputConfig.execSingleBatches()
Do inference and return the results (in batches) for the single output variable specified.
|
Map<String,INDArray> |
BatchOutputConfig.output()
Do inference and return the results
|
| Modifier and Type | Method and Description |
|---|---|
BatchOutputConfig |
BatchOutputConfig.input(@NonNull SDVariable variable,
@NonNull INDArray placeholder)
|
BatchOutputConfig |
BatchOutputConfig.input(@NonNull String variable,
@NonNull INDArray placeholder)
Add a placeholder value for a specified variable
|
| Modifier and Type | Method and Description |
|---|---|
BatchOutputConfig |
BatchOutputConfig.inputs(Map<String,INDArray> placeholders)
Calls
BatchOutputConfig.input(String, INDArray) on each entry in the map. |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
SessionMemMgr.allocate(boolean detached,
DataType dataType,
long... shape)
Allocate an array with the specified datatype and shape.
NOTE: This array should be assumed to be uninitialized - i.e., contains random values. |
INDArray |
SessionMemMgr.allocate(boolean detached,
LongShapeDescriptor descriptor)
As per
SessionMemMgr.allocate(boolean, DataType, long...) but from a LongShapeDescriptor instead |
INDArray[] |
InferenceSession.doExec(DifferentialFunction op,
OpContext opContext,
FrameIter outputFrameIter,
Set<AbstractSession.VarId> opInputs,
Set<AbstractSession.VarId> allIterInputs,
Set<String> constAndPhInputs) |
INDArray |
SessionMemMgr.dup(INDArray arr)
Duplicate the specified array, to an array that is managed/allocated by the session memory manager
|
protected INDArray |
InferenceSession.getArray(SDVariable sdv,
Collection<AbstractSession.VarId> opInputs,
Collection<AbstractSession.VarId> allIterInputs) |
INDArray |
InferenceSession.getConstantOrVariable(String variableName) |
INDArray[] |
InferenceSession.getOutputs(Pair<SameDiffOp,OpContext> opPair,
FrameIter outputFrameIter,
Set<AbstractSession.VarId> opInputs,
Set<AbstractSession.VarId> allIterInputs,
Set<String> constAndPhInputs,
List<Listener> listeners,
At at,
MultiDataSet batch,
Set<String> allReqVariables) |
INDArray[] |
TrainingSession.getOutputs(Pair<SameDiffOp,OpContext> opPair,
FrameIter outputFrameIter,
Set<AbstractSession.VarId> opInputs,
Set<AbstractSession.VarId> allIterInputs,
Set<String> constAndPhInputs,
List<Listener> listeners,
At at,
MultiDataSet batch,
Set<String> allReqVariables) |
INDArray[] |
InferenceSession.getOutputsHelperTensorArrayOps(DifferentialFunction op,
FrameIter outputFrameIter,
Set<AbstractSession.VarId> opInputs,
Set<AbstractSession.VarId> allIterInputs)
Forward pass for TensorArray ops
|
INDArray |
SessionMemMgr.ulike(INDArray arr)
Allocate an uninitialized array with the same datatype and shape as the specified array
|
| Modifier and Type | Method and Description |
|---|---|
protected Map<String,INDArray> |
InferenceSession.postProcessOutput(Map<String,INDArray> output) |
protected Map<String,INDArray> |
InferenceSession.preprocessPlaceholders(Map<String,INDArray> placeholders,
At at) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
SessionMemMgr.dup(INDArray arr)
Duplicate the specified array, to an array that is managed/allocated by the session memory manager
|
void |
SessionMemMgr.release(INDArray array)
Release the array.
|
INDArray |
SessionMemMgr.ulike(INDArray arr)
Allocate an uninitialized array with the same datatype and shape as the specified array
|
| Modifier and Type | Method and Description |
|---|---|
Pair<SameDiffOp,OpContext> |
InferenceSession.getAndParameterizeOp(String opName,
FrameIter frameIter,
Set<AbstractSession.VarId> opInputs,
Set<AbstractSession.VarId> allIterInputs,
Set<String> constAndPhInputs,
Map<String,INDArray> placeholderValues,
Set<String> allReqVariables) |
protected Map<String,INDArray> |
InferenceSession.postProcessOutput(Map<String,INDArray> output) |
protected Map<String,INDArray> |
InferenceSession.preprocessPlaceholders(Map<String,INDArray> placeholders,
At at) |
Loss |
TrainingSession.trainingIteration(TrainingConfig config,
Map<String,INDArray> placeholders,
Set<String> paramsToTrain,
Map<String,GradientUpdater> updaters,
MultiDataSet batch,
List<String> lossVariables,
List<Listener> listeners,
At at)
Perform one iteration of training - i.e., do forward and backward passes, and update the parameters
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
ArrayCacheMemoryMgr.allocate(boolean detached,
DataType dataType,
long... shape) |
INDArray |
ArrayCloseMemoryMgr.allocate(boolean detached,
DataType dataType,
long... shape) |
INDArray |
NoOpMemoryMgr.allocate(boolean detached,
DataType dataType,
long... shape) |
INDArray |
CloseValidationMemoryMgr.allocate(boolean detached,
DataType dataType,
long... shape) |
INDArray |
ArrayCacheMemoryMgr.allocate(boolean detached,
LongShapeDescriptor descriptor) |
INDArray |
ArrayCloseMemoryMgr.allocate(boolean detached,
LongShapeDescriptor descriptor) |
INDArray |
NoOpMemoryMgr.allocate(boolean detached,
LongShapeDescriptor descriptor) |
INDArray |
CloseValidationMemoryMgr.allocate(boolean detached,
LongShapeDescriptor descriptor) |
INDArray |
AbstractMemoryMgr.dup(@NonNull INDArray arr) |
INDArray |
AbstractMemoryMgr.ulike(@NonNull INDArray arr) |
| Modifier and Type | Method and Description |
|---|---|
protected Set<INDArray> |
CloseValidationMemoryMgr.identitySetAllConstPhVar() |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
AbstractMemoryMgr.dup(@NonNull INDArray arr) |
void |
ArrayCacheMemoryMgr.release(@NonNull INDArray array) |
void |
ArrayCloseMemoryMgr.release(@NonNull INDArray array) |
void |
NoOpMemoryMgr.release(@NonNull INDArray array) |
void |
CloseValidationMemoryMgr.release(INDArray array) |
INDArray |
AbstractMemoryMgr.ulike(@NonNull INDArray arr) |
| Modifier and Type | Method and Description |
|---|---|
OptimizationHelper |
OptimizationHelper.arrayRecoveryFunction(String arrayName,
Supplier<INDArray> fn) |
| Modifier and Type | Method and Description |
|---|---|
static List<INDArray> |
TrainingUtils.getSingleOutput(List<Map<String,INDArray>> outputs,
String output)
Get a list of batch outputs for a single variable from a list of batch outputs for all variables
|
static List<INDArray> |
SameDiffUtils.getSingleOutput(List<Map<String,INDArray>> outputs,
String output)
Get a list of batch outputs for a single variable from a list of batch outputs for all variables
|
static Map<String,INDArray> |
TrainingUtils.stackOutputs(List<Map<String,INDArray>> outputs)
Stack batch outputs, like an output from
SameDiff.output(MultiDataSetIterator, String...) |
static Map<String,INDArray> |
SameDiffUtils.stackOutputs(List<Map<String,INDArray>> outputs)
Stack batch outputs, like an output from
SameDiff.output(MultiDataSetIterator, String...) |
| Modifier and Type | Method and Description |
|---|---|
static ExternalErrorsFunction |
SameDiffUtils.externalErrors(SameDiff sameDiff,
Map<String,INDArray> externalGradients,
SDVariable... inputs) |
static List<INDArray> |
TrainingUtils.getSingleOutput(List<Map<String,INDArray>> outputs,
String output)
Get a list of batch outputs for a single variable from a list of batch outputs for all variables
|
static List<INDArray> |
SameDiffUtils.getSingleOutput(List<Map<String,INDArray>> outputs,
String output)
Get a list of batch outputs for a single variable from a list of batch outputs for all variables
|
static Map<String,INDArray> |
TrainingUtils.stackOutputs(List<Map<String,INDArray>> outputs)
Stack batch outputs, like an output from
SameDiff.output(MultiDataSetIterator, String...) |
static Map<String,INDArray> |
SameDiffUtils.stackOutputs(List<Map<String,INDArray>> outputs)
Stack batch outputs, like an output from
SameDiff.output(MultiDataSetIterator, String...) |
| Modifier and Type | Method and Description |
|---|---|
Map<String,INDArray> |
TestCase.gradCheckMask() |
| Modifier and Type | Method and Description |
|---|---|
TestCase |
TestCase.expected(@NonNull SDVariable var,
@NonNull INDArray output)
Validate the output (forward pass) for a single variable using INDArray.equals(INDArray)
|
TestCase |
TestCase.expected(@NonNull String name,
@NonNull INDArray output)
Validate the output (forward pass) for a single variable using INDArray.equals(INDArray)
|
OpTestCase |
OpTestCase.expectedOutput(int outputNum,
INDArray expected)
Validate the op output using INDArray.equals(INDArray)
|
OpTestCase |
OpTestCase.expectedOutput(int outputNum,
INDArray expected,
double eps)
Validate the op output using INDArray.equals(INDArray)
|
TestCase |
TestCase.expectedOutput(@NonNull String name,
@NonNull INDArray expected)
Validate the output (forward pass) for a single variable using INDArray.equals(INDArray)
|
TestCase |
TestCase.expectedOutput(@NonNull String name,
@NonNull INDArray expected,
double eps)
Validate the output (forward pass) for a single variable using INDArray.equals(INDArray)
|
OpTestCase |
OpTestCase.expectedOutputRelError(int outputNum,
@NonNull INDArray expected,
double maxRelError,
double minAbsError)
Validate the output for a single variable using element-wise relative error:
relError = abs(x-y)/(abs(x)+abs(y)), with x=y=0 case defined to be 0.0.
|
TestCase |
TestCase.expectedOutputRelError(@NonNull String name,
@NonNull INDArray expected,
double maxRelError,
double minAbsError)
Validate the output (forward pass) for a single variable using element-wise relative error:
relError = abs(x-y)/(abs(x)+abs(y)), with x=y=0 case defined to be 0.0.
|
void |
ActivationGradientCheckListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs) |
TestCase |
TestCase.placeholderValue(String variable,
INDArray value) |
| Modifier and Type | Method and Description |
|---|---|
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
boolean print,
boolean exitOnFirstFailure) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure,
boolean skipValidation,
boolean debugMode,
Set<String> skipVariables,
Map<String,INDArray> gradCheckMask) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure,
boolean skipValidation,
boolean debugMode,
Set<String> skipVariables,
Map<String,INDArray> gradCheckMask) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure,
boolean skipValidation,
boolean debugMode,
Set<String> skipVariables,
Map<String,INDArray> gradCheckMask,
int maxPerParam,
GradCheckUtil.Subset subset) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure,
boolean skipValidation,
boolean debugMode,
Set<String> skipVariables,
Map<String,INDArray> gradCheckMask,
int maxPerParam,
GradCheckUtil.Subset subset) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
String... skipVariables) |
TestCase |
TestCase.expected(SDVariable var,
Function<INDArray,String> validationFn) |
TestCase |
TestCase.expected(String name,
Function<INDArray,String> validationFn) |
TestCase |
TestCase.placeholderValues(Map<String,INDArray> placeholderValues) |
| Modifier and Type | Method and Description |
|---|---|
String |
RelErrorFn.apply(INDArray actual) |
String |
EqualityFn.apply(INDArray actual) |
| Modifier and Type | Field and Description |
|---|---|
protected INDArray[] |
NonInplaceValidationListener.opInputs |
protected INDArray[] |
NonInplaceValidationListener.opInputsOrig |
| Modifier and Type | Method and Description |
|---|---|
void |
NonInplaceValidationListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
OpContext opContext,
INDArray[] outputs) |
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
EvaluationUtils.reshapeTimeSeriesMaskToVector(INDArray timeSeriesMask)
Reshape time series mask arrays.
|
static INDArray |
EvaluationUtils.reshapeTimeSeriesTo2d(INDArray labels) |
| Modifier and Type | Method and Description |
|---|---|
static Pair<INDArray,INDArray> |
EvaluationUtils.extractNonMaskedTimeSteps(INDArray labels,
INDArray predicted,
INDArray outputMask) |
static Pair<INDArray,INDArray> |
EvaluationUtils.extractNonMaskedTimeSteps(INDArray labels,
INDArray predicted,
INDArray outputMask) |
static Triple<INDArray,INDArray,INDArray> |
BaseEvaluation.reshapeAndExtractNotMasked(INDArray labels,
INDArray predictions,
INDArray mask,
int axis) |
static Triple<INDArray,INDArray,INDArray> |
BaseEvaluation.reshapeAndExtractNotMasked(INDArray labels,
INDArray predictions,
INDArray mask,
int axis) |
static Triple<INDArray,INDArray,INDArray> |
BaseEvaluation.reshapeAndExtractNotMasked(INDArray labels,
INDArray predictions,
INDArray mask,
int axis) |
| Modifier and Type | Method and Description |
|---|---|
void |
IEvaluation.eval(INDArray labels,
INDArray networkPredictions) |
void |
BaseEvaluation.eval(INDArray labels,
INDArray networkPredictions) |
void |
IEvaluation.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray) |
void |
BaseEvaluation.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray) |
void |
IEvaluation.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray,
List<? extends Serializable> recordMetaData) |
void |
IEvaluation.eval(INDArray labels,
INDArray networkPredictions,
List<? extends Serializable> recordMetaData) |
void |
BaseEvaluation.eval(@NonNull INDArray labels,
@NonNull INDArray predictions,
List<? extends Serializable> recordMetaData) |
void |
BaseEvaluation.eval(@NonNull INDArray labels,
@NonNull INDArray predictions,
List<? extends Serializable> recordMetaData) |
void |
IEvaluation.evalTimeSeries(INDArray labels,
INDArray predicted)
Deprecated.
|
void |
BaseEvaluation.evalTimeSeries(INDArray labels,
INDArray predicted) |
void |
IEvaluation.evalTimeSeries(INDArray labels,
INDArray predicted,
INDArray labelsMaskArray)
Deprecated.
|
void |
BaseEvaluation.evalTimeSeries(INDArray labels,
INDArray predictions,
INDArray labelsMask) |
static Pair<INDArray,INDArray> |
EvaluationUtils.extractNonMaskedTimeSteps(INDArray labels,
INDArray predicted,
INDArray outputMask) |
static Triple<INDArray,INDArray,INDArray> |
BaseEvaluation.reshapeAndExtractNotMasked(INDArray labels,
INDArray predictions,
INDArray mask,
int axis) |
static INDArray |
EvaluationUtils.reshapeTimeSeriesMaskToVector(INDArray timeSeriesMask)
Reshape time series mask arrays.
|
static INDArray |
EvaluationUtils.reshapeTimeSeriesTo2d(INDArray labels) |
| Modifier and Type | Field and Description |
|---|---|
protected INDArray |
Evaluation.costArray |
| Modifier and Type | Method and Description |
|---|---|
protected INDArray |
ROC.getProbAndLabelUsed() |
| Modifier and Type | Method and Description |
|---|---|
void |
Evaluation.eval(INDArray realOutcomes,
INDArray guesses)
Collects statistics on the real outcomes vs the
guesses.
|
void |
EvaluationCalibration.eval(INDArray labels,
INDArray networkPredictions) |
void |
EvaluationBinary.eval(INDArray labels,
INDArray networkPredictions) |
void |
EvaluationCalibration.eval(INDArray labels,
INDArray predictions,
INDArray mask) |
void |
EvaluationBinary.eval(INDArray labelsArr,
INDArray predictionsArr,
INDArray maskArr) |
void |
ROC.eval(INDArray labels,
INDArray predictions,
INDArray mask,
List<? extends Serializable> recordMetaData)
Evaluate (collect statistics for) the given minibatch of data.
|
void |
Evaluation.eval(INDArray labels,
INDArray predictions,
INDArray mask,
List<? extends Serializable> recordMetaData)
Evaluate the network, with optional metadata
|
void |
EvaluationCalibration.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray,
List<? extends Serializable> recordMetaData) |
void |
EvaluationBinary.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray,
List<? extends Serializable> recordMetaData) |
void |
ROCBinary.eval(INDArray labels,
INDArray predictions,
INDArray mask,
List<? extends Serializable> recordMetaData) |
void |
ROCMultiClass.eval(INDArray labels,
INDArray predictions,
INDArray mask,
List<? extends Serializable> recordMetaData)
Evaluate the network, with optional metadata
|
| Constructor and Description |
|---|
Evaluation(INDArray costArray)
Created evaluation instance with the specified cost array.
|
Evaluation(int axis,
Integer binaryPositiveClass,
int topN,
List<String> labelsList,
Double binaryDecisionThreshold,
INDArray costArray,
int maxWarningClassesToPrint) |
Evaluation(List<String> labels,
INDArray costArray)
Created evaluation instance with the specified cost array.
|
EvaluationBinary(INDArray decisionThreshold)
Create an EvaulationBinary instance with an optional decision threshold array.
|
EvaluationBinary(int axis,
ROCBinary rocBinary,
List<String> labels,
INDArray decisionThreshold) |
| Modifier and Type | Method and Description |
|---|---|
T |
EvaluationLambda.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray,
List<? extends Serializable> recordMetaData) |
void |
CustomEvaluation.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray,
List<? extends Serializable> recordMetaData) |
| Modifier and Type | Method and Description |
|---|---|
void |
RegressionEvaluation.eval(INDArray labels,
INDArray predictions) |
void |
RegressionEvaluation.eval(INDArray labelsArr,
INDArray predictionsArr,
INDArray maskArr) |
void |
RegressionEvaluation.eval(INDArray labels,
INDArray networkPredictions,
INDArray maskArray,
List<? extends Serializable> recordMetaData) |
| Modifier and Type | Method and Description |
|---|---|
long |
LogFileWriter.writeHistogramEventCustomBins(String name,
LogFileWriter.EventSubtype subtype,
long time,
int iteration,
int epoch,
INDArray bins,
INDArray y) |
long |
LogFileWriter.writeHistogramEventDiscrete(@NonNull String name,
LogFileWriter.EventSubtype subtype,
long time,
int iteration,
int epoch,
List<String> binLabels,
@NonNull INDArray y) |
long |
LogFileWriter.writeHistogramEventEqualSpacing(String name,
LogFileWriter.EventSubtype subtype,
long time,
int iteration,
int epoch,
double min,
double max,
INDArray y) |
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
TFGraphMapper.getArrayFrom(NodeDef nodeDef,
GraphDef graph)
Deprecated.
|
static INDArray |
TFGraphMapper.getNDArrayFromTensor(NodeDef node)
Convert the tensor to an NDArray (if possible and if array is available)
|
static INDArray |
TFGraphMapper.mapTensorProto(TensorProto tfTensor)
Convert a TensorProto to an INDArray
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
TFTensorMappers.BoolTensorMapper.arrayFor(long[] shape,
boolean[] jArr) |
INDArray |
TFTensorMappers.Float64TensorMapper.arrayFor(long[] shape,
double[] jArr) |
INDArray |
TFTensorMappers.Float16TensorMapper.arrayFor(long[] shape,
float[] jArr) |
INDArray |
TFTensorMappers.Float32TensorMapper.arrayFor(long[] shape,
float[] jArr) |
INDArray |
TFTensorMappers.BFloat16TensorMapper.arrayFor(long[] shape,
float[] jArr) |
INDArray |
TFTensorMappers.Int8TensorMapper.arrayFor(long[] shape,
int[] jArr) |
INDArray |
TFTensorMappers.Int16TensorMapper.arrayFor(long[] shape,
int[] jArr) |
INDArray |
TFTensorMappers.Int32TensorMapper.arrayFor(long[] shape,
int[] jArr) |
INDArray |
TFTensorMappers.UInt8TensorMapper.arrayFor(long[] shape,
int[] jArr) |
INDArray |
TFTensorMappers.UInt16TensorMapper.arrayFor(long[] shape,
int[] jArr) |
INDArray |
TFTensorMapper.arrayFor(long[] shape,
J jArr) |
INDArray |
TFTensorMappers.Int64TensorMapper.arrayFor(long[] shape,
long[] jArr) |
INDArray |
TFTensorMappers.UInt32TensorMapper.arrayFor(long[] shape,
long[] jArr) |
INDArray |
TFTensorMappers.UInt64TensorMapper.arrayFor(long[] shape,
long[] jArr) |
INDArray |
TFTensorMappers.StringTensorMapper.arrayFor(long[] shape,
String[] jArr) |
INDArray |
TFTensorMappers.BaseTensorMapper.toNDArray() |
INDArray |
TFTensorMapper.toNDArray() |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
IActivation.getActivation(INDArray in,
boolean training)
Carry out activation function on the input array (usually known as 'preOut' or 'z')
Implementations must overwrite "in", transform in place and return "in"
Can support separate behaviour during test
|
| Modifier and Type | Method and Description |
|---|---|
Pair<INDArray,INDArray> |
IActivation.backprop(INDArray in,
INDArray epsilon)
Backpropagate the errors through the activation function, given input z and epsilon dL/da.
Returns 2 INDArrays: (a) The gradient dL/dz, calculated from dL/da, and (b) The parameter gradients dL/dW, where w is the weights in the activation function. |
Pair<INDArray,INDArray> |
IActivation.backprop(INDArray in,
INDArray epsilon)
Backpropagate the errors through the activation function, given input z and epsilon dL/da.
Returns 2 INDArrays: (a) The gradient dL/dz, calculated from dL/da, and (b) The parameter gradients dL/dW, where w is the weights in the activation function. |
| Modifier and Type | Method and Description |
|---|---|
protected void |
BaseActivationFunction.assertShape(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
IActivation.backprop(INDArray in,
INDArray epsilon)
Backpropagate the errors through the activation function, given input z and epsilon dL/da.
Returns 2 INDArrays: (a) The gradient dL/dz, calculated from dL/da, and (b) The parameter gradients dL/dW, where w is the weights in the activation function. |
INDArray |
IActivation.getActivation(INDArray in,
boolean training)
Carry out activation function on the input array (usually known as 'preOut' or 'z')
Implementations must overwrite "in", transform in place and return "in"
Can support separate behaviour during test
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
ActivationSwish.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationCube.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSoftmax.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationHardTanH.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSigmoid.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationReLU6.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSoftSign.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationLReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationRReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationIdentity.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationRationalTanh.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationTanH.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationGELU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSELU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationHardSigmoid.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationELU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSoftPlus.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationRectifiedTanh.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationThresholdedReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationPReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationMish.getActivation(INDArray in,
boolean training) |
| Modifier and Type | Method and Description |
|---|---|
Pair<INDArray,INDArray> |
ActivationSwish.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationCube.backprop(@NonNull INDArray in,
@NonNull INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationCube.backprop(@NonNull INDArray in,
@NonNull INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationSoftmax.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationHardTanH.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationSigmoid.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationReLU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationReLU6.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationSoftSign.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationLReLU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationRReLU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationIdentity.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationRationalTanh.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationTanH.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationGELU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationSELU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationHardSigmoid.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationELU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationSoftPlus.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationRectifiedTanh.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationThresholdedReLU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationPReLU.backprop(INDArray in,
INDArray epsilon) |
Pair<INDArray,INDArray> |
ActivationMish.backprop(INDArray in,
INDArray epsilon) |
INDArray |
ActivationSwish.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationCube.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSoftmax.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationHardTanH.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSigmoid.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationReLU6.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSoftSign.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationLReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationRReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationIdentity.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationRationalTanh.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationTanH.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationGELU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSELU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationHardSigmoid.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationELU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationSoftPlus.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationRectifiedTanh.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationThresholdedReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationPReLU.getActivation(INDArray in,
boolean training) |
INDArray |
ActivationMish.getActivation(INDArray in,
boolean training) |
| Constructor and Description |
|---|
ActivationPReLU(INDArray alpha,
long[] sharedAxes) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
Lapack.getLFactor(INDArray A)
extracts the L (lower triangular) matrix from the LU factor result
L will be the same dimensions as A
|
INDArray |
Lapack.getPFactor(int M,
INDArray ipiv)
This method takes one of the ipiv returns from LAPACK and creates
the permutation matrix.
|
INDArray |
Lapack.getrf(INDArray A)
LU decomposiiton of a matrix
Factorize a matrix A
The matrix A is overridden by the L & U combined.
|
INDArray |
Lapack.getUFactor(INDArray A)
extracts the U (upper triangular) matrix from the LU factor result
U will be n x n matrix where n = num cols in A
|
| Modifier and Type | Method and Description |
|---|---|
double |
Level1.asum(INDArray arr)
computes the sum of magnitudes of all vector elements or, for a complex vector x, the sum
|
void |
Level1.axpy(long N,
double alpha,
INDArray x,
INDArray y)
computes a vector-scalar product and adds the result to a vector.
|
void |
Level1.copy(INDArray x,
INDArray y)
copy a vector to another vector.
|
double |
Level1.dot(long N,
double alpha,
INDArray X,
INDArray Y)
computes a vector-vector dot product.
|
void |
Level2.gbmv(char order,
char TransA,
int KL,
int KU,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
gbmv computes a matrix-vector product using a general band matrix and performs one of the following matrix-vector operations:
y := alpha*a*x + beta*y for trans = 'N'or'n';
y := alpha*a'*x + beta*y for trans = 'T'or't';
y := alpha*conjg(a')*x + beta*y for trans = 'C'or'c'.
|
void |
Level3.gemm(char Order,
char TransA,
char TransB,
double alpha,
INDArray A,
INDArray B,
double beta,
INDArray C)
gemm performs a matrix-matrix operation
c := alpha*op(a)*op(b) + beta*c,
where c is an m-by-n matrix,
op(a) is an m-by-k matrix,
op(b) is a k-by-n matrix.
|
void |
Level3.gemm(INDArray A,
INDArray B,
INDArray C,
boolean transposeA,
boolean transposeB,
double alpha,
double beta)
A convenience method for matrix-matrix operations with transposes.
|
void |
Level2.gemv(char order,
char transA,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
gemv computes a matrix-vector product using a general matrix and performs one of the following matrix-vector operations:
y := alpha*a*x + beta*y for trans = 'N'or'n';
y := alpha*a'*x + beta*y for trans = 'T'or't';
y := alpha*conjg(a')*x + beta*y for trans = 'C'or'c'.
|
void |
Lapack.geqrf(INDArray A,
INDArray R)
QR decomposiiton of a matrix
Factorize a matrix A such that A = QR
The matrix A is overwritten by the Q component (i.e.
|
void |
Level2.ger(char order,
double alpha,
INDArray X,
INDArray Y,
INDArray A)
performs a rank-1 update of a general m-by-n matrix a:
a := alpha*x*y' + a.
|
void |
Lapack.gesvd(INDArray A,
INDArray S,
INDArray U,
INDArray VT)
SVD decomposiiton of a matrix
Factorize a matrix into its singular vectors and eigenvalues
The decomposition is such that:
A = U x S x VT
gesvd = singular value decomposition (SVD) of a general matrix (GE)
|
static long |
BlasBufferUtil.getBlasOffset(INDArray arr)
Get blas stride for the
given array
|
static int |
BlasBufferUtil.getBlasStride(INDArray arr)
Get blas stride for the
given array
|
static char |
BlasBufferUtil.getCharForTranspose(INDArray arr)
Returns the proper character for
how to interpret a buffer (fortran being N C being T)
|
static long |
BlasBufferUtil.getDimension(INDArray arr,
boolean defaultRows)
Get the dimension associated with
the given ordering.
|
static double[] |
BlasBufferUtil.getDoubleData(INDArray buf)
Returns the double data
for this ndarray.
|
static float[] |
BlasBufferUtil.getFloatData(INDArray buf)
Returns the float data
for this ndarray.
|
static int |
BlasBufferUtil.getLd(INDArray arr)
Get the leading dimension
for a blas invocation.
|
INDArray |
Lapack.getLFactor(INDArray A)
extracts the L (lower triangular) matrix from the LU factor result
L will be the same dimensions as A
|
INDArray |
Lapack.getPFactor(int M,
INDArray ipiv)
This method takes one of the ipiv returns from LAPACK and creates
the permutation matrix.
|
INDArray |
Lapack.getrf(INDArray A)
LU decomposiiton of a matrix
Factorize a matrix A
The matrix A is overridden by the L & U combined.
|
void |
Lapack.getri(int N,
INDArray A,
int lda,
int[] IPIV,
INDArray WORK,
int lwork,
int INFO)
Generate inverse ggiven LU decomp
|
static int |
BlasBufferUtil.getStrideForOrdering(INDArray arr)
Return the proper stride
through a vector
relative to the ordering of the array
This is for incX/incY parameters in BLAS.
|
INDArray |
Lapack.getUFactor(INDArray A)
extracts the U (upper triangular) matrix from the LU factor result
U will be n x n matrix where n = num cols in A
|
int |
Level1.iamax(INDArray arr)
finds the element of a
vector that has the largest absolute value.
|
int |
Level1.iamax(long N,
INDArray arr,
int stride)
finds the element of a
vector that has the largest absolute value.
|
int |
Level1.iamin(INDArray arr)
finds the element of a vector that has the minimum absolute value.
|
double |
Level1.nrm2(INDArray arr)
computes the Euclidean norm of a vector.
|
void |
Lapack.potrf(INDArray A,
boolean lower)
Triangular decomposiiton of a positive definite matrix ( cholesky )
Factorize a matrix A such that A = LL* (assuming lower==true) or
A = U*U (a * represents conjugate i.e.
|
void |
Level1.rot(long N,
INDArray X,
INDArray Y,
double c,
double s)
performs rotation of points in the plane.
|
void |
Level1.rotg(INDArray a,
INDArray b,
INDArray c,
INDArray s)
computes parameters for a Givens rotation.
|
void |
Level1.rotmg(INDArray d1,
INDArray d2,
INDArray b1,
double b2,
INDArray P)
computes the modified parameters for a Givens rotation.
|
void |
Level2.sbmv(char order,
char Uplo,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
sbmv computes a matrix-vector product using a symmetric band matrix:
y := alpha*a*x + beta*y.
|
void |
Level1.scal(long N,
double alpha,
INDArray X)
computes a vector by a scalar product.
|
static void |
BlasBufferUtil.setData(double[] data,
INDArray toSet)
Set the data for the underlying array.
|
static void |
BlasBufferUtil.setData(float[] data,
INDArray toSet)
Set the data for the underlying array.
|
void |
Level2.spmv(char order,
char Uplo,
double alpha,
INDArray Ap,
INDArray X,
double beta,
INDArray Y) |
void |
Level2.spr(char order,
char Uplo,
double alpha,
INDArray X,
INDArray Ap)
spr performs a rank-1 update of an n-by-n packed symmetric matrix a:
a := alpha*x*x' + a.
|
void |
Level2.spr2(char order,
char Uplo,
double alpha,
INDArray X,
INDArray Y,
INDArray A)
?spr2 performs a rank-2 update of an n-by-n packed symmetric matrix a:
a := alpha*x*y' + alpha*y*x' + a.
|
void |
Level1.swap(INDArray x,
INDArray y)
swaps a vector with another vector.
|
int |
Lapack.syev(char jobz,
char uplo,
INDArray A,
INDArray V)
Caclulate the eigenvalues and vectors of a symmetric matrix.
|
void |
Level3.symm(char Order,
char Side,
char Uplo,
double alpha,
INDArray A,
INDArray B,
double beta,
INDArray C)
her2k performs a rank-2k update of an n-by-n Hermitian matrix c, that is, one of the following operations:
c := alpha*a*conjg(b') + conjg(alpha)*b*conjg(a') + beta*c, for trans = 'N'or'n'
c := alpha*conjg(b')*a + conjg(alpha)*conjg(a')*b + beta*c, for trans = 'C'or'c'
where c is an n-by-n Hermitian matrix;
a and b are n-by-k matrices if trans = 'N'or'n',
a and b are k-by-n matrices if trans = 'C'or'c'.
|
void |
Level2.symv(char order,
char Uplo,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
symv computes a matrix-vector product for a symmetric matrix:
y := alpha*a*x + beta*y.
|
void |
Level2.syr(char order,
char Uplo,
int N,
double alpha,
INDArray X,
INDArray A)
syr performs a rank-1 update of an n-by-n symmetric matrix a:
a := alpha*x*x' + a.
|
void |
Level2.syr2(char order,
char Uplo,
double alpha,
INDArray X,
INDArray Y,
INDArray A) |
void |
Level3.syr2k(char Order,
char Uplo,
char Trans,
double alpha,
INDArray A,
INDArray B,
double beta,
INDArray C)
yr2k performs a rank-2k update of an n-by-n symmetric matrix c, that is, one of the following operations:
c := alpha*a*b' + alpha*b*a' + beta*c for trans = 'N'or'n'
c := alpha*a'*b + alpha*b'*a + beta*c for trans = 'T'or't',
where c is an n-by-n symmetric matrix;
a and b are n-by-k matrices, if trans = 'N'or'n',
a and b are k-by-n matrices, if trans = 'T'or't'.
|
void |
Level3.syrk(char Order,
char Uplo,
char Trans,
double alpha,
INDArray A,
double beta,
INDArray C)
syrk performs a rank-n update of an n-by-n symmetric matrix c, that is, one of the following operations:
c := alpha*a*a' + beta*c for trans = 'N'or'n'
c := alpha*a'*a + beta*c for trans = 'T'or't','C'or'c',
where c is an n-by-n symmetric matrix;
a is an n-by-k matrix, if trans = 'N'or'n',
a is a k-by-n matrix, if trans = 'T'or't','C'or'c'.
|
void |
Level2.tbmv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
syr2 performs a rank-2 update of an n-by-n symmetric matrix a:
a := alpha*x*y' + alpha*y*x' + a.
|
void |
Level2.tbsv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
?tbsv solves a system of linear equations whose coefficients are in a triangular band matrix.
|
void |
Level2.tpmv(char order,
char Uplo,
char TransA,
char Diag,
INDArray Ap,
INDArray X)
tpmv computes a matrix-vector product using a triangular packed matrix.
|
void |
Level2.tpsv(char order,
char Uplo,
char TransA,
char Diag,
INDArray Ap,
INDArray X)
tpsv solves a system of linear equations whose coefficients are in a triangular packed matrix.
|
void |
Level3.trmm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
double alpha,
INDArray A,
INDArray B,
INDArray C)
syr2k performs a rank-2k update of an n-by-n symmetric matrix c, that is, one of the following operations:
c := alpha*a*b' + alpha*b*a' + beta*c for trans = 'N'or'n'
c := alpha*a'*b + alpha*b'*a + beta*c for trans = 'T'or't',
where c is an n-by-n symmetric matrix;
a and b are n-by-k matrices, if trans = 'N'or'n',
a and b are k-by-n matrices, if trans = 'T'or't'.
|
void |
Level2.trmv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
trmv computes a matrix-vector product using a triangular matrix.
|
void |
Level3.trsm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
double alpha,
INDArray A,
INDArray B)
?trsm solves one of the following matrix equations:
op(a)*x = alpha*b or x*op(a) = alpha*b,
where x and b are m-by-n general matrices, and a is triangular;
op(a) must be an m-by-m matrix, if side = 'L'or'l'
op(a) must be an n-by-n matrix, if side = 'R'or'r'.
|
void |
Level2.trsv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
trsv solves a system of linear equations whose coefficients are in a triangular matrix.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BaseLapack.getLFactor(INDArray A) |
INDArray |
BaseLapack.getPFactor(int M,
INDArray ipiv) |
INDArray |
BaseLapack.getrf(INDArray A) |
INDArray |
BaseLapack.getUFactor(INDArray A) |
| Modifier and Type | Method and Description |
|---|---|
double |
BaseLevel1.asum(INDArray arr)
computes the sum of magnitudes of all vector elements or, for a complex vector x, the sum
|
void |
BaseLevel1.axpy(long n,
double alpha,
INDArray x,
INDArray y)
computes a vector-scalar product and adds the result to a vector.
|
void |
BaseLevel1.copy(INDArray x,
INDArray y)
swaps a vector with another vector.
|
protected abstract double |
BaseLevel1.dasum(long N,
INDArray X,
int incX) |
protected abstract void |
BaseLevel1.daxpy(long N,
double alpha,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel1.dcopy(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract double |
BaseLevel1.ddot(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel2.dgbmv(char order,
char TransA,
int M,
int N,
int KL,
int KU,
double alpha,
INDArray A,
int lda,
INDArray X,
int incX,
double beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel3.dgemm(char Order,
char TransA,
char TransB,
int M,
int N,
int K,
double alpha,
INDArray A,
int lda,
INDArray B,
int ldb,
double beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel2.dgemv(char order,
char TransA,
int M,
int N,
double alpha,
INDArray A,
int lda,
INDArray X,
int incX,
double beta,
INDArray Y,
int incY) |
abstract void |
BaseLapack.dgeqrf(int M,
int N,
INDArray A,
INDArray R,
INDArray INFO) |
protected abstract void |
BaseLevel2.dger(char order,
int M,
int N,
double alpha,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray A,
int lda) |
abstract void |
BaseLapack.dgesvd(byte jobu,
byte jobvt,
int M,
int N,
INDArray A,
INDArray S,
INDArray U,
INDArray VT,
INDArray INFO) |
abstract void |
BaseLapack.dgetrf(int M,
int N,
INDArray A,
INDArray IPIV,
INDArray INFO) |
protected abstract double |
BaseLevel1.dnrm2(long N,
INDArray X,
int incX) |
double |
BaseLevel1.dot(long n,
double alpha,
INDArray X,
INDArray Y)
computes a vector-vector dot product.
|
abstract void |
BaseLapack.dpotrf(byte uplo,
int N,
INDArray A,
INDArray INFO) |
protected abstract void |
BaseLevel1.drot(long N,
INDArray X,
int incX,
INDArray Y,
int incY,
double c,
double s) |
protected abstract void |
BaseLevel1.drotm(long N,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray P) |
protected abstract void |
BaseLevel1.drotmg(double d1,
double d2,
double b1,
double b2,
INDArray P) |
protected abstract void |
BaseLevel2.dsbmv(char order,
char Uplo,
int N,
int K,
double alpha,
INDArray A,
int lda,
INDArray X,
int incX,
double beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel1.dscal(long N,
double alpha,
INDArray X,
int incX) |
protected abstract double |
BaseLevel1.dsdot(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel2.dspmv(char order,
char Uplo,
int N,
double alpha,
INDArray Ap,
INDArray X,
int incX,
double beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel2.dspr(char order,
char Uplo,
int N,
double alpha,
INDArray X,
int incX,
INDArray Ap) |
protected abstract void |
BaseLevel2.dspr2(char order,
char Uplo,
int N,
double alpha,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray A) |
protected abstract void |
BaseLevel1.dswap(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
abstract int |
BaseLapack.dsyev(char jobz,
char uplo,
int N,
INDArray A,
INDArray R) |
protected abstract void |
BaseLevel3.dsymm(char Order,
char Side,
char Uplo,
int M,
int N,
double alpha,
INDArray A,
int lda,
INDArray B,
int ldb,
double beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel2.dsymv(char order,
char Uplo,
int N,
double alpha,
INDArray A,
int lda,
INDArray X,
int incX,
double beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel2.dsyr(char order,
char Uplo,
int N,
double alpha,
INDArray X,
int incX,
INDArray A,
int lda) |
protected abstract void |
BaseLevel2.dsyr2(char order,
char Uplo,
int N,
double alpha,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray A,
int lda) |
protected abstract void |
BaseLevel3.dsyr2k(char Order,
char Uplo,
char Trans,
int N,
int K,
double alpha,
INDArray A,
int lda,
INDArray B,
int ldb,
double beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel3.dsyrk(char Order,
char Uplo,
char Trans,
int N,
int K,
double alpha,
INDArray A,
int lda,
double beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel2.dtbmv(char order,
char Uplo,
char TransA,
char Diag,
int N,
int K,
INDArray A,
int lda,
INDArray X,
int incX) |
protected abstract void |
BaseLevel2.dtbsv(char order,
char Uplo,
char TransA,
char Diag,
int N,
int K,
INDArray A,
int lda,
INDArray X,
int incX) |
protected abstract void |
BaseLevel2.dtpmv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray Ap,
INDArray X,
int incX) |
protected abstract void |
BaseLevel2.dtpsv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray Ap,
INDArray X,
int incX) |
protected abstract void |
BaseLevel3.dtrmm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
int M,
int N,
double alpha,
INDArray A,
int lda,
INDArray B,
int ldb) |
protected abstract void |
BaseLevel2.dtrmv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray A,
int lda,
INDArray X,
int incX) |
protected abstract void |
BaseLevel3.dtrsm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
int M,
int N,
double alpha,
INDArray A,
int lda,
INDArray B,
int ldb) |
protected abstract void |
BaseLevel2.dtrsv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray A,
int lda,
INDArray X,
int incX) |
void |
BaseLevel2.gbmv(char order,
char TransA,
int KL,
int KU,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
gbmv computes a matrix-vector product using a general band matrix and performs one of the following matrix-vector operations:
y := alpha*a*x + beta*y for trans = 'N'or'n';
y := alpha*a'*x + beta*y for trans = 'T'or't';
y := alpha*conjg(a')*x + beta*y for trans = 'C'or'c'.
|
void |
BaseLevel3.gemm(char Order,
char TransA,
char TransB,
double alpha,
INDArray A,
INDArray B,
double beta,
INDArray C)
gemm performs a matrix-matrix operation
c := alpha*op(a)*op(b) + beta*c,
where c is an m-by-n matrix,
op(a) is an m-by-k matrix,
op(b) is a k-by-n matrix.
|
void |
BaseLevel3.gemm(INDArray A,
INDArray B,
INDArray C,
boolean transposeA,
boolean transposeB,
double alpha,
double beta)
A convenience method for matrix-matrix operations with transposes.
|
void |
BaseLevel2.gemv(char order,
char transA,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
gemv computes a matrix-vector product using a general matrix and performs one of the following matrix-vector operations:
y := alpha*a*x + beta*y for trans = 'N'or'n';
y := alpha*a'*x + beta*y for trans = 'T'or't';
y := alpha*conjg(a')*x + beta*y for trans = 'C'or'c'.
|
void |
BaseLapack.geqrf(INDArray A,
INDArray R) |
void |
BaseLevel2.ger(char order,
double alpha,
INDArray X,
INDArray Y,
INDArray A)
performs a rank-1 update of a general m-by-n matrix a:
a := alpha*x*y' + a.
|
void |
BaseLapack.gesvd(INDArray A,
INDArray S,
INDArray U,
INDArray VT) |
INDArray |
BaseLapack.getLFactor(INDArray A) |
INDArray |
BaseLapack.getPFactor(int M,
INDArray ipiv) |
INDArray |
BaseLapack.getrf(INDArray A) |
INDArray |
BaseLapack.getUFactor(INDArray A) |
protected abstract float |
BaseLevel1.hasum(long N,
INDArray X,
int incX) |
protected abstract void |
BaseLevel1.haxpy(long N,
float alpha,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract float |
BaseLevel1.hdot(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel3.hgemm(char Order,
char TransA,
char TransB,
int M,
int N,
int K,
float alpha,
INDArray A,
int lda,
INDArray B,
int ldb,
float beta,
INDArray C,
int ldc) |
int |
BaseLevel1.iamax(INDArray arr)
finds the element of a
vector that has the largest absolute value.
|
int |
BaseLevel1.iamax(long n,
INDArray arr,
int stride) |
int |
BaseLevel1.iamin(INDArray arr)
finds the element of a vector that has the minimum absolute value.
|
protected abstract int |
BaseLevel1.idamax(long N,
INDArray X,
int incX) |
protected abstract int |
BaseLevel1.isamax(long N,
INDArray X,
int incX) |
double |
BaseLevel1.nrm2(INDArray arr)
computes the Euclidean norm of a vector.
|
void |
BaseLapack.potrf(INDArray A,
boolean lower) |
void |
BaseLevel1.rot(long N,
INDArray X,
INDArray Y,
double c,
double s)
performs rotation of points in the plane.
|
void |
BaseLevel1.rotg(INDArray a,
INDArray b,
INDArray c,
INDArray s)
computes parameters for a Givens rotation.
|
void |
BaseLevel1.rotmg(INDArray d1,
INDArray d2,
INDArray b1,
double b2,
INDArray P)
computes the modified parameters for a Givens rotation.
|
protected abstract float |
BaseLevel1.sasum(long N,
INDArray X,
int incX) |
protected abstract void |
BaseLevel1.saxpy(long N,
float alpha,
INDArray X,
int incX,
INDArray Y,
int incY) |
void |
BaseLevel2.sbmv(char order,
char Uplo,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
sbmv computes a matrix-vector product using a symmetric band matrix:
y := alpha*a*x + beta*y.
|
void |
BaseLevel1.scal(long N,
double alpha,
INDArray X)
computes a vector by a scalar product.
|
protected abstract void |
BaseLevel1.scopy(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract float |
BaseLevel1.sdot(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract float |
BaseLevel1.sdsdot(long N,
float alpha,
INDArray X,
int incX,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel2.sgbmv(char order,
char TransA,
int M,
int N,
int KL,
int KU,
float alpha,
INDArray A,
int lda,
INDArray X,
int incX,
float beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel3.sgemm(char Order,
char TransA,
char TransB,
int M,
int N,
int K,
float alpha,
INDArray A,
int lda,
INDArray B,
int ldb,
float beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel2.sgemv(char order,
char TransA,
int M,
int N,
float alpha,
INDArray A,
int lda,
INDArray X,
int incX,
float beta,
INDArray Y,
int incY) |
abstract void |
BaseLapack.sgeqrf(int M,
int N,
INDArray A,
INDArray R,
INDArray INFO)
Float/Double versions of QR decomp.
|
protected abstract void |
BaseLevel2.sger(char order,
int M,
int N,
float alpha,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray A,
int lda) |
abstract void |
BaseLapack.sgesvd(byte jobu,
byte jobvt,
int M,
int N,
INDArray A,
INDArray S,
INDArray U,
INDArray VT,
INDArray INFO) |
abstract void |
BaseLapack.sgetrf(int M,
int N,
INDArray A,
INDArray IPIV,
INDArray INFO)
Float/Double versions of LU decomp.
|
protected abstract float |
BaseLevel1.snrm2(long N,
INDArray X,
int incX) |
void |
BaseLevel2.spmv(char order,
char Uplo,
double alpha,
INDArray Ap,
INDArray X,
double beta,
INDArray Y) |
abstract void |
BaseLapack.spotrf(byte uplo,
int N,
INDArray A,
INDArray INFO)
Float/Double versions of cholesky decomp for positive definite matrices
A = LL*
|
void |
BaseLevel2.spr(char order,
char Uplo,
double alpha,
INDArray X,
INDArray Ap)
spr performs a rank-1 update of an n-by-n packed symmetric matrix a:
a := alpha*x*x' + a.
|
void |
BaseLevel2.spr2(char order,
char Uplo,
double alpha,
INDArray X,
INDArray Y,
INDArray A)
?spr2 performs a rank-2 update of an n-by-n packed symmetric matrix a:
a := alpha*x*y' + alpha*y*x' + a.
|
protected abstract void |
BaseLevel1.srot(long N,
INDArray X,
int incX,
INDArray Y,
int incY,
float c,
float s) |
protected abstract void |
BaseLevel1.srotm(long N,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray P) |
protected abstract void |
BaseLevel1.srotmg(float d1,
float d2,
float b1,
float b2,
INDArray P) |
protected abstract void |
BaseLevel2.ssbmv(char order,
char Uplo,
int N,
int K,
float alpha,
INDArray A,
int lda,
INDArray X,
int incX,
float beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel1.sscal(long N,
float alpha,
INDArray X,
int incX) |
protected abstract void |
BaseLevel2.sspmv(char order,
char Uplo,
int N,
float alpha,
INDArray Ap,
INDArray X,
int incX,
float beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel2.sspr(char order,
char Uplo,
int N,
float alpha,
INDArray X,
int incX,
INDArray Ap) |
protected abstract void |
BaseLevel2.sspr2(char order,
char Uplo,
int N,
float alpha,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray A) |
protected abstract void |
BaseLevel1.sswap(long N,
INDArray X,
int incX,
INDArray Y,
int incY) |
abstract int |
BaseLapack.ssyev(char jobz,
char uplo,
int N,
INDArray A,
INDArray R)
Float/Double versions of eigen value/vector calc.
|
protected abstract void |
BaseLevel3.ssymm(char Order,
char Side,
char Uplo,
int M,
int N,
float alpha,
INDArray A,
int lda,
INDArray B,
int ldb,
float beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel2.ssymv(char order,
char Uplo,
int N,
float alpha,
INDArray A,
int lda,
INDArray X,
int incX,
float beta,
INDArray Y,
int incY) |
protected abstract void |
BaseLevel2.ssyr(char order,
char Uplo,
int N,
float alpha,
INDArray X,
int incX,
INDArray A,
int lda) |
protected abstract void |
BaseLevel2.ssyr2(char order,
char Uplo,
int N,
float alpha,
INDArray X,
int incX,
INDArray Y,
int incY,
INDArray A,
int lda) |
protected abstract void |
BaseLevel3.ssyr2k(char Order,
char Uplo,
char Trans,
int N,
int K,
float alpha,
INDArray A,
int lda,
INDArray B,
int ldb,
float beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel3.ssyrk(char Order,
char Uplo,
char Trans,
int N,
int K,
float alpha,
INDArray A,
int lda,
float beta,
INDArray C,
int ldc) |
protected abstract void |
BaseLevel2.stbmv(char order,
char Uplo,
char TransA,
char Diag,
int N,
int K,
INDArray A,
int lda,
INDArray X,
int incX) |
protected abstract void |
BaseLevel2.stbsv(char order,
char Uplo,
char TransA,
char Diag,
int N,
int K,
INDArray A,
int lda,
INDArray X,
int incX) |
protected abstract void |
BaseLevel2.stpmv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray Ap,
INDArray X,
int incX) |
protected abstract void |
BaseLevel2.stpsv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray Ap,
INDArray X,
int incX) |
protected abstract void |
BaseLevel3.strmm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
int M,
int N,
float alpha,
INDArray A,
int lda,
INDArray B,
int ldb) |
protected abstract void |
BaseLevel2.strmv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray A,
int lda,
INDArray X,
int incX) |
protected abstract void |
BaseLevel3.strsm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
int M,
int N,
float alpha,
INDArray A,
int lda,
INDArray B,
int ldb) |
protected abstract void |
BaseLevel2.strsv(char order,
char Uplo,
char TransA,
char Diag,
int N,
INDArray A,
int lda,
INDArray X,
int incX) |
void |
BaseLevel1.swap(INDArray x,
INDArray y)
swaps a vector with another vector.
|
int |
BaseLapack.syev(char jobz,
char uplo,
INDArray A,
INDArray V) |
void |
BaseLevel3.symm(char Order,
char Side,
char Uplo,
double alpha,
INDArray A,
INDArray B,
double beta,
INDArray C)
her2k performs a rank-2k update of an n-by-n Hermitian matrix c, that is, one of the following operations:
c := alpha*a*conjg(b') + conjg(alpha)*b*conjg(a') + beta*c, for trans = 'N'or'n'
c := alpha*conjg(b')*a + conjg(alpha)*conjg(a')*b + beta*c, for trans = 'C'or'c'
where c is an n-by-n Hermitian matrix;
a and b are n-by-k matrices if trans = 'N'or'n',
a and b are k-by-n matrices if trans = 'C'or'c'.
|
void |
BaseLevel2.symv(char order,
char Uplo,
double alpha,
INDArray A,
INDArray X,
double beta,
INDArray Y)
symv computes a matrix-vector product for a symmetric matrix:
y := alpha*a*x + beta*y.
|
void |
BaseLevel2.syr(char order,
char Uplo,
int N,
double alpha,
INDArray X,
INDArray A)
syr performs a rank-1 update of an n-by-n symmetric matrix a:
a := alpha*x*x' + a.
|
void |
BaseLevel2.syr2(char order,
char Uplo,
double alpha,
INDArray X,
INDArray Y,
INDArray A) |
void |
BaseLevel3.syr2k(char Order,
char Uplo,
char Trans,
double alpha,
INDArray A,
INDArray B,
double beta,
INDArray C)
yr2k performs a rank-2k update of an n-by-n symmetric matrix c, that is, one of the following operations:
c := alpha*a*b' + alpha*b*a' + beta*c for trans = 'N'or'n'
c := alpha*a'*b + alpha*b'*a + beta*c for trans = 'T'or't',
where c is an n-by-n symmetric matrix;
a and b are n-by-k matrices, if trans = 'N'or'n',
a and b are k-by-n matrices, if trans = 'T'or't'.
|
void |
BaseLevel3.syrk(char Order,
char Uplo,
char Trans,
double alpha,
INDArray A,
double beta,
INDArray C)
syrk performs a rank-n update of an n-by-n symmetric matrix c, that is, one of the following operations:
c := alpha*a*a' + beta*c for trans = 'N'or'n'
c := alpha*a'*a + beta*c for trans = 'T'or't','C'or'c',
where c is an n-by-n symmetric matrix;
a is an n-by-k matrix, if trans = 'N'or'n',
a is a k-by-n matrix, if trans = 'T'or't','C'or'c'.
|
void |
BaseLevel2.tbmv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
syr2 performs a rank-2 update of an n-by-n symmetric matrix a:
a := alpha*x*y' + alpha*y*x' + a.
|
void |
BaseLevel2.tbsv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
?tbsv solves a system of linear equations whose coefficients are in a triangular band matrix.
|
void |
BaseLevel2.tpmv(char order,
char Uplo,
char TransA,
char Diag,
INDArray Ap,
INDArray X)
tpmv computes a matrix-vector product using a triangular packed matrix.
|
void |
BaseLevel2.tpsv(char order,
char Uplo,
char TransA,
char Diag,
INDArray Ap,
INDArray X)
tpsv solves a system of linear equations whose coefficients are in a triangular packed matrix.
|
void |
BaseLevel3.trmm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
double alpha,
INDArray A,
INDArray B,
INDArray C)
syr2k performs a rank-2k update of an n-by-n symmetric matrix c, that is, one of the following operations:
c := alpha*a*b' + alpha*b*a' + beta*c for trans = 'N'or'n'
c := alpha*a'*b + alpha*b'*a + beta*c for trans = 'T'or't',
where c is an n-by-n symmetric matrix;
a and b are n-by-k matrices, if trans = 'N'or'n',
a and b are k-by-n matrices, if trans = 'T'or't'.
|
void |
BaseLevel2.trmv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
trmv computes a matrix-vector product using a triangular matrix.
|
void |
BaseLevel3.trsm(char Order,
char Side,
char Uplo,
char TransA,
char Diag,
double alpha,
INDArray A,
INDArray B)
?trsm solves one of the following matrix equations:
op(a)*x = alpha*b or x*op(a) = alpha*b,
where x and b are m-by-n general matrices, and a is triangular;
op(a) must be an m-by-m matrix, if side = 'L'or'l'
op(a) must be an n-by-n matrix, if side = 'R'or'r'.
|
void |
BaseLevel2.trsv(char order,
char Uplo,
char TransA,
char Diag,
INDArray A,
INDArray X)
trsv solves a system of linear equations whose coefficients are in a triangular matrix.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
MMulTranspose.exec(INDArray a,
INDArray b,
INDArray result)
Execute the matrix multiplication: A x B
Note that if a or b have transposeA/B == true, then this is done internally.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
MMulTranspose.exec(INDArray a,
INDArray b,
INDArray result)
Execute the matrix multiplication: A x B
Note that if a or b have transposeA/B == true, then this is done internally.
|
| Constructor and Description |
|---|
GemmParams(INDArray a,
INDArray b,
INDArray c) |
GemmParams(INDArray a,
INDArray b,
INDArray c,
boolean transposeA,
boolean transposeB) |
GemvParameters(INDArray a,
INDArray x,
INDArray y) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
DistributedINDArray.entry()
This method returns INDArray for the current device
PLEASE NOTE: if you use more than one thread per device you'd better not use this method unless you're 100% sure
|
INDArray |
DistributedINDArray.entry(int entry)
This method returns INDArray for specific entry (i.e.
|
INDArray |
AffinityManager.replicateToDevice(Integer deviceId,
INDArray array)
This method replicates given INDArray, and places it to target device.
|
INDArray |
BasicAffinityManager.replicateToDevice(Integer deviceId,
INDArray array)
This method replicates given INDArray, and places it to target device.
|
| Modifier and Type | Method and Description |
|---|---|
void |
AffinityManager.ensureLocation(INDArray array,
AffinityManager.Location location)
This method propagates given INDArray to specified location
|
void |
BasicAffinityManager.ensureLocation(INDArray array,
AffinityManager.Location location) |
AffinityManager.Location |
AffinityManager.getActiveLocation(INDArray array)
This method returns last-updated location for the given INDArray
|
AffinityManager.Location |
BasicAffinityManager.getActiveLocation(INDArray array) |
Integer |
AffinityManager.getDeviceForArray(INDArray array)
This method returns id of current device for a given INDArray
|
Integer |
BasicAffinityManager.getDeviceForArray(INDArray array) |
void |
DistributedINDArray.propagate(INDArray array)
This method propagates given INDArray to all entries as is
|
INDArray |
AffinityManager.replicateToDevice(Integer deviceId,
INDArray array)
This method replicates given INDArray, and places it to target device.
|
INDArray |
BasicAffinityManager.replicateToDevice(Integer deviceId,
INDArray array)
This method replicates given INDArray, and places it to target device.
|
void |
AffinityManager.tagLocation(INDArray array,
AffinityManager.Location location)
This method tags specific INDArray as "recent" on specified location
|
void |
BasicAffinityManager.tagLocation(INDArray array,
AffinityManager.Location location) |
void |
AffinityManager.touch(INDArray array)
Utility method, to associate INDArray with specific device (backend-specific)
|
| Constructor and Description |
|---|
FirstAxisIterator(INDArray iterateOver) |
INDArrayIterator(INDArray iterateOver) |
| Modifier and Type | Method and Description |
|---|---|
void |
MemoryManager.collect(INDArray... arrays)
This method detaches off-heap memory from passed INDArray instances, and optionally stores them in cache for future reuse
PLEASE NOTE: Cache options depend on specific implementations
|
void |
BasicMemoryManager.collect(INDArray... arrays)
This method detaches off-heap memory from passed INDArray instances, and optionally stores them in cache for future reuse
PLEASE NOTE: Cache options depend on specific implementations
|
void |
MemoryManager.memset(INDArray array)
This method fills given INDArray with zeroes.
|
| Modifier and Type | Field and Description |
|---|---|
protected Map<T,INDArray> |
BasicStash.stash |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BasicStash.get(T key) |
INDArray |
Stash.get(T key) |
| Modifier and Type | Method and Description |
|---|---|
void |
BasicStash.put(T key,
INDArray object) |
void |
Stash.put(T key,
INDArray object) |
| Modifier and Type | Class and Description |
|---|---|
class |
BaseNDArray |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
INDArray.add(INDArray other)
Element-wise copy addition of two NDArrays
|
INDArray |
BaseNDArray.add(INDArray other) |
INDArray |
INDArray.add(INDArray other,
INDArray result)
Element-wise copy addition of two NDArrays
|
INDArray |
BaseNDArray.add(INDArray other,
INDArray result) |
INDArray |
INDArray.add(Number n)
Scalar addition (cloning)
|
INDArray |
BaseNDArray.add(Number n) |
INDArray |
INDArray.add(Number n,
INDArray result)
Addition of this ndarray.
|
INDArray |
BaseNDArray.add(Number n,
INDArray result) |
INDArray |
INDArray.addColumnVector(INDArray columnVector)
Addition of a column vector (copy)
|
INDArray |
BaseNDArray.addColumnVector(INDArray columnVector) |
INDArray |
INDArray.addi(INDArray other)
in place (element wise) addition of two NDArrays
|
INDArray |
BaseNDArray.addi(INDArray other) |
INDArray |
INDArray.addi(INDArray other,
INDArray result)
in place (element wise) addition of two NDArrays
|
INDArray |
BaseNDArray.addi(INDArray other,
INDArray result) |
INDArray |
INDArray.addi(Number n)
In place scalar addition
|
INDArray |
BaseNDArray.addi(Number n) |
INDArray |
INDArray.addi(Number n,
INDArray result)
In place addition
|
INDArray |
BaseNDArray.addi(Number n,
INDArray result) |
INDArray |
INDArray.addiColumnVector(INDArray columnVector)
In place addition of a column vector
|
INDArray |
BaseNDArray.addiColumnVector(INDArray columnVector) |
INDArray |
INDArray.addiRowVector(INDArray rowVector)
In place addition of a row vector
|
INDArray |
BaseNDArray.addiRowVector(INDArray rowVector) |
INDArray |
INDArray.addRowVector(INDArray rowVector)
Addition of a row vector (copy)
|
INDArray |
BaseNDArray.addRowVector(INDArray rowVector) |
INDArray |
INDArray.amax(int... dimension)
Returns the absolute overall max of this ndarray along given dimensions
|
INDArray |
BaseNDArray.amax(int... dimension) |
INDArray |
INDArray.amean(int... dimension)
Returns the absolute overall mean of this ndarray
|
INDArray |
BaseNDArray.amean(int... dimension) |
INDArray |
INDArray.amin(int... dimension)
Returns minimum (absolute) value in this INDArray, along the specified dimensions
|
INDArray |
BaseNDArray.amin(int... dimension) |
INDArray |
INDArray.argMax(int... dimension)
This method returns index of highest value along specified dimension(s)
|
INDArray |
BaseNDArray.argMax(int... dimension) |
INDArray |
INDArray.assign(boolean value)
Set all entries of the ndarray to the specified value
|
INDArray |
BaseNDArray.assign(boolean value) |
INDArray |
INDArray.assign(INDArray arr)
Assign all of the elements in the given ndarray to this ndarray
|
INDArray |
BaseNDArray.assign(INDArray arr) |
INDArray |
INDArray.assign(Number value)
Set all entries of the ndarray to the specified value
|
INDArray |
BaseNDArray.assign(Number value) |
INDArray |
INDArray.assignIf(INDArray arr,
Condition condition)
Assign all elements from given ndarray that are matching given condition,
ndarray to this ndarray
|
INDArray |
BaseNDArray.assignIf(INDArray arr,
Condition condition) |
INDArray |
INDArray.broadcast(INDArray result)
Broadcasts this ndarray to be the specified shape
|
INDArray |
BaseNDArray.broadcast(INDArray result) |
INDArray |
INDArray.broadcast(long... shape)
Broadcasts this ndarray to be the specified shape
|
INDArray |
BaseNDArray.broadcast(long... shape) |
INDArray |
INDArray.castTo(DataType dataType)
This method cast elements of this INDArray to new data type
|
INDArray |
BaseNDArray.castTo(DataType dataType) |
INDArray |
INDArray.cond(Condition condition)
Returns a binary INDArray with value 'true' if the element matches the specified condition and 'false' otherwise
|
INDArray |
BaseNDArray.cond(Condition condition) |
protected INDArray |
BaseNDArray.create(DataBuffer buffer) |
protected INDArray |
BaseNDArray.create(DataBuffer data,
int[] shape,
int[] strides) |
protected INDArray |
BaseNDArray.create(DataBuffer data,
int[] newShape,
int[] newStrides,
long offset) |
protected INDArray |
BaseNDArray.create(DataBuffer data,
int[] newShape,
int[] newStrides,
long offset,
char ordering) |
protected INDArray |
BaseNDArray.create(DataBuffer data,
int[] shape,
long offset) |
protected INDArray |
BaseNDArray.create(DataBuffer data,
long[] newShape,
long[] newStrides,
long offset,
char ordering) |
protected INDArray |
BaseNDArray.create(int[] shape) |
protected INDArray |
BaseNDArray.create(int[] shape,
char ordering) |
protected INDArray |
BaseNDArray.create(int[] shape,
int[] strides,
long offset) |
protected INDArray |
BaseNDArray.create(int rows,
int length) |
protected INDArray |
BaseNDArray.createScalar(double d) |
protected INDArray |
BaseNDArray.createScalarForIndex(long i,
boolean applyOffset) |
INDArray |
INDArray.cumsum(int dimension)
Returns the cumulative sum along a dimension.
|
INDArray |
BaseNDArray.cumsum(int dimension) |
INDArray |
INDArray.cumsumi(int dimension)
Returns the cumulative sum along a dimension.
|
INDArray |
BaseNDArray.cumsumi(int dimension) |
INDArray |
INDArray.detach()
This method detaches INDArray from Workspace, returning copy.
|
INDArray |
BaseNDArray.detach() |
INDArray |
INDArray.dimShuffle(Object[] rearrange,
int[] newOrder,
boolean[] broadCastable)
Dimshuffle: an extension of permute that adds the ability
to broadcast various dimensions.
|
INDArray |
BaseNDArray.dimShuffle(Object[] rearrange,
int[] newOrder,
boolean[] broadCastable)
Deprecated.
|
INDArray |
INDArray.dimShuffle(Object[] rearrange,
long[] newOrder,
boolean[] broadCastable)
See {@link #dimShuffle(Object[], int[], boolean[])
|
INDArray |
BaseNDArray.dimShuffle(Object[] rearrange,
long[] newOrder,
boolean[] broadCastable)
Dimshuffle: an extension of permute that adds the ability
to broadcast various dimensions.
|
INDArray |
INDArray.div(INDArray other)
Copy (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.div(INDArray other) |
INDArray |
INDArray.div(INDArray other,
INDArray result)
copy (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.div(INDArray other,
INDArray result) |
INDArray |
INDArray.div(Number n)
Division by a number
|
INDArray |
BaseNDArray.div(Number n) |
INDArray |
INDArray.div(Number n,
INDArray result)
Division if ndarray by number
|
INDArray |
BaseNDArray.div(Number n,
INDArray result) |
INDArray |
INDArray.divColumnVector(INDArray columnVector)
Division of a column vector (copy)
|
INDArray |
BaseNDArray.divColumnVector(INDArray columnVector) |
INDArray |
INDArray.divi(INDArray other)
in place (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.divi(INDArray other) |
INDArray |
INDArray.divi(INDArray other,
INDArray result)
in place (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.divi(INDArray other,
INDArray result) |
INDArray |
INDArray.divi(Number n)
In place scalar division
|
INDArray |
BaseNDArray.divi(Number n) |
INDArray |
INDArray.divi(Number n,
INDArray result)
In place division of this ndarray
|
INDArray |
BaseNDArray.divi(Number n,
INDArray result) |
INDArray |
INDArray.diviColumnVector(INDArray columnVector)
In place division of a column vector
|
INDArray |
BaseNDArray.diviColumnVector(INDArray columnVector) |
INDArray |
INDArray.diviRowVector(INDArray rowVector)
In place division of a row vector
|
INDArray |
BaseNDArray.diviRowVector(INDArray rowVector) |
INDArray |
INDArray.divRowVector(INDArray rowVector)
Division of a row vector (copy)
|
INDArray |
BaseNDArray.divRowVector(INDArray rowVector) |
protected INDArray |
BaseNDArray.doColumnWise(INDArray columnVector,
char operation)
Do a row wise op (a,s,m,d)
a : add
s : subtract
m : multiply
d : divide
h : reverse subtraction
t : reverse division
|
protected INDArray |
BaseNDArray.doRowWise(INDArray rowVector,
char operation)
Do a row wise op (a,s,m,d)
a : add
s : subtract
m : multiply
d : divide
h : reverse subtraction
t : reverse division
|
INDArray |
INDArray.dup()
Returns a copy of this ndarray
|
INDArray |
BaseNDArray.dup() |
INDArray |
INDArray.dup(char order)
Returns a copy of this ndarray, where the returned ndarray has the specified order
|
INDArray |
BaseNDArray.dup(char order) |
INDArray |
INDArray.entropy(int... dimension)
Returns entropy value for this INDArray along specified dimension(s)
|
INDArray |
BaseNDArray.entropy(int... dimension) |
INDArray |
INDArray.eps(INDArray other)
Returns the binary ndarray for "Epsilon equals" comparison.
|
INDArray |
BaseNDArray.eps(INDArray other) |
INDArray |
INDArray.eps(Number other)
Returns the binary ndarray for "Epsilon equals" comparison.
|
INDArray |
BaseNDArray.eps(Number other) |
INDArray |
INDArray.eq(INDArray other)
Returns the binary ndarray for "Equals" comparison.
|
INDArray |
BaseNDArray.eq(INDArray other) |
INDArray |
INDArray.eq(Number other)
Returns the binary ndarray for "Equals" comparison.
|
INDArray |
BaseNDArray.eq(Number other) |
INDArray |
INDArray.fmod(INDArray denominator)
remainder of division
|
INDArray |
BaseNDArray.fmod(INDArray denominator) |
INDArray |
INDArray.fmod(INDArray denominator,
INDArray result)
remainder of division
|
INDArray |
BaseNDArray.fmod(INDArray denominator,
INDArray result) |
INDArray |
INDArray.fmod(Number denominator)
remainder of division by scalar.
|
INDArray |
BaseNDArray.fmod(Number denominator) |
INDArray |
INDArray.fmod(Number denominator,
INDArray result)
remainder of division by scalar.
|
INDArray |
BaseNDArray.fmod(Number denominator,
INDArray result) |
INDArray |
INDArray.fmodi(INDArray denominator)
In place fmod
|
INDArray |
BaseNDArray.fmodi(INDArray denominator) |
INDArray |
INDArray.fmodi(Number denominator)
In place fmod
|
INDArray |
BaseNDArray.fmodi(Number denominator) |
INDArray |
INDArray.get(INDArray indices)
Get the elements from this ndarray based on the specified indices
|
INDArray |
BaseNDArray.get(INDArray indices) |
INDArray |
INDArray.get(INDArrayIndex... indexes)
Returns a subset of this array based on the specified indexes
|
INDArray |
BaseNDArray.get(INDArrayIndex... indexes) |
INDArray |
INDArray.getColumn(long i)
Returns the specified column.
|
INDArray |
BaseNDArray.getColumn(long c) |
INDArray |
INDArray.getColumn(long i,
boolean keepDim)
Returns the specified column.
|
INDArray |
BaseNDArray.getColumn(long c,
boolean keepDim) |
INDArray |
INDArray.getColumns(int... columns)
Get an INDArray comprised of the specified columns only.
|
INDArray |
BaseNDArray.getColumns(int... cindices) |
INDArray |
INDArray.getRow(long i)
Returns the specified row as a 1D vector.
|
INDArray |
BaseNDArray.getRow(long r) |
INDArray |
INDArray.getRow(long i,
boolean keepDim)
Returns the specified row.
|
INDArray |
BaseNDArray.getRow(long r,
boolean keepDim) |
INDArray |
INDArray.getRows(int... rows)
Get an INDArray comprised of the specified rows only.
|
INDArray |
BaseNDArray.getRows(int[] rindices) |
INDArray |
INDArray.getScalar(int... indices)
Returns the elements at the specified indices
|
INDArray |
BaseNDArray.getScalar(int[] indexes) |
INDArray |
INDArray.getScalar(long... indices)
See
getScalar(int[]) |
INDArray |
BaseNDArray.getScalar(long... indexes) |
INDArray |
INDArray.getScalar(long i)
Returns the element at the specified index
|
INDArray |
BaseNDArray.getScalar(long i) |
INDArray |
INDArray.getScalar(long row,
long column)
Returns the element at the specified row/column
|
INDArray |
BaseNDArray.getScalar(long row,
long column) |
INDArray |
INDArray.getWhere(INDArray comp,
Condition condition)
Boolean indexing:
Return the element if it fulfills the condition in
result array
|
INDArray |
BaseNDArray.getWhere(INDArray comp,
Condition condition) |
INDArray |
INDArray.getWhere(Number comp,
Condition condition)
Boolean indexing:
Return the element if it fulfills the condition in
result array
|
INDArray |
BaseNDArray.getWhere(Number comp,
Condition condition) |
INDArray |
INDArray.gt(INDArray other)
Returns the binary ndarray for "Greater Than" comparison.
|
INDArray |
BaseNDArray.gt(INDArray other) |
INDArray |
INDArray.gt(Number other)
Returns the binary ndarray for "Greater" comparison.
|
INDArray |
BaseNDArray.gt(Number other) |
INDArray |
INDArray.gte(Number other)
Returns binary ndarray for "Greter or equals" comparison.
|
INDArray |
BaseNDArray.gte(Number other) |
INDArray |
INDArray.isInfinite()
Returns the binary NDArray with value true where this array's entries are infinite, or false where they
are not infinite
|
INDArray |
BaseNDArray.isInfinite() |
INDArray |
INDArray.isNaN()
Returns the binary NDArray with value true where this array's entries are NaN, or false where they
are not infinite
|
INDArray |
BaseNDArray.isNaN() |
INDArray |
INDArray.leverage()
This method detaches INDArray from current Workspace, and attaches it to Workspace above, if any.
|
INDArray |
BaseNDArray.leverage() |
INDArray |
INDArray.leverageOrDetach(String id)
This method detaches INDArray from current Workspace, and attaches it to Workspace with a given Id, if a workspace
with the given ID is open and active.
|
INDArray |
BaseNDArray.leverageOrDetach(String id) |
INDArray |
INDArray.leverageTo(String id)
This method detaches INDArray from current Workspace, and attaches it to Workspace with a given Id - if a workspace
with that ID exists.
|
INDArray |
BaseNDArray.leverageTo(String id) |
INDArray |
INDArray.leverageTo(String id,
boolean enforceExistence)
This method detaches INDArray from current Workspace, and attaches it to Workspace with a given Id.
|
INDArray |
BaseNDArray.leverageTo(String id,
boolean enforceExistence) |
INDArray |
INDArray.like()
This method returns empty array with the same dtype/order/shape as this one
|
INDArray |
BaseNDArray.like() |
INDArray |
INDArray.logEntropy(int... dimension)
Returns log entropy value for this INDArray along specified dimension(s)
|
INDArray |
BaseNDArray.logEntropy(int... dimension) |
INDArray |
INDArray.lt(INDArray other)
Returns the binary ndarray for "Less" comparison.
|
INDArray |
BaseNDArray.lt(INDArray other) |
INDArray |
INDArray.lt(Number other)
Returns the binary ndarray for "Less" comparison.
|
INDArray |
BaseNDArray.lt(Number other) |
INDArray |
INDArray.lte(Number other)
Returns the binary ndarray for "Less or equals" comparison.
|
INDArray |
BaseNDArray.lte(Number other) |
INDArray |
INDArray.match(INDArray comp,
Condition condition)
Return a mask on whether each element matches the given condition
|
INDArray |
BaseNDArray.match(INDArray comp,
Condition condition) |
INDArray |
INDArray.match(Number comp,
Condition condition)
Returns a mask
|
INDArray |
BaseNDArray.match(Number comp,
Condition condition) |
INDArray |
INDArray.max(boolean keepDims,
int... dimension)
Returns the overall max of this ndarray along given dimensions
|
INDArray |
BaseNDArray.max(boolean keepDims,
int... dimension) |
INDArray |
INDArray.max(int... dimension)
Returns the overall max of this ndarray along given dimensions
|
INDArray |
BaseNDArray.max(int... dimension) |
INDArray |
INDArray.mean(boolean keepDims,
int... dimension)
Returns the overall mean of this ndarray
|
INDArray |
BaseNDArray.mean(boolean keepDims,
int... dimension) |
INDArray |
INDArray.mean(INDArray result,
boolean keepDims,
int... dimension)
Returns the overall mean of this ndarray
|
INDArray |
BaseNDArray.mean(@NonNull INDArray result,
boolean keepDims,
int... dimension) |
INDArray |
INDArray.mean(INDArray result,
int... dimension)
Returns the overall mean of this ndarray
|
INDArray |
BaseNDArray.mean(@NonNull INDArray result,
int... dimension) |
INDArray |
INDArray.mean(int... dimension)
Returns the overall mean of this ndarray
|
INDArray |
BaseNDArray.mean(int... dimension) |
INDArray |
INDArray.median(int... dimension)
This method returns median along given dimension(s)
|
INDArray |
BaseNDArray.median(int... dimension) |
INDArray |
INDArray.migrate()
This method pulls this INDArray into current Workspace.
|
INDArray |
BaseNDArray.migrate() |
INDArray |
INDArray.migrate(boolean detachOnNoWs)
This method pulls this INDArray into current Workspace, or optionally detaches if no workspace is present.
That is: If current workspace is present/active, INDArray is migrated to it. If no current workspace is present/active, one of two things occur: 1. |
INDArray |
BaseNDArray.migrate(boolean detachOnNoWs) |
INDArray |
INDArray.min(boolean keepDims,
int... dimension)
Returns the overall min of this ndarray
|
INDArray |
BaseNDArray.min(boolean keepDims,
int... dimension) |
INDArray |
INDArray.min(int... dimension)
Returns the overall min of this ndarray
|
INDArray |
BaseNDArray.min(int... dimension) |
INDArray |
INDArray.mmul(INDArray other)
Perform a copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other) |
INDArray |
INDArray.mmul(INDArray other,
char resultOrder)
Perform a copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
char resultOrder) |
INDArray |
INDArray.mmul(INDArray other,
INDArray result)
Perform an copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
INDArray result)
Perform an copy matrix multiplication
|
INDArray |
INDArray.mmul(INDArray other,
INDArray result,
MMulTranspose mMulTranspose)
Perform an copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
INDArray result,
MMulTranspose mMulTranspose) |
INDArray |
INDArray.mmul(INDArray other,
MMulTranspose mMulTranspose)
Perform a copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
MMulTranspose mMulTranspose) |
INDArray |
INDArray.mmuli(INDArray other)
Perform an inplace matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other) |
INDArray |
INDArray.mmuli(INDArray other,
INDArray result)
Perform an inplace matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other,
INDArray result) |
INDArray |
INDArray.mmuli(INDArray other,
INDArray result,
MMulTranspose transpose)
Perform an in place matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other,
INDArray result,
MMulTranspose transpose) |
INDArray |
INDArray.mmuli(INDArray other,
MMulTranspose transpose)
Perform an copy matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other,
MMulTranspose transpose) |
INDArray |
INDArray.mul(INDArray other)
copy (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.mul(INDArray other) |
INDArray |
INDArray.mul(INDArray other,
INDArray result)
copy (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.mul(INDArray other,
INDArray result) |
INDArray |
INDArray.mul(Number n)
Scalar multiplication (copy)
|
INDArray |
BaseNDArray.mul(Number n) |
INDArray |
INDArray.mul(Number n,
INDArray result)
Multiplication of ndarray.
|
INDArray |
BaseNDArray.mul(Number n,
INDArray result) |
INDArray |
INDArray.mulColumnVector(INDArray columnVector)
Multiplication of a column vector (copy)
|
INDArray |
BaseNDArray.mulColumnVector(INDArray columnVector) |
INDArray |
INDArray.muli(INDArray other)
in place (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.muli(INDArray other) |
INDArray |
INDArray.muli(INDArray other,
INDArray result)
in place (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.muli(INDArray other,
INDArray result) |
INDArray |
INDArray.muli(Number n)
In place scalar multiplication
|
INDArray |
BaseNDArray.muli(Number n) |
INDArray |
INDArray.muli(Number n,
INDArray result)
In place multiplication of this ndarray
|
INDArray |
BaseNDArray.muli(Number n,
INDArray result) |
INDArray |
INDArray.muliColumnVector(INDArray columnVector)
In place multiplication of a column vector
|
INDArray |
BaseNDArray.muliColumnVector(INDArray columnVector) |
INDArray |
INDArray.muliRowVector(INDArray rowVector)
In place multiplication of a row vector
|
INDArray |
BaseNDArray.muliRowVector(INDArray rowVector) |
INDArray |
INDArray.mulRowVector(INDArray rowVector)
Multiplication of a row vector (copy)
|
INDArray |
BaseNDArray.mulRowVector(INDArray rowVector) |
INDArray |
INDArray.neg()
Returns the ndarray negative (cloned)
|
INDArray |
BaseNDArray.neg() |
INDArray |
INDArray.negi()
In place setting of the negative version of this ndarray
|
INDArray |
BaseNDArray.negi() |
INDArray |
INDArray.neq(INDArray other)
Returns the binary ndarray for "Not equals" comparison.
|
INDArray |
BaseNDArray.neq(INDArray other) |
INDArray |
INDArray.neq(Number other)
Returns the binary ndarray for "Not equals" comparison.
|
INDArray |
BaseNDArray.neq(Number other) |
protected INDArray |
BaseNDArray.newShape(long[] newShape,
char ordering) |
INDArray |
INDArray.norm1(boolean keepDims,
int... dimension)
Returns the norm1 (L1 norm, i.e., sum of absolute values; also known as Taxicab or Manhattan norm) along the
specified dimension
|
INDArray |
BaseNDArray.norm1(boolean keepDims,
int... dimension) |
INDArray |
INDArray.norm1(int... dimension)
Returns the norm1 (L1 norm, i.e., sum of absolute values; also known as Taxicab or Manhattan norm) along the
specified dimension
|
INDArray |
BaseNDArray.norm1(int... dimension) |
INDArray |
INDArray.norm2(boolean keepDims,
int... dimension)
Returns the norm2 (L2 norm, sqrt(sum(x_i^2), also known as Euclidean norm) along the specified dimension(s)
|
INDArray |
BaseNDArray.norm2(boolean keepDims,
int... dimension) |
INDArray |
INDArray.norm2(int... dimension)
Returns the norm2 (L2 norm, sqrt(sum(x_i^2), also known as Euclidean norm) along the specified dimension(s)
|
INDArray |
BaseNDArray.norm2(int... dimension) |
INDArray |
INDArray.normmax(boolean keepDims,
int... dimension)
Returns the max norm (aka infinity norm, equal to the maximum absolute value) along the specified dimension(s)
|
INDArray |
BaseNDArray.normmax(boolean keepDims,
int... dimension) |
INDArray |
INDArray.normmax(int... dimension)
Returns the max norm (aka infinity norm, equal to the maximum absolute value) along the specified dimension(s)
|
INDArray |
BaseNDArray.normmax(int... dimension) |
INDArray |
INDArray.percentile(Number percentile,
int... dimension)
This method returns percentile along given dimension(s)
|
INDArray |
BaseNDArray.percentile(Number quantile,
int... dimension) |
INDArray |
INDArray.permute(int... rearrange)
See: http://www.mathworks.com/help/matlab/ref/permute.html
|
INDArray |
BaseNDArray.permute(int... rearrange) |
INDArray |
INDArray.permutei(int... rearrange)
An in-place version of permute.
|
INDArray |
BaseNDArray.permutei(int... rearrange) |
INDArray |
INDArray.prod(boolean keepDims,
int... dimension)
Returns the product along a given dimension
|
INDArray |
BaseNDArray.prod(boolean keepDims,
int... dimension) |
INDArray |
INDArray.prod(int... dimension)
Returns the product along a given dimension
|
INDArray |
BaseNDArray.prod(int... dimension) |
INDArray |
INDArray.put(INDArray indices,
INDArray element)
Put element in to the indices denoted by
the indices ndarray.
|
INDArray |
BaseNDArray.put(INDArray indices,
INDArray element) |
INDArray |
INDArray.put(INDArrayIndex[] indices,
INDArray element)
Put the elements of the ndarray in to the specified indices
|
INDArray |
BaseNDArray.put(INDArrayIndex[] indices,
INDArray element) |
INDArray |
INDArray.put(INDArrayIndex[] indices,
Number element)
Put the elements of the ndarray in to the specified indices
|
INDArray |
BaseNDArray.put(INDArrayIndex[] indices,
Number element) |
INDArray |
INDArray.put(int[] indices,
INDArray element)
Inserts the element at the specified index
|
INDArray |
BaseNDArray.put(int[] indices,
INDArray element) |
INDArray |
INDArray.put(int i,
INDArray element)
Inserts the element at the specified index
|
INDArray |
BaseNDArray.put(int i,
INDArray element) |
INDArray |
INDArray.put(int i,
int j,
INDArray element)
Inserts the element at the specified index
|
INDArray |
BaseNDArray.put(int i,
int j,
INDArray element) |
INDArray |
INDArray.put(int i,
int j,
Number element)
Inserts the element at the specified index
|
INDArray |
BaseNDArray.put(int i,
int j,
Number element) |
INDArray |
INDArray.putColumn(int column,
INDArray toPut)
Insert a column in to this array
Will throw an exception if this ndarray is not a matrix
|
INDArray |
BaseNDArray.putColumn(int column,
INDArray toPut) |
INDArray |
INDArray.putiColumnVector(INDArray columnVector)
In place assignment of a column vector
|
INDArray |
BaseNDArray.putiColumnVector(INDArray columnVector) |
INDArray |
INDArray.putiRowVector(INDArray rowVector)
in place assignment of row vector, to each row of this array
|
INDArray |
BaseNDArray.putiRowVector(INDArray rowVector) |
INDArray |
INDArray.putRow(long row,
INDArray toPut)
Insert a row in to this array
Will throw an exception if this ndarray is not a matrix
|
INDArray |
BaseNDArray.putRow(long row,
INDArray toPut) |
INDArray |
INDArray.putScalar(int[] i,
double value)
Insert the item at the specified indices
|
INDArray |
BaseNDArray.putScalar(int[] indexes,
double value) |
INDArray |
INDArray.putScalar(int[] indexes,
float value)
Put the specified float value at the specified indices in this array
|
INDArray |
BaseNDArray.putScalar(int[] indexes,
float value) |
INDArray |
INDArray.putScalar(int[] indexes,
int value)
Put the specified integer value at the specified indices in this array
|
INDArray |
BaseNDArray.putScalar(int[] indexes,
int value) |
INDArray |
INDArray.putScalar(long[] i,
double value)
|
INDArray |
BaseNDArray.putScalar(long[] indexes,
double value) |
INDArray |
INDArray.putScalar(long[] i,
float value)
|
INDArray |
BaseNDArray.putScalar(long[] indexes,
float value) |
INDArray |
INDArray.putScalar(long[] i,
int value)
|
INDArray |
BaseNDArray.putScalar(long[] indexes,
int value) |
INDArray |
INDArray.putScalar(long i,
double value)
Insert the number linearly in to the ndarray
|
INDArray |
BaseNDArray.putScalar(long i,
double value) |
INDArray |
INDArray.putScalar(long i,
float value)
Insert a scalar float at the specified index
|
INDArray |
BaseNDArray.putScalar(long i,
float value) |
INDArray |
INDArray.putScalar(long i,
int value)
Insert a scalar int at the specified index
|
INDArray |
BaseNDArray.putScalar(long i,
int value) |
INDArray |
INDArray.putScalar(long row,
long col,
double value)
Insert the value at the specified indices, in a 2d (rank 2) NDArray
Equivalent to putScalar(int[], double) but avoids int[] creation |
INDArray |
BaseNDArray.putScalar(long row,
long col,
double value) |
INDArray |
INDArray.putScalar(long dim0,
long dim1,
long dim2,
double value)
Insert the value at the specified indices, in a 3d (rank 3) NDArray
Equivalent to putScalar(int[], double) but avoids int[] creation |
INDArray |
BaseNDArray.putScalar(long dim0,
long dim1,
long dim2,
double value) |
INDArray |
INDArray.putScalar(long dim0,
long dim1,
long dim2,
long dim3,
double value)
Insert the value at the specified indices, in a 4d (rank 4) NDArray
Equivalent to putScalar(int[], double) but avoids int[] creation |
INDArray |
BaseNDArray.putScalar(long dim0,
long dim1,
long dim2,
long dim3,
double value) |
INDArray |
INDArray.putScalarUnsafe(long offset,
double value)
Insert a scalar at the given linear offset
|
INDArray |
BaseNDArray.putScalarUnsafe(long offset,
double value) |
INDArray |
INDArray.putSlice(int slice,
INDArray put)
Assigns the given matrix (put) to the specified slice
|
INDArray |
BaseNDArray.putSlice(int slice,
INDArray put) |
INDArray |
INDArray.putWhere(INDArray comp,
INDArray put,
Condition condition)
Assign the element according to the comparison array
|
INDArray |
BaseNDArray.putWhere(INDArray comp,
INDArray put,
Condition condition) |
INDArray |
INDArray.putWhere(Number comp,
INDArray put,
Condition condition)
Assign the element according to the comparison array
|
INDArray |
BaseNDArray.putWhere(Number comp,
INDArray put,
Condition condition) |
INDArray |
INDArray.putWhere(Number comp,
Number put,
Condition condition)
Assign the element according to the comparison array
|
INDArray |
BaseNDArray.putWhere(Number comp,
Number put,
Condition condition) |
INDArray |
INDArray.putWhereWithMask(INDArray mask,
INDArray put)
Use a pre computed mask for assigning arrays
|
INDArray |
BaseNDArray.putWhereWithMask(INDArray mask,
INDArray put) |
INDArray |
INDArray.putWhereWithMask(INDArray mask,
Number put)
Use a pre computed mask for assigning arrays
|
INDArray |
BaseNDArray.putWhereWithMask(INDArray mask,
Number put) |
INDArray |
INDArray.ravel()
Returns a flattened version (row vector) of this ndarray
|
INDArray |
BaseNDArray.ravel() |
INDArray |
INDArray.ravel(char order)
Returns a flattened version (row vector) of this ndarray
|
INDArray |
BaseNDArray.ravel(char ordering) |
INDArray |
INDArray.rdiv(INDArray other)
Reverse division, elements wise.
|
INDArray |
BaseNDArray.rdiv(INDArray other) |
INDArray |
INDArray.rdiv(INDArray other,
INDArray result)
Reverse division
|
INDArray |
BaseNDArray.rdiv(INDArray other,
INDArray result) |
INDArray |
INDArray.rdiv(Number n)
Reverse division with a scalar - i.e., (n / thisArrayValues)
|
INDArray |
BaseNDArray.rdiv(Number n) |
INDArray |
INDArray.rdiv(Number n,
INDArray result)
Reverse division (number / ndarray)
|
INDArray |
BaseNDArray.rdiv(Number n,
INDArray result) |
INDArray |
INDArray.rdivColumnVector(INDArray columnVector)
Reverse division of a column vector (copy)
|
INDArray |
BaseNDArray.rdivColumnVector(INDArray columnVector) |
INDArray |
INDArray.rdivi(INDArray other)
Reverse divsion (in place).
|
INDArray |
BaseNDArray.rdivi(INDArray other) |
INDArray |
INDArray.rdivi(INDArray other,
INDArray result)
Reverse division (in-place)
|
INDArray |
BaseNDArray.rdivi(INDArray other,
INDArray result) |
INDArray |
INDArray.rdivi(Number n)
In place reverse division - i.e., (n / thisArrayValues)
|
INDArray |
BaseNDArray.rdivi(Number n) |
INDArray |
INDArray.rdivi(Number n,
INDArray result)
Reverse in place division
|
INDArray |
BaseNDArray.rdivi(Number n,
INDArray result) |
INDArray |
INDArray.rdiviColumnVector(INDArray columnVector)
In place reverse divison of a column vector
|
INDArray |
BaseNDArray.rdiviColumnVector(INDArray columnVector) |
INDArray |
INDArray.rdiviRowVector(INDArray rowVector)
In place reverse division of a column vector
|
INDArray |
BaseNDArray.rdiviRowVector(INDArray rowVector) |
INDArray |
INDArray.rdivRowVector(INDArray rowVector)
Reverse division of a column vector (copy)
|
INDArray |
BaseNDArray.rdivRowVector(INDArray rowVector) |
INDArray |
INDArray.remainder(INDArray denominator)
Remainder operator
|
INDArray |
BaseNDArray.remainder(INDArray denominator) |
INDArray |
INDArray.remainder(INDArray denominator,
INDArray result)
Remainder operator
|
INDArray |
BaseNDArray.remainder(INDArray denominator,
INDArray result) |
INDArray |
INDArray.remainder(Number denominator)
The scalar remainder
|
INDArray |
BaseNDArray.remainder(Number denominator) |
INDArray |
INDArray.remainder(Number denominator,
INDArray result)
The scalar remainder
|
INDArray |
BaseNDArray.remainder(Number denominator,
INDArray result) |
INDArray |
INDArray.remainderi(INDArray denominator)
In place remainder
|
INDArray |
BaseNDArray.remainderi(INDArray denominator) |
INDArray |
INDArray.remainderi(Number denominator)
In place remainder
|
INDArray |
BaseNDArray.remainderi(Number denominator) |
INDArray |
INDArray.repeat(int dimension,
long... repeats)
Repeat elements along a specified dimension.
|
INDArray |
BaseNDArray.repeat(int dimension,
long... repeats) |
INDArray |
INDArray.replaceWhere(INDArray arr,
Condition condition)
Replaces all elements in this ndarray that are matching give condition, with corresponding elements from given array
|
INDArray |
BaseNDArray.replaceWhere(INDArray arr,
Condition condition) |
INDArray |
INDArray.repmat(int... shape)
Deprecated.
|
INDArray |
BaseNDArray.repmat(int[] shape)
Deprecated.
|
INDArray |
INDArray.repmat(long... shape)
Replicate and tile array to fill out to the given shape
See:
https://github.com/numpy/numpy/blob/master/numpy/matlib.py#L310-L358
|
INDArray |
BaseNDArray.repmat(long[] shape) |
INDArray |
INDArray.reshape(char order,
boolean enforceView,
long... newShape)
Reshapes the ndarray (note: it's not possible to change the length of the ndarray).
|
INDArray |
BaseNDArray.reshape(char order,
boolean enforceView,
long... newShape) |
INDArray |
INDArray.reshape(char order,
int... newShape)
Reshapes the ndarray (can't change the length of the ndarray).
|
INDArray |
BaseNDArray.reshape(char order,
int... newShape)
Deprecated.
|
INDArray |
INDArray.reshape(char order,
int rows,
int columns)
Reshapes the ndarray (can't change the length of the ndarray).
|
INDArray |
BaseNDArray.reshape(char order,
int rows,
int columns) |
INDArray |
INDArray.reshape(char order,
long... newShape)
Reshapes the ndarray (can't change the length of the ndarray).
|
INDArray |
BaseNDArray.reshape(char order,
long... newShape) |
INDArray |
INDArray.reshape(int[] shape)
See
reshape(long[]) |
INDArray |
BaseNDArray.reshape(int[] shape)
Reshape the ndarray in to the specified dimensions,
possible errors being thrown for invalid shapes
Note here that one dimension can be -1.
|
INDArray |
INDArray.reshape(long... newShape)
Reshapes the ndarray (can't change the length of the ndarray).
|
INDArray |
BaseNDArray.reshape(long... shape) |
INDArray |
INDArray.reshape(long rows,
long columns)
Reshapes the ndarray (can't change the length of the ndarray).
|
INDArray |
BaseNDArray.reshape(long newRows,
long newColumns) |
INDArray |
INDArray.rsub(INDArray other)
Element-wise reverse subtraction (copy op).
|
INDArray |
BaseNDArray.rsub(INDArray other) |
INDArray |
INDArray.rsub(INDArray other,
INDArray result)
Reverse subtraction
|
INDArray |
BaseNDArray.rsub(INDArray other,
INDArray result) |
INDArray |
INDArray.rsub(Number n)
Reverse subtraction with duplicates - i.e., (n - thisArrayValues)
|
INDArray |
BaseNDArray.rsub(Number n) |
INDArray |
INDArray.rsub(Number n,
INDArray result)
Reverse subtraction
|
INDArray |
BaseNDArray.rsub(Number n,
INDArray result) |
INDArray |
INDArray.rsubColumnVector(INDArray columnVector)
Reverse subtraction of a column vector (copy)
|
INDArray |
BaseNDArray.rsubColumnVector(INDArray columnVector) |
INDArray |
INDArray.rsubi(INDArray other)
Element-wise reverse subtraction (in the place op) - i.e., other - this
|
INDArray |
BaseNDArray.rsubi(INDArray other) |
INDArray |
INDArray.rsubi(INDArray other,
INDArray result)
Reverse subtraction (in-place)
|
INDArray |
BaseNDArray.rsubi(INDArray other,
INDArray result) |
INDArray |
INDArray.rsubi(Number n)
Reverse subtraction in place - i.e., (n - thisArrayValues)
|
INDArray |
BaseNDArray.rsubi(Number n) |
INDArray |
INDArray.rsubi(Number n,
INDArray result)
Reverse in place subtraction
|
INDArray |
BaseNDArray.rsubi(Number n,
INDArray result) |
INDArray |
INDArray.rsubiColumnVector(INDArray columnVector)
In place reverse subtraction of a column vector
|
INDArray |
BaseNDArray.rsubiColumnVector(INDArray columnVector) |
INDArray |
INDArray.rsubiRowVector(INDArray rowVector)
In place reverse subtraction of a row vector
|
INDArray |
BaseNDArray.rsubiRowVector(INDArray rowVector) |
INDArray |
INDArray.rsubRowVector(INDArray rowVector)
Reverse subtraction of a row vector (copy)
|
INDArray |
BaseNDArray.rsubRowVector(INDArray rowVector) |
INDArray |
INDArray.shannonEntropy(int... dimension)
Returns Shannon entropy value for this INDArray along specified dimension(s)
|
INDArray |
BaseNDArray.shannonEntropy(int... dimension) |
INDArray |
INDArray.slice(long i)
Returns the specified slice of this ndarray
|
INDArray |
BaseNDArray.slice(long slice) |
INDArray |
INDArray.slice(long i,
int dimension)
Returns the slice of this from the specified dimension
|
INDArray |
BaseNDArray.slice(long slice,
int dimension) |
INDArray |
INDArray.std(boolean biasCorrected,
boolean keepDims,
int... dimension)
Standard deviation of an ndarray along a dimension
|
INDArray |
BaseNDArray.std(boolean biasCorrected,
boolean keepDims,
int... dimension) |
INDArray |
INDArray.std(boolean biasCorrected,
int... dimension)
Standard deviation of an ndarray along a dimension
|
INDArray |
BaseNDArray.std(boolean biasCorrected,
int... dimension) |
INDArray |
INDArray.std(int... dimension)
Standard deviation of an INDArray along one or more dimensions
|
INDArray |
BaseNDArray.std(int... dimension) |
INDArray |
INDArray.sub(INDArray other)
copy subtraction of two NDArrays
|
INDArray |
BaseNDArray.sub(INDArray other) |
INDArray |
INDArray.sub(INDArray other,
INDArray result)
copy subtraction of two NDArrays
|
INDArray |
BaseNDArray.sub(INDArray other,
INDArray result) |
INDArray |
INDArray.sub(Number n)
Scalar subtraction (copied)
|
INDArray |
BaseNDArray.sub(Number n) |
INDArray |
INDArray.sub(Number n,
INDArray result)
Subtraction of this ndarray
|
INDArray |
BaseNDArray.sub(Number n,
INDArray result) |
INDArray |
INDArray.subColumnVector(INDArray columnVector)
Subtraction of a column vector (copy)
|
INDArray |
BaseNDArray.subColumnVector(INDArray columnVector) |
INDArray |
INDArray.subi(INDArray other)
in place (element wise) subtraction of two NDArrays
|
INDArray |
BaseNDArray.subi(INDArray other) |
INDArray |
INDArray.subi(INDArray other,
INDArray result)
in place (element wise) subtraction of two NDArrays
|
INDArray |
BaseNDArray.subi(INDArray other,
INDArray result)
in place subtraction of two matrices
|
INDArray |
INDArray.subi(Number n)
In place scalar subtraction
|
INDArray |
BaseNDArray.subi(Number n) |
INDArray |
INDArray.subi(Number n,
INDArray result)
In place subtraction of this ndarray
|
INDArray |
BaseNDArray.subi(Number n,
INDArray result) |
INDArray |
INDArray.subiColumnVector(INDArray columnVector)
In place subtraction of a column vector
|
INDArray |
BaseNDArray.subiColumnVector(INDArray columnVector) |
INDArray |
INDArray.subiRowVector(INDArray rowVector)
In place subtraction of a row vector
|
INDArray |
BaseNDArray.subiRowVector(INDArray rowVector) |
INDArray |
INDArray.subRowVector(INDArray rowVector)
Subtraction of a row vector (copy)
|
INDArray |
BaseNDArray.subRowVector(INDArray rowVector) |
INDArray |
INDArray.sum(boolean keepDims,
int... dimension)
Returns the sum along the last dimension of this ndarray
|
INDArray |
BaseNDArray.sum(boolean keepDim,
int... dimension) |
INDArray |
INDArray.sum(INDArray result,
boolean keepDims,
int... dimension)
Returns the sum along the last dimension of this ndarray
|
INDArray |
BaseNDArray.sum(@NonNull INDArray result,
boolean keepDims,
int... dimension) |
INDArray |
INDArray.sum(INDArray result,
int... dimension)
Returns the sum along the last dimension of this ndarray
|
INDArray |
BaseNDArray.sum(@NonNull INDArray result,
int... dimension) |
INDArray |
INDArray.sum(int... dimension)
Returns the sum along the last dimension of this ndarray
|
INDArray |
BaseNDArray.sum(int... dimension) |
INDArray |
INDArray.swapAxes(int dimension,
int with)
Mainly here for people coming from numpy.
|
INDArray |
BaseNDArray.swapAxes(int dimension,
int with) |
INDArray |
INDArray.tensorAlongDimension(long index,
int... dimension)
Get the vector along a particular dimension
|
INDArray |
BaseNDArray.tensorAlongDimension(long index,
int... dimension) |
INDArray |
INDArray.transpose()
Flip the rows and columns of a matrix
|
INDArray |
BaseNDArray.transpose() |
INDArray |
INDArray.transposei()
Flip the rows and columns of a matrix, in-place
|
INDArray |
BaseNDArray.transposei()
Return transposed version of this matrix.
|
INDArray |
INDArray.ulike()
This method returns uninitialized array with the same dtype/order/shape as this one
|
INDArray |
BaseNDArray.ulike() |
INDArray |
INDArray.unsafeDuplication()
Perform efficient (but unsafe) duplication.
|
INDArray |
INDArray.unsafeDuplication(boolean blocking)
Perform efficient (but unsafe) duplication.
|
INDArray |
INDArray.var(boolean biasCorrected,
int... dimension)
Returns the overall variance of this ndarray
|
INDArray |
BaseNDArray.var(boolean biasCorrected,
int... dimension) |
INDArray |
INDArray.var(int... dimension)
Returns the overall variance of this ndarray
|
INDArray |
BaseNDArray.var(int... dimension) |
INDArray |
INDArray.vectorAlongDimension(int index,
int dimension)
Get the vector along a particular dimension
|
INDArray |
BaseNDArray.vectorAlongDimension(int index,
int dimension) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
INDArray.add(INDArray other)
Element-wise copy addition of two NDArrays
|
INDArray |
BaseNDArray.add(INDArray other) |
INDArray |
INDArray.add(INDArray other,
INDArray result)
Element-wise copy addition of two NDArrays
|
INDArray |
BaseNDArray.add(INDArray other,
INDArray result) |
INDArray |
INDArray.add(Number n,
INDArray result)
Addition of this ndarray.
|
INDArray |
BaseNDArray.add(Number n,
INDArray result) |
INDArray |
INDArray.addColumnVector(INDArray columnVector)
Addition of a column vector (copy)
|
INDArray |
BaseNDArray.addColumnVector(INDArray columnVector) |
INDArray |
INDArray.addi(INDArray other)
in place (element wise) addition of two NDArrays
|
INDArray |
BaseNDArray.addi(INDArray other) |
INDArray |
INDArray.addi(INDArray other,
INDArray result)
in place (element wise) addition of two NDArrays
|
INDArray |
BaseNDArray.addi(INDArray other,
INDArray result) |
INDArray |
INDArray.addi(Number n,
INDArray result)
In place addition
|
INDArray |
BaseNDArray.addi(Number n,
INDArray result) |
INDArray |
INDArray.addiColumnVector(INDArray columnVector)
In place addition of a column vector
|
INDArray |
BaseNDArray.addiColumnVector(INDArray columnVector) |
INDArray |
INDArray.addiRowVector(INDArray rowVector)
In place addition of a row vector
|
INDArray |
BaseNDArray.addiRowVector(INDArray rowVector) |
INDArray |
INDArray.addRowVector(INDArray rowVector)
Addition of a row vector (copy)
|
INDArray |
BaseNDArray.addRowVector(INDArray rowVector) |
protected void |
BaseNDArray.assertSlice(INDArray put,
long slice) |
INDArray |
INDArray.assign(INDArray arr)
Assign all of the elements in the given ndarray to this ndarray
|
INDArray |
BaseNDArray.assign(INDArray arr) |
INDArray |
INDArray.assignIf(INDArray arr,
Condition condition)
Assign all elements from given ndarray that are matching given condition,
ndarray to this ndarray
|
INDArray |
BaseNDArray.assignIf(INDArray arr,
Condition condition) |
INDArray |
INDArray.broadcast(INDArray result)
Broadcasts this ndarray to be the specified shape
|
INDArray |
BaseNDArray.broadcast(INDArray result) |
double |
INDArray.distance1(INDArray other)
Returns the (1-norm) distance.
|
double |
BaseNDArray.distance1(INDArray other) |
double |
INDArray.distance2(INDArray other)
Returns the (euclidean) distance.
|
double |
BaseNDArray.distance2(INDArray other) |
INDArray |
INDArray.div(INDArray other)
Copy (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.div(INDArray other) |
INDArray |
INDArray.div(INDArray other,
INDArray result)
copy (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.div(INDArray other,
INDArray result) |
INDArray |
INDArray.div(Number n,
INDArray result)
Division if ndarray by number
|
INDArray |
BaseNDArray.div(Number n,
INDArray result) |
INDArray |
INDArray.divColumnVector(INDArray columnVector)
Division of a column vector (copy)
|
INDArray |
BaseNDArray.divColumnVector(INDArray columnVector) |
INDArray |
INDArray.divi(INDArray other)
in place (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.divi(INDArray other) |
INDArray |
INDArray.divi(INDArray other,
INDArray result)
in place (element wise) division of two NDArrays
|
INDArray |
BaseNDArray.divi(INDArray other,
INDArray result) |
INDArray |
INDArray.divi(Number n,
INDArray result)
In place division of this ndarray
|
INDArray |
BaseNDArray.divi(Number n,
INDArray result) |
INDArray |
INDArray.diviColumnVector(INDArray columnVector)
In place division of a column vector
|
INDArray |
BaseNDArray.diviColumnVector(INDArray columnVector) |
INDArray |
INDArray.diviRowVector(INDArray rowVector)
In place division of a row vector
|
INDArray |
BaseNDArray.diviRowVector(INDArray rowVector) |
INDArray |
INDArray.divRowVector(INDArray rowVector)
Division of a row vector (copy)
|
INDArray |
BaseNDArray.divRowVector(INDArray rowVector) |
protected INDArray |
BaseNDArray.doColumnWise(INDArray columnVector,
char operation)
Do a row wise op (a,s,m,d)
a : add
s : subtract
m : multiply
d : divide
h : reverse subtraction
t : reverse division
|
protected INDArray |
BaseNDArray.doRowWise(INDArray rowVector,
char operation)
Do a row wise op (a,s,m,d)
a : add
s : subtract
m : multiply
d : divide
h : reverse subtraction
t : reverse division
|
INDArray |
INDArray.eps(INDArray other)
Returns the binary ndarray for "Epsilon equals" comparison.
|
INDArray |
BaseNDArray.eps(INDArray other) |
INDArray |
INDArray.eq(INDArray other)
Returns the binary ndarray for "Equals" comparison.
|
INDArray |
BaseNDArray.eq(INDArray other) |
boolean |
INDArray.equalShapes(INDArray other)
This method checks 2 INDArrays for equal shapes.
Shapes are considered equal if: (a) Both arrays have equal rank, and (b) size(0)...size(rank()-1) are equal for both arrays |
boolean |
BaseNDArray.equalShapes(@NonNull INDArray other) |
INDArray |
INDArray.fmod(INDArray denominator)
remainder of division
|
INDArray |
BaseNDArray.fmod(INDArray denominator) |
INDArray |
INDArray.fmod(INDArray denominator,
INDArray result)
remainder of division
|
INDArray |
BaseNDArray.fmod(INDArray denominator,
INDArray result) |
INDArray |
INDArray.fmod(Number denominator,
INDArray result)
remainder of division by scalar.
|
INDArray |
BaseNDArray.fmod(Number denominator,
INDArray result) |
INDArray |
INDArray.fmodi(INDArray denominator)
In place fmod
|
INDArray |
BaseNDArray.fmodi(INDArray denominator) |
INDArray |
INDArray.get(INDArray indices)
Get the elements from this ndarray based on the specified indices
|
INDArray |
BaseNDArray.get(INDArray indices) |
protected double |
BaseNDArray.getPercentile(Number quantile,
INDArray sorted) |
INDArray |
INDArray.getWhere(INDArray comp,
Condition condition)
Boolean indexing:
Return the element if it fulfills the condition in
result array
|
INDArray |
BaseNDArray.getWhere(INDArray comp,
Condition condition) |
INDArray |
INDArray.gt(INDArray other)
Returns the binary ndarray for "Greater Than" comparison.
|
INDArray |
BaseNDArray.gt(INDArray other) |
INDArray |
INDArray.lt(INDArray other)
Returns the binary ndarray for "Less" comparison.
|
INDArray |
BaseNDArray.lt(INDArray other) |
INDArray |
INDArray.match(INDArray comp,
Condition condition)
Return a mask on whether each element matches the given condition
|
INDArray |
BaseNDArray.match(INDArray comp,
Condition condition) |
INDArray |
INDArray.mean(INDArray result,
boolean keepDims,
int... dimension)
Returns the overall mean of this ndarray
|
INDArray |
BaseNDArray.mean(@NonNull INDArray result,
boolean keepDims,
int... dimension) |
INDArray |
INDArray.mean(INDArray result,
int... dimension)
Returns the overall mean of this ndarray
|
INDArray |
BaseNDArray.mean(@NonNull INDArray result,
int... dimension) |
INDArray |
INDArray.mmul(INDArray other)
Perform a copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other) |
INDArray |
INDArray.mmul(INDArray other,
char resultOrder)
Perform a copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
char resultOrder) |
INDArray |
INDArray.mmul(INDArray other,
INDArray result)
Perform an copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
INDArray result)
Perform an copy matrix multiplication
|
INDArray |
INDArray.mmul(INDArray other,
INDArray result,
MMulTranspose mMulTranspose)
Perform an copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
INDArray result,
MMulTranspose mMulTranspose) |
INDArray |
INDArray.mmul(INDArray other,
MMulTranspose mMulTranspose)
Perform a copy matrix multiplication
|
INDArray |
BaseNDArray.mmul(INDArray other,
MMulTranspose mMulTranspose) |
INDArray |
INDArray.mmuli(INDArray other)
Perform an inplace matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other) |
INDArray |
INDArray.mmuli(INDArray other,
INDArray result)
Perform an inplace matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other,
INDArray result) |
INDArray |
INDArray.mmuli(INDArray other,
INDArray result,
MMulTranspose transpose)
Perform an in place matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other,
INDArray result,
MMulTranspose transpose) |
INDArray |
INDArray.mmuli(INDArray other,
MMulTranspose transpose)
Perform an copy matrix multiplication
|
INDArray |
BaseNDArray.mmuli(INDArray other,
MMulTranspose transpose) |
INDArray |
INDArray.mul(INDArray other)
copy (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.mul(INDArray other) |
INDArray |
INDArray.mul(INDArray other,
INDArray result)
copy (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.mul(INDArray other,
INDArray result) |
INDArray |
INDArray.mul(Number n,
INDArray result)
Multiplication of ndarray.
|
INDArray |
BaseNDArray.mul(Number n,
INDArray result) |
INDArray |
INDArray.mulColumnVector(INDArray columnVector)
Multiplication of a column vector (copy)
|
INDArray |
BaseNDArray.mulColumnVector(INDArray columnVector) |
INDArray |
INDArray.muli(INDArray other)
in place (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.muli(INDArray other) |
INDArray |
INDArray.muli(INDArray other,
INDArray result)
in place (element wise) multiplication of two NDArrays
|
INDArray |
BaseNDArray.muli(INDArray other,
INDArray result) |
INDArray |
INDArray.muli(Number n,
INDArray result)
In place multiplication of this ndarray
|
INDArray |
BaseNDArray.muli(Number n,
INDArray result) |
INDArray |
INDArray.muliColumnVector(INDArray columnVector)
In place multiplication of a column vector
|
INDArray |
BaseNDArray.muliColumnVector(INDArray columnVector) |
INDArray |
INDArray.muliRowVector(INDArray rowVector)
In place multiplication of a row vector
|
INDArray |
BaseNDArray.muliRowVector(INDArray rowVector) |
INDArray |
INDArray.mulRowVector(INDArray rowVector)
Multiplication of a row vector (copy)
|
INDArray |
BaseNDArray.mulRowVector(INDArray rowVector) |
INDArray |
INDArray.neq(INDArray other)
Returns the binary ndarray for "Not equals" comparison.
|
INDArray |
BaseNDArray.neq(INDArray other) |
INDArray |
INDArray.put(INDArray indices,
INDArray element)
Put element in to the indices denoted by
the indices ndarray.
|
INDArray |
BaseNDArray.put(INDArray indices,
INDArray element) |
INDArray |
INDArray.put(INDArrayIndex[] indices,
INDArray element)
Put the elements of the ndarray in to the specified indices
|
INDArray |
BaseNDArray.put(INDArrayIndex[] indices,
INDArray element) |
INDArray |
INDArray.put(int[] indices,
INDArray element)
Inserts the element at the specified index
|
INDArray |
BaseNDArray.put(int[] indices,
INDArray element) |
INDArray |
INDArray.put(int i,
INDArray element)
Inserts the element at the specified index
|
INDArray |
BaseNDArray.put(int i,
INDArray element) |
INDArray |
INDArray.put(int i,
int j,
INDArray element)
Inserts the element at the specified index
|
INDArray |
BaseNDArray.put(int i,
int j,
INDArray element) |
INDArray |
INDArray.putColumn(int column,
INDArray toPut)
Insert a column in to this array
Will throw an exception if this ndarray is not a matrix
|
INDArray |
BaseNDArray.putColumn(int column,
INDArray toPut) |
INDArray |
INDArray.putiColumnVector(INDArray columnVector)
In place assignment of a column vector
|
INDArray |
BaseNDArray.putiColumnVector(INDArray columnVector) |
INDArray |
INDArray.putiRowVector(INDArray rowVector)
in place assignment of row vector, to each row of this array
|
INDArray |
BaseNDArray.putiRowVector(INDArray rowVector) |
INDArray |
INDArray.putRow(long row,
INDArray toPut)
Insert a row in to this array
Will throw an exception if this ndarray is not a matrix
|
INDArray |
BaseNDArray.putRow(long row,
INDArray toPut) |
INDArray |
INDArray.putSlice(int slice,
INDArray put)
Assigns the given matrix (put) to the specified slice
|
INDArray |
BaseNDArray.putSlice(int slice,
INDArray put) |
INDArray |
INDArray.putWhere(INDArray comp,
INDArray put,
Condition condition)
Assign the element according to the comparison array
|
INDArray |
BaseNDArray.putWhere(INDArray comp,
INDArray put,
Condition condition) |
INDArray |
INDArray.putWhere(Number comp,
INDArray put,
Condition condition)
Assign the element according to the comparison array
|
INDArray |
BaseNDArray.putWhere(Number comp,
INDArray put,
Condition condition) |
INDArray |
INDArray.putWhereWithMask(INDArray mask,
INDArray put)
Use a pre computed mask for assigning arrays
|
INDArray |
BaseNDArray.putWhereWithMask(INDArray mask,
INDArray put) |
INDArray |
INDArray.putWhereWithMask(INDArray mask,
Number put)
Use a pre computed mask for assigning arrays
|
INDArray |
BaseNDArray.putWhereWithMask(INDArray mask,
Number put) |
INDArray |
INDArray.rdiv(INDArray other)
Reverse division, elements wise.
|
INDArray |
BaseNDArray.rdiv(INDArray other) |
INDArray |
INDArray.rdiv(INDArray other,
INDArray result)
Reverse division
|
INDArray |
BaseNDArray.rdiv(INDArray other,
INDArray result) |
INDArray |
INDArray.rdiv(Number n,
INDArray result)
Reverse division (number / ndarray)
|
INDArray |
BaseNDArray.rdiv(Number n,
INDArray result) |
INDArray |
INDArray.rdivColumnVector(INDArray columnVector)
Reverse division of a column vector (copy)
|
INDArray |
BaseNDArray.rdivColumnVector(INDArray columnVector) |
INDArray |
INDArray.rdivi(INDArray other)
Reverse divsion (in place).
|
INDArray |
BaseNDArray.rdivi(INDArray other) |
INDArray |
INDArray.rdivi(INDArray other,
INDArray result)
Reverse division (in-place)
|
INDArray |
BaseNDArray.rdivi(INDArray other,
INDArray result) |
INDArray |
INDArray.rdivi(Number n,
INDArray result)
Reverse in place division
|
INDArray |
BaseNDArray.rdivi(Number n,
INDArray result) |
INDArray |
INDArray.rdiviColumnVector(INDArray columnVector)
In place reverse divison of a column vector
|
INDArray |
BaseNDArray.rdiviColumnVector(INDArray columnVector) |
INDArray |
INDArray.rdiviRowVector(INDArray rowVector)
In place reverse division of a column vector
|
INDArray |
BaseNDArray.rdiviRowVector(INDArray rowVector) |
INDArray |
INDArray.rdivRowVector(INDArray rowVector)
Reverse division of a column vector (copy)
|
INDArray |
BaseNDArray.rdivRowVector(INDArray rowVector) |
INDArray |
INDArray.remainder(INDArray denominator)
Remainder operator
|
INDArray |
BaseNDArray.remainder(INDArray denominator) |
INDArray |
INDArray.remainder(INDArray denominator,
INDArray result)
Remainder operator
|
INDArray |
BaseNDArray.remainder(INDArray denominator,
INDArray result) |
INDArray |
INDArray.remainder(Number denominator,
INDArray result)
The scalar remainder
|
INDArray |
BaseNDArray.remainder(Number denominator,
INDArray result) |
INDArray |
INDArray.remainderi(INDArray denominator)
In place remainder
|
INDArray |
BaseNDArray.remainderi(INDArray denominator) |
INDArray |
INDArray.replaceWhere(INDArray arr,
Condition condition)
Replaces all elements in this ndarray that are matching give condition, with corresponding elements from given array
|
INDArray |
BaseNDArray.replaceWhere(INDArray arr,
Condition condition) |
INDArray |
INDArray.rsub(INDArray other)
Element-wise reverse subtraction (copy op).
|
INDArray |
BaseNDArray.rsub(INDArray other) |
INDArray |
INDArray.rsub(INDArray other,
INDArray result)
Reverse subtraction
|
INDArray |
BaseNDArray.rsub(INDArray other,
INDArray result) |
INDArray |
INDArray.rsub(Number n,
INDArray result)
Reverse subtraction
|
INDArray |
BaseNDArray.rsub(Number n,
INDArray result) |
INDArray |
INDArray.rsubColumnVector(INDArray columnVector)
Reverse subtraction of a column vector (copy)
|
INDArray |
BaseNDArray.rsubColumnVector(INDArray columnVector) |
INDArray |
INDArray.rsubi(INDArray other)
Element-wise reverse subtraction (in the place op) - i.e., other - this
|
INDArray |
BaseNDArray.rsubi(INDArray other) |
INDArray |
INDArray.rsubi(INDArray other,
INDArray result)
Reverse subtraction (in-place)
|
INDArray |
BaseNDArray.rsubi(INDArray other,
INDArray result) |
INDArray |
INDArray.rsubi(Number n,
INDArray result)
Reverse in place subtraction
|
INDArray |
BaseNDArray.rsubi(Number n,
INDArray result) |
INDArray |
INDArray.rsubiColumnVector(INDArray columnVector)
In place reverse subtraction of a column vector
|
INDArray |
BaseNDArray.rsubiColumnVector(INDArray columnVector) |
INDArray |
INDArray.rsubiRowVector(INDArray rowVector)
In place reverse subtraction of a row vector
|
INDArray |
BaseNDArray.rsubiRowVector(INDArray rowVector) |
INDArray |
INDArray.rsubRowVector(INDArray rowVector)
Reverse subtraction of a row vector (copy)
|
INDArray |
BaseNDArray.rsubRowVector(INDArray rowVector) |
double |
INDArray.squaredDistance(INDArray other)
Returns the square of the Euclidean distance.
|
double |
BaseNDArray.squaredDistance(INDArray other) |
INDArray |
INDArray.sub(INDArray other)
copy subtraction of two NDArrays
|
INDArray |
BaseNDArray.sub(INDArray other) |
INDArray |
INDArray.sub(INDArray other,
INDArray result)
copy subtraction of two NDArrays
|
INDArray |
BaseNDArray.sub(INDArray other,
INDArray result) |
INDArray |
INDArray.sub(Number n,
INDArray result)
Subtraction of this ndarray
|
INDArray |
BaseNDArray.sub(Number n,
INDArray result) |
INDArray |
INDArray.subColumnVector(INDArray columnVector)
Subtraction of a column vector (copy)
|
INDArray |
BaseNDArray.subColumnVector(INDArray columnVector) |
INDArray |
INDArray.subi(INDArray other)
in place (element wise) subtraction of two NDArrays
|
INDArray |
BaseNDArray.subi(INDArray other) |
INDArray |
INDArray.subi(INDArray other,
INDArray result)
in place (element wise) subtraction of two NDArrays
|
INDArray |
BaseNDArray.subi(INDArray other,
INDArray result)
in place subtraction of two matrices
|
INDArray |
INDArray.subi(Number n,
INDArray result)
In place subtraction of this ndarray
|
INDArray |
BaseNDArray.subi(Number n,
INDArray result) |
INDArray |
INDArray.subiColumnVector(INDArray columnVector)
In place subtraction of a column vector
|
INDArray |
BaseNDArray.subiColumnVector(INDArray columnVector) |
INDArray |
INDArray.subiRowVector(INDArray rowVector)
In place subtraction of a row vector
|
INDArray |
BaseNDArray.subiRowVector(INDArray rowVector) |
INDArray |
INDArray.subRowVector(INDArray rowVector)
Subtraction of a row vector (copy)
|
INDArray |
BaseNDArray.subRowVector(INDArray rowVector) |
INDArray |
INDArray.sum(INDArray result,
boolean keepDims,
int... dimension)
Returns the sum along the last dimension of this ndarray
|
INDArray |
BaseNDArray.sum(@NonNull INDArray result,
boolean keepDims,
int... dimension) |
INDArray |
INDArray.sum(INDArray result,
int... dimension)
Returns the sum along the last dimension of this ndarray
|
INDArray |
BaseNDArray.sum(@NonNull INDArray result,
int... dimension) |
| Modifier and Type | Method and Description |
|---|---|
void |
INDArray.sliceVectors(List<INDArray> list)
Flattens the array for linear indexing in list.
|
void |
BaseNDArray.sliceVectors(List<INDArray> list) |
| Constructor and Description |
|---|
BaseNDArrayProxy(INDArray anInstance) |
| Constructor and Description |
|---|
BaseNDArray(List<INDArray> slices,
int[] shape)
Create an ndarray from the specified slices.
|
BaseNDArray(List<INDArray> slices,
int[] shape,
char ordering)
Create an ndarray from the specified slices.
|
BaseNDArray(List<INDArray> slices,
int[] shape,
int[] stride)
Create an ndarray from the specified slices.
|
BaseNDArray(List<INDArray> slices,
int[] shape,
int[] stride,
char ordering)
Create an ndarray from the specified slices.
|
BaseNDArray(List<INDArray> slices,
long[] shape) |
BaseNDArray(List<INDArray> slices,
long[] shape,
char ordering) |
BaseNDArray(List<INDArray> slices,
long[] shape,
long[] stride) |
BaseNDArray(List<INDArray> slices,
long[] shape,
long[] stride,
char ordering) |
| Modifier and Type | Field and Description |
|---|---|
protected INDArray |
BaseOp.dimensionz |
protected INDArray |
BaseOp.x |
protected INDArray |
BaseOp.y |
protected INDArray |
BaseOp.z |
| Modifier and Type | Field and Description |
|---|---|
protected Map<Integer,INDArray> |
BaseOpContext.fastpath_in |
protected Map<Integer,INDArray> |
BaseOpContext.fastpath_out |
protected List<INDArray> |
DynamicCustomOp.inputArguments |
protected List<INDArray> |
DynamicCustomOp.outputArguments |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BroadcastOp.dimensions() |
INDArray |
IndexAccumulation.dimensions()
This method returns dimensions for this op
|
INDArray |
ReduceOp.dimensions()
This method returns dimensions for this op
|
INDArray |
BaseReduceSameOp.dimensions() |
INDArray |
BaseOp.dimensions() |
INDArray |
ScalarOp.dimensions()
This method returns target dimensions for this op
|
INDArray |
DynamicCustomOp.generateFake(DataType dataType,
long... shape)
Generate fake data for
DynamicCustomOp.computeArrays()
of the the given shape with the given data type |
INDArray |
DynamicCustomOp.generateFake(long... shape)
Generate fake data for
DynamicCustomOp.computeArrays()
of the the given shape with the data type Nd4j.defaultFloatingPointType() |
INDArray |
CustomOp.getInputArgument(int index) |
INDArray |
DynamicCustomOp.getInputArgument(int index) |
INDArray |
BaseOp.getInputArgument(int index) |
INDArray |
OpContext.getInputArray(int idx) |
INDArray |
BaseOpContext.getInputArray(int idx) |
INDArray |
CustomOp.getOutputArgument(int index) |
INDArray |
DynamicCustomOp.getOutputArgument(int index) |
INDArray |
OpContext.getOutputArray(int i) |
INDArray |
BaseOpContext.getOutputArray(int i) |
INDArray |
LossFunction.input()
The true
|
INDArray |
BaseReduceOp.noOp() |
INDArray |
ReduceOp.noOp()
Returns the no op version
of the input
Basically when a reduce can't happen (eg: sum(0) on a row vector)
you have a no op state for a given reduction.
|
INDArray |
LossFunction.output()
The guess
|
INDArray |
BaseScalarBoolOp.scalar() |
INDArray |
BaseScalarOp.scalar() |
INDArray |
ScalarOp.scalar()
The normal scalar
|
protected static INDArray[] |
DynamicCustomOp.wrapOrNull(INDArray in) |
INDArray |
BaseOp.x() |
INDArray |
Op.x()
The origin ndarray
|
INDArray |
BaseOp.y() |
INDArray |
Op.y()
The pairwise op ndarray
|
INDArray |
BaseScalarBoolOp.z() |
INDArray |
BaseScalarOp.z() |
INDArray |
BaseTransformOp.z() |
INDArray |
BaseOp.z() |
INDArray |
Op.z()
The resulting ndarray
|
| Modifier and Type | Method and Description |
|---|---|
List<INDArray> |
OpContext.getInputArrays()
This method returns List of input arrays defined within this context
|
List<INDArray> |
BaseOpContext.getInputArrays() |
List<INDArray> |
OpContext.getOutputArrays()
This method returns List of output arrays defined within this context
|
List<INDArray> |
BaseOpContext.getOutputArrays() |
List<INDArray> |
CustomOp.inputArguments() |
List<INDArray> |
DynamicCustomOp.inputArguments() |
List<INDArray> |
CustomOp.outputArguments() |
List<INDArray> |
DynamicCustomOp.outputArguments() |
| Modifier and Type | Method and Description |
|---|---|
void |
CustomOp.addInputArgument(INDArray... arg) |
void |
DynamicCustomOp.addInputArgument(INDArray... arg) |
DynamicCustomOp.DynamicCustomOpsBuilder |
DynamicCustomOp.DynamicCustomOpsBuilder.addInputs(INDArray... inputs)
This method
takes arbitrary number of input INDArrays in, as Op input
Note that this ACCUMULATES arguments.
|
void |
CustomOp.addOutputArgument(INDArray... arg) |
void |
DynamicCustomOp.addOutputArgument(INDArray... arg) |
DynamicCustomOp.DynamicCustomOpsBuilder |
DynamicCustomOp.DynamicCustomOpsBuilder.addOutputs(INDArray... outputs)
This method takes arbitrary number of
output INDArrays in, to store operation result
Note that this ACCUMULATES arguments.
|
void |
CustomOp.removeInputArgument(INDArray arg) |
void |
DynamicCustomOp.removeInputArgument(INDArray arg) |
void |
CustomOp.removeOutputArgument(INDArray arg) |
void |
DynamicCustomOp.removeOutputArgument(INDArray arg) |
void |
DynamicCustomOp.setInputArgument(int index,
INDArray input) |
void |
DynamicCustomOp.setInputArguments(INDArray... inputs) |
void |
OpContext.setInputArray(int index,
INDArray array)
This method adds INDArray as input argument for future op call
|
void |
BaseOpContext.setInputArray(int index,
@NonNull INDArray array) |
void |
OpContext.setInputArrays(INDArray... arrays)
This method sets provided arrays as input arrays
|
void |
BaseOpContext.setInputArrays(INDArray... arrays) |
void |
DynamicCustomOp.setOutputArgument(int index,
INDArray output) |
void |
OpContext.setOutputArray(int index,
INDArray array)
This method adds INDArray as output for future op call
|
void |
BaseOpContext.setOutputArray(int index,
@NonNull INDArray array) |
void |
OpContext.setOutputArrays(INDArray... arrays)
This method sets provided arrays as output arrays
|
void |
BaseOpContext.setOutputArrays(INDArray... arrays) |
void |
BaseScalarBoolOp.setScalar(INDArray scalar) |
void |
BaseScalarOp.setScalar(INDArray scalar) |
void |
ScalarOp.setScalar(INDArray scalar) |
void |
BaseOp.setX(INDArray x) |
void |
Op.setX(INDArray x)
set x (the input ndarray)
|
void |
BaseOp.setY(INDArray y) |
void |
Op.setY(INDArray y)
set y(the pairwise ndarray)
|
void |
BaseOp.setZ(INDArray z) |
void |
Op.setZ(INDArray z)
set z (the solution ndarray)
|
protected static INDArray[] |
DynamicCustomOp.wrapOrNull(INDArray in) |
| Modifier and Type | Method and Description |
|---|---|
void |
OpContext.setInputArrays(List<INDArray> arrays)
This method sets provided arrays as input arrays
|
void |
OpContext.setOutputArrays(List<INDArray> arrays)
This method sets provided arrays as output arrays
|
| Constructor and Description |
|---|
BaseBroadcastBoolOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BaseBroadcastOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BaseIndexAccumulation(INDArray x,
boolean keepDims,
int[] dimensions) |
BaseIndexAccumulation(INDArray x,
INDArray z,
int[] dimensions) |
BaseIndexAccumulation(INDArray x,
int[] dimensions) |
BaseOp(INDArray x)
An op for one ndarray
|
BaseOp(INDArray x,
INDArray z)
Specify an alternative result array
|
BaseOp(INDArray x,
INDArray y,
INDArray z) |
BaseReduceBoolOp(INDArray x,
boolean keepDims,
int... dimensions) |
BaseReduceBoolOp(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
BaseReduceBoolOp(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
BaseReduceBoolOp(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
BaseReduceBoolOp(INDArray x,
INDArray z,
int... dimensions) |
BaseReduceBoolOp(INDArray x,
int... dimensions) |
BaseReduceFloatOp(INDArray x,
boolean keepDims,
int... dimensions) |
BaseReduceFloatOp(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
BaseReduceFloatOp(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
BaseReduceFloatOp(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
BaseReduceFloatOp(INDArray x,
INDArray z,
int... dimensions) |
BaseReduceFloatOp(INDArray x,
int... dimensions) |
BaseReduceLongOp(INDArray x,
boolean keepDims,
int... dimensions) |
BaseReduceLongOp(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
BaseReduceLongOp(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
BaseReduceLongOp(INDArray x,
INDArray z,
int... dimensions) |
BaseReduceLongOp(INDArray x,
int... dimensions) |
BaseReduceOp(INDArray x,
boolean keepDims,
int... dimensions) |
BaseReduceOp(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
BaseReduceOp(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
BaseReduceOp(INDArray x,
INDArray y,
int... dimensions) |
BaseReduceOp(INDArray x,
int... dimensions) |
BaseReduceSameOp(INDArray x,
boolean keepDims,
int... dimensions) |
BaseReduceSameOp(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
BaseReduceSameOp(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
BaseReduceSameOp(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
BaseReduceSameOp(INDArray x,
INDArray y,
int... dimensions) |
BaseReduceSameOp(INDArray x,
int... dimensions) |
BaseScalarBoolOp(INDArray x,
INDArray y,
INDArray z,
Number num) |
BaseScalarBoolOp(INDArray x,
INDArray z,
Number set) |
BaseScalarBoolOp(INDArray x,
Number num) |
BaseScalarOp(INDArray x,
INDArray y,
INDArray z,
Number num) |
BaseScalarOp(INDArray x,
INDArray z,
Number set) |
BaseScalarOp(INDArray x,
Number num) |
BaseTransformAnyOp(INDArray x) |
BaseTransformAnyOp(INDArray x,
INDArray z) |
BaseTransformAnyOp(INDArray x,
INDArray y,
INDArray z) |
BaseTransformBoolOp(INDArray x) |
BaseTransformBoolOp(INDArray x,
INDArray z) |
BaseTransformBoolOp(INDArray x,
INDArray y,
INDArray z) |
BaseTransformFloatOp(INDArray x) |
BaseTransformFloatOp(INDArray x,
INDArray z) |
BaseTransformOp(INDArray x) |
BaseTransformOp(INDArray x,
INDArray z) |
BaseTransformOp(INDArray x,
INDArray y,
INDArray z) |
BaseTransformSameOp(INDArray x) |
BaseTransformSameOp(INDArray x,
INDArray z) |
BaseTransformSameOp(INDArray x,
INDArray y,
INDArray z) |
BaseTransformStrictOp(INDArray x) |
BaseTransformStrictOp(INDArray x,
INDArray z) |
DynamicCustomOp(INDArray[] inputs,
INDArray[] outputs)
Initialize this operation for execution (pre created ndarrays)
|
DynamicCustomOp(INDArray[] inputs,
INDArray[] outputs)
Initialize this operation for execution (pre created ndarrays)
|
DynamicCustomOp(String opName,
INDArray[] inputs,
INDArray[] outputs)
Initialize this operation for execution (pre created ndarrays)
|
DynamicCustomOp(String opName,
INDArray[] inputs,
INDArray[] outputs)
Initialize this operation for execution (pre created ndarrays)
|
DynamicCustomOp(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments) |
DynamicCustomOp(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments) |
DynamicCustomOp(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments)
Initialize this custom op with all of the
inputs, outputs, and respective
arguments for execution
|
DynamicCustomOp(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments)
Initialize this custom op with all of the
inputs, outputs, and respective
arguments for execution
|
DynamicCustomOp(String opName,
INDArray input,
INDArray output,
List<Double> tArguments,
int[] iArguments) |
NoOp(INDArray in) |
| Modifier and Type | Field and Description |
|---|---|
protected List<INDArray> |
BaseAggregate.arguments |
| Modifier and Type | Method and Description |
|---|---|
List<INDArray> |
Aggregate.getArguments() |
List<INDArray> |
BaseAggregate.getArguments() |
| Constructor and Description |
|---|
AggregateAxpy(@NonNull INDArray x,
@NonNull INDArray y,
double alpha)
Deprecated.
|
AggregateAxpy(@NonNull INDArray x,
@NonNull INDArray y,
double alpha)
Deprecated.
|
AggregateGEMM(int Order,
int TransA,
int TransB,
int M,
int N,
int K,
double alpha,
@NonNull INDArray A,
int lda,
@NonNull INDArray B,
int ldb,
double beta,
@NonNull INDArray C,
int ldc) |
AggregateGEMM(int Order,
int TransA,
int TransB,
int M,
int N,
int K,
double alpha,
@NonNull INDArray A,
int lda,
@NonNull INDArray B,
int ldb,
double beta,
@NonNull INDArray C,
int ldc) |
AggregateGEMM(int Order,
int TransA,
int TransB,
int M,
int N,
int K,
double alpha,
@NonNull INDArray A,
int lda,
@NonNull INDArray B,
int ldb,
double beta,
@NonNull INDArray C,
int ldc) |
| Constructor and Description |
|---|
CompatSparseToDense(INDArray indices,
INDArray shape,
INDArray values) |
CompatSparseToDense(INDArray indices,
INDArray shape,
INDArray values,
INDArray defaultValue) |
CompatStringSplit(INDArray strings,
INDArray delimiter) |
CompatStringSplit(INDArray strings,
INDArray delimiter,
INDArray indices,
INDArray values) |
| Constructor and Description |
|---|
DecodeBitmap(@NonNull INDArray encoded,
@NonNull INDArray updates) |
DecodeBitmap(@NonNull INDArray encoded,
@NonNull INDArray updates) |
DecodeThreshold(@NonNull INDArray encoded,
@NonNull INDArray updates) |
DecodeThreshold(@NonNull INDArray encoded,
@NonNull INDArray updates) |
EncodeBitmap(@NonNull INDArray updates,
float threshold) |
EncodeBitmap(@NonNull INDArray updates,
@NonNull INDArray encoded,
@NonNull INDArray counter,
float threshold) |
EncodeBitmap(@NonNull INDArray updates,
@NonNull INDArray encoded,
@NonNull INDArray counter,
float threshold) |
EncodeBitmap(@NonNull INDArray updates,
@NonNull INDArray encoded,
@NonNull INDArray counter,
float threshold) |
EncodeThreshold(@NonNull INDArray updates,
float threshold) |
EncodeThreshold(@NonNull INDArray updates,
float threshold,
@NonNull Integer boundary) |
EncodeThreshold(@NonNull INDArray updates,
@NonNull INDArray encoded,
float threshold,
@NonNull Integer boundary) |
EncodeThreshold(@NonNull INDArray updates,
@NonNull INDArray encoded,
float threshold,
@NonNull Integer boundary) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BarnesHutSymmetrize.getSymmetrizedCols() |
INDArray |
BarnesHutSymmetrize.getSymmetrizedValues() |
| Constructor and Description |
|---|
AdjustContrast(@NonNull INDArray in,
double factor) |
AdjustContrast(@NonNull INDArray in,
double factor,
INDArray out) |
AdjustContrast(@NonNull INDArray in,
double factor,
INDArray out) |
AdjustHue(@NonNull INDArray in,
double delta) |
AdjustHue(@NonNull INDArray in,
double delta,
INDArray out) |
AdjustHue(@NonNull INDArray in,
double delta,
INDArray out) |
AdjustSaturation(@NonNull INDArray in,
double factor) |
AdjustSaturation(@NonNull INDArray in,
double factor,
INDArray out) |
AdjustSaturation(@NonNull INDArray in,
double factor,
INDArray out) |
BarnesEdgeForces(INDArray rowP,
INDArray colP,
INDArray valP,
INDArray dataP,
long N,
INDArray output) |
BarnesHutGains(INDArray output,
INDArray input,
INDArray gradx,
INDArray epsilon) |
BarnesHutSymmetrize(INDArray rowP,
INDArray colP,
INDArray valP,
long N,
INDArray outRows) |
BetaInc(@NonNull INDArray a_input,
@NonNull INDArray b_input,
@NonNull INDArray x_input) |
BetaInc(@NonNull INDArray a_input,
@NonNull INDArray b_input,
@NonNull INDArray x_input) |
BetaInc(@NonNull INDArray a_input,
@NonNull INDArray b_input,
@NonNull INDArray x_input) |
BetaInc(@NonNull INDArray a_input,
@NonNull INDArray b_input,
@NonNull INDArray x_input,
INDArray output) |
BetaInc(@NonNull INDArray a_input,
@NonNull INDArray b_input,
@NonNull INDArray x_input,
INDArray output) |
BetaInc(@NonNull INDArray a_input,
@NonNull INDArray b_input,
@NonNull INDArray x_input,
INDArray output) |
BetaInc(@NonNull INDArray a_input,
@NonNull INDArray b_input,
@NonNull INDArray x_input,
INDArray output) |
BitCast(INDArray in,
DataType dataType) |
BitCast(INDArray in,
DataType dataType,
INDArray out) |
BitCast(INDArray in,
int dataType) |
BitCast(INDArray in,
int dataType,
INDArray out) |
CompareAndBitpack(INDArray in,
double threshold) |
CompareAndBitpack(INDArray in,
double threshold,
INDArray out) |
Digamma(@NonNull INDArray x) |
DivideNoNan(INDArray in1,
INDArray in2) |
DivideNoNan(INDArray in1,
INDArray in2,
INDArray out) |
DrawBoundingBoxes(INDArray images,
INDArray boxes,
INDArray colors) |
DrawBoundingBoxes(INDArray images,
INDArray boxes,
INDArray colors,
INDArray output) |
FakeQuantWithMinMaxVarsPerChannel(INDArray x,
INDArray min,
INDArray max) |
FakeQuantWithMinMaxVarsPerChannel(INDArray x,
INDArray min,
INDArray max,
boolean narrow) |
FakeQuantWithMinMaxVarsPerChannel(INDArray x,
INDArray min,
INDArray max,
int num_bits) |
FakeQuantWithMinMaxVarsPerChannel(INDArray x,
INDArray min,
INDArray max,
int num_bits,
boolean narrow) |
Flatten(char order,
INDArray... inputs) |
Flatten(INDArray output,
INDArray... inputs) |
Flatten(INDArray output,
INDArray... inputs) |
FusedBatchNorm(@NonNull INDArray x,
@NonNull INDArray scale,
@NonNull INDArray offset,
int dataFormat,
int isTraining,
INDArray yOut,
INDArray batchMeanOut,
INDArray batchMeanVar) |
FusedBatchNorm(@NonNull INDArray x,
@NonNull INDArray scale,
@NonNull INDArray offset,
int dataFormat,
int isTraining,
INDArray yOut,
INDArray batchMeanOut,
INDArray batchMeanVar) |
FusedBatchNorm(@NonNull INDArray x,
@NonNull INDArray scale,
@NonNull INDArray offset,
int dataFormat,
int isTraining,
INDArray yOut,
INDArray batchMeanOut,
INDArray batchMeanVar) |
FusedBatchNorm(@NonNull INDArray x,
@NonNull INDArray scale,
@NonNull INDArray offset,
int dataFormat,
int isTraining,
INDArray yOut,
INDArray batchMeanOut,
INDArray batchMeanVar) |
HsvToRgb(INDArray input) |
Igamma(@NonNull INDArray n,
@NonNull INDArray x) |
Igamma(@NonNull INDArray n,
@NonNull INDArray x) |
Igamma(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
Igamma(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
Igamma(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
Igammac(@NonNull INDArray n,
@NonNull INDArray x) |
Igammac(@NonNull INDArray n,
@NonNull INDArray x) |
Igammac(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
Igammac(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
Igammac(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
KnnMinDistance(INDArray point,
INDArray lowest,
INDArray highest,
INDArray distance) |
Lgamma(@NonNull INDArray x) |
Lgamma(@NonNull INDArray x,
INDArray output) |
Lgamma(@NonNull INDArray x,
INDArray output) |
LinearSolve(INDArray a,
INDArray b) |
LinearSolve(INDArray a,
INDArray b,
boolean adjoint) |
Logdet(INDArray input) |
Lstsq(@NonNull INDArray matrix,
@NonNull INDArray rhs) |
Lstsq(@NonNull INDArray matrix,
@NonNull INDArray rhs) |
Lstsq(@NonNull INDArray matrix,
@NonNull INDArray rhs,
double l2_regularizer,
boolean fast) |
Lstsq(@NonNull INDArray matrix,
@NonNull INDArray rhs,
double l2_regularizer,
boolean fast) |
Lu(INDArray input) |
MatrixBandPart(@NonNull INDArray input,
int minLower,
int maxUpper) |
Polygamma(@NonNull INDArray n,
@NonNull INDArray x) |
Polygamma(@NonNull INDArray n,
@NonNull INDArray x) |
Polygamma(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
Polygamma(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
Polygamma(@NonNull INDArray n,
@NonNull INDArray x,
INDArray output) |
RandomCrop(@NonNull INDArray input,
@NonNull INDArray shape) |
RandomCrop(@NonNull INDArray input,
@NonNull INDArray shape) |
RgbToGrayscale(INDArray image) |
RgbToHsv(INDArray input) |
RgbToYiq(INDArray input) |
RgbToYuv(INDArray input) |
Roll(@NonNull INDArray input,
@NonNull INDArray shifts,
@NonNull INDArray axes) |
Roll(@NonNull INDArray input,
@NonNull INDArray shifts,
@NonNull INDArray axes) |
Roll(@NonNull INDArray input,
@NonNull INDArray shifts,
@NonNull INDArray axes) |
Roll(@NonNull INDArray input,
int shift) |
SpTreeCell(INDArray corner,
INDArray width,
INDArray point,
long N,
boolean contains) |
ToggleBits(@NonNull INDArray input) |
ToggleBits(@NonNull INDArray input,
INDArray output) |
ToggleBits(@NonNull INDArray input,
INDArray output) |
TriangularSolve(INDArray matrix,
INDArray rhs,
boolean lower,
boolean adjoint) |
Triu(INDArray input,
int diag) |
YiqToRgb(INDArray input) |
YuvToRgb(INDArray input) |
| Modifier and Type | Method and Description |
|---|---|
INDArray[] |
DefaultOpExecutioner.allocateOutputArrays(CustomOp op) |
INDArray[] |
OpExecutioner.allocateOutputArrays(CustomOp op)
Equivalent to calli
|
INDArray |
DefaultOpExecutioner.bitmapDecode(INDArray encoded,
INDArray target) |
INDArray |
OpExecutioner.bitmapDecode(INDArray encoded,
INDArray target) |
INDArray |
DefaultOpExecutioner.bitmapEncode(INDArray indArray,
double threshold) |
INDArray |
OpExecutioner.bitmapEncode(INDArray indArray,
double threshold) |
abstract INDArray |
DefaultOpExecutioner.exec(BroadcastOp broadcast) |
INDArray |
OpExecutioner.exec(BroadcastOp broadcast)
Execute a broadcast op, possibly along one or more dimensions
|
INDArray[] |
DefaultOpExecutioner.exec(CustomOp op) |
INDArray[] |
OpExecutioner.exec(CustomOp op) |
INDArray[] |
DefaultOpExecutioner.exec(CustomOp op,
OpContext context) |
INDArray[] |
OpExecutioner.exec(CustomOp op,
OpContext context)
This method executes op with given context
|
abstract INDArray |
DefaultOpExecutioner.exec(IndexAccumulation op) |
INDArray |
OpExecutioner.exec(IndexAccumulation indexAccum)
Execute an index accumulation along one or more dimensions
|
abstract INDArray |
DefaultOpExecutioner.exec(Op op) |
INDArray |
OpExecutioner.exec(Op op)
Execute the operation
|
abstract INDArray |
DefaultOpExecutioner.exec(Op op,
OpContext opContext) |
INDArray |
OpExecutioner.exec(Op op,
OpContext opContext)
Execute the operation
|
INDArray |
DefaultOpExecutioner.exec(RandomOp op)
This method executes specified RandomOp using default RNG available via Nd4j.getRandom()
|
INDArray |
OpExecutioner.exec(RandomOp op)
This method executes specified RandomOp using default RNG available via Nd4j.getRandom()
|
abstract INDArray |
DefaultOpExecutioner.exec(RandomOp op,
Random rng)
This method executes specific RandomOp against specified RNG
|
INDArray |
OpExecutioner.exec(RandomOp op,
Random rng)
This method executes specific RandomOp against specified RNG
|
abstract INDArray |
DefaultOpExecutioner.exec(ReduceOp op) |
INDArray |
OpExecutioner.exec(ReduceOp reduceOp)
Execute a reduceOp, possibly along one or more dimensions
|
abstract INDArray |
DefaultOpExecutioner.exec(ScalarOp op) |
INDArray |
OpExecutioner.exec(ScalarOp broadcast)
Execute ScalarOp
|
abstract INDArray |
DefaultOpExecutioner.exec(Variance accumulation) |
INDArray |
OpExecutioner.exec(Variance accumulation)
Execute an variance accumulation op, possibly along one or more dimensions
|
INDArray |
DefaultOpExecutioner.getX(Op op,
OpContext oc) |
INDArray |
DefaultOpExecutioner.getY(Op op,
OpContext oc) |
INDArray |
DefaultOpExecutioner.getZ(Op op,
OpContext oc) |
INDArray |
DefaultOpExecutioner.thresholdDecode(INDArray encoded,
INDArray target) |
INDArray |
OpExecutioner.thresholdDecode(INDArray encoded,
INDArray target)
This method decodes thresholds array, and puts it into target array
|
INDArray |
DefaultOpExecutioner.thresholdEncode(INDArray input,
double threshold) |
INDArray |
OpExecutioner.thresholdEncode(INDArray input,
double threshold)
This method encodes array as thresholds, updating input array at the same time
|
INDArray |
DefaultOpExecutioner.thresholdEncode(INDArray input,
double threshold,
Integer boundary) |
INDArray |
OpExecutioner.thresholdEncode(INDArray input,
double threshold,
Integer boundary)
This method encodes array as thresholds, updating input array at the same time
|
| Modifier and Type | Method and Description |
|---|---|
Map<String,INDArray> |
DefaultOpExecutioner.executeGraph(long id,
Map<String,INDArray> map,
Map<String,Integer> reverseMap) |
Map<String,INDArray> |
OpExecutioner.executeGraph(long id,
Map<String,INDArray> map,
Map<String,Integer> reverseMap) |
| Modifier and Type | Method and Description |
|---|---|
String |
DefaultOpExecutioner.arrayInfo(INDArray arr) |
INDArray |
DefaultOpExecutioner.bitmapDecode(INDArray encoded,
INDArray target) |
INDArray |
OpExecutioner.bitmapDecode(INDArray encoded,
INDArray target) |
INDArray |
DefaultOpExecutioner.bitmapEncode(INDArray indArray,
double threshold) |
INDArray |
OpExecutioner.bitmapEncode(INDArray indArray,
double threshold) |
long |
DefaultOpExecutioner.bitmapEncode(INDArray indArray,
INDArray target,
double threshold) |
long |
OpExecutioner.bitmapEncode(INDArray indArray,
INDArray target,
double threshold)
This method returns number of elements affected by encoder
|
static void |
OpExecutionerUtil.checkForAny(INDArray z) |
static void |
OpExecutionerUtil.checkForInf(INDArray z) |
static void |
OpExecutionerUtil.checkForNaN(INDArray z) |
protected void |
DefaultOpExecutioner.checkWorkspace(String opName,
INDArray array) |
protected static String |
DefaultOpExecutioner.firstX(INDArray array,
int x) |
INDArrayStatistics |
DefaultOpExecutioner.inspectArray(INDArray array) |
INDArrayStatistics |
OpExecutioner.inspectArray(INDArray array) |
void |
DefaultOpExecutioner.scatterUpdate(ScatterUpdate.UpdateOp op,
INDArray array,
INDArray indices,
INDArray updates,
int[] axis) |
void |
OpExecutioner.scatterUpdate(ScatterUpdate.UpdateOp op,
@NonNull INDArray array,
@NonNull INDArray indices,
@NonNull INDArray updates,
int[] axis)
Deprecated.
|
void |
OpExecutioner.scatterUpdate(ScatterUpdate.UpdateOp op,
@NonNull INDArray array,
@NonNull INDArray indices,
@NonNull INDArray updates,
int[] axis)
Deprecated.
|
void |
OpExecutioner.scatterUpdate(ScatterUpdate.UpdateOp op,
@NonNull INDArray array,
@NonNull INDArray indices,
@NonNull INDArray updates,
int[] axis)
Deprecated.
|
void |
DefaultOpExecutioner.setX(INDArray x,
Op op,
OpContext oc) |
void |
DefaultOpExecutioner.setY(INDArray y,
Op op,
OpContext oc) |
void |
DefaultOpExecutioner.setZ(INDArray z,
Op op,
OpContext oc) |
TadPack |
DefaultOpExecutioner.tadShapeInfoAndOffsets(INDArray array,
int[] dimension) |
TadPack |
OpExecutioner.tadShapeInfoAndOffsets(INDArray array,
int[] dimension)
This method returns host/device tad buffers
|
INDArray |
DefaultOpExecutioner.thresholdDecode(INDArray encoded,
INDArray target) |
INDArray |
OpExecutioner.thresholdDecode(INDArray encoded,
INDArray target)
This method decodes thresholds array, and puts it into target array
|
INDArray |
DefaultOpExecutioner.thresholdEncode(INDArray input,
double threshold) |
INDArray |
OpExecutioner.thresholdEncode(INDArray input,
double threshold)
This method encodes array as thresholds, updating input array at the same time
|
INDArray |
DefaultOpExecutioner.thresholdEncode(INDArray input,
double threshold,
Integer boundary) |
INDArray |
OpExecutioner.thresholdEncode(INDArray input,
double threshold,
Integer boundary)
This method encodes array as thresholds, updating input array at the same time
|
static void |
DefaultOpExecutioner.validateDataType(DataType expectedType,
Object op,
INDArray... operands) |
| Modifier and Type | Method and Description |
|---|---|
Map<String,INDArray> |
DefaultOpExecutioner.executeGraph(long id,
Map<String,INDArray> map,
Map<String,Integer> reverseMap) |
Map<String,INDArray> |
OpExecutioner.executeGraph(long id,
Map<String,INDArray> map,
Map<String,Integer> reverseMap) |
| Constructor and Description |
|---|
BiasAdd(@NonNull INDArray input,
@NonNull INDArray bias,
boolean nchw) |
BiasAdd(@NonNull INDArray input,
@NonNull INDArray bias,
boolean nchw) |
BiasAdd(@NonNull INDArray input,
@NonNull INDArray bias,
INDArray output,
boolean nchw) |
BiasAdd(@NonNull INDArray input,
@NonNull INDArray bias,
INDArray output,
boolean nchw) |
BiasAdd(@NonNull INDArray input,
@NonNull INDArray bias,
INDArray output,
boolean nchw) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient,
boolean nchw) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient,
boolean nchw) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient,
boolean nchw) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient,
INDArray output) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient,
INDArray output) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient,
INDArray output) |
BiasAddGrad(@NonNull INDArray input,
@NonNull INDArray bias,
@NonNull INDArray gradient,
INDArray output) |
BroadcastAddOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastAMax(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastAMin(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastCopyOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastDivOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastGradientArgs(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastMax(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastMin(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastMulOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastRDivOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastRSubOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastSubOp(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastTo(@NonNull INDArray input,
@NonNull INDArray shape,
@NonNull INDArray output) |
BroadcastTo(@NonNull INDArray input,
@NonNull INDArray shape,
@NonNull INDArray output) |
BroadcastTo(@NonNull INDArray input,
@NonNull INDArray shape,
@NonNull INDArray output) |
BroadcastTo(@NonNull INDArray input,
@NonNull long[] shape,
@NonNull INDArray output) |
BroadcastTo(@NonNull INDArray input,
@NonNull long[] shape,
@NonNull INDArray output) |
| Constructor and Description |
|---|
BroadcastEqualTo(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastGreaterThan(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastGreaterThanOrEqual(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastLessThan(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastLessThanOrEqual(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
BroadcastNotEqual(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
| Constructor and Description |
|---|
Select(INDArray[] inputs,
INDArray[] outputs) |
Select(INDArray[] inputs,
INDArray[] outputs) |
Select(INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
Select(INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
Where(INDArray condition) |
Where(INDArray[] inputs,
INDArray[] outputs) |
Where(INDArray[] inputs,
INDArray[] outputs) |
Where(INDArray x,
INDArray condition) |
Where(INDArray x,
INDArray y,
INDArray condition) |
Where(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
Where(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
WhereNumpy(INDArray[] inputs,
INDArray[] outputs) |
WhereNumpy(INDArray[] inputs,
INDArray[] outputs) |
WhereNumpy(INDArray x,
INDArray y,
INDArray condition) |
WhereNumpy(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
WhereNumpy(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
| Constructor and Description |
|---|
BaseCompatOp(INDArray... inputs) |
Merge(INDArray... inputs) |
Switch(INDArray input,
INDArray predicate) |
| Constructor and Description |
|---|
BaseGridOp(INDArray x,
INDArray y) |
FreeGridOp(INDArray x,
INDArray y) |
| Constructor and Description |
|---|
CropAndResize(@NonNull INDArray image,
@NonNull INDArray cropBoxes,
@NonNull INDArray boxIndices,
@NonNull INDArray cropOutSize,
@NonNull CropAndResize.Method method,
double extrapolationValue,
INDArray output) |
CropAndResize(@NonNull INDArray image,
@NonNull INDArray cropBoxes,
@NonNull INDArray boxIndices,
@NonNull INDArray cropOutSize,
@NonNull CropAndResize.Method method,
double extrapolationValue,
INDArray output) |
CropAndResize(@NonNull INDArray image,
@NonNull INDArray cropBoxes,
@NonNull INDArray boxIndices,
@NonNull INDArray cropOutSize,
@NonNull CropAndResize.Method method,
double extrapolationValue,
INDArray output) |
CropAndResize(@NonNull INDArray image,
@NonNull INDArray cropBoxes,
@NonNull INDArray boxIndices,
@NonNull INDArray cropOutSize,
@NonNull CropAndResize.Method method,
double extrapolationValue,
INDArray output) |
CropAndResize(@NonNull INDArray image,
@NonNull INDArray cropBoxes,
@NonNull INDArray boxIndices,
@NonNull INDArray cropOutSize,
@NonNull CropAndResize.Method method,
double extrapolationValue,
INDArray output) |
CropAndResize(INDArray image,
INDArray cropBoxes,
INDArray boxIndices,
INDArray cropOutSize,
double extrapolationValue) |
ExtractImagePatches(@NonNull INDArray input,
@NonNull int[] kSizes,
@NonNull int[] strides,
@NonNull int[] rates,
boolean sameMode) |
ExtractImagePatches(INDArray input,
int kH,
int kW,
int sH,
int sW,
int rH,
int rW,
boolean sameMode) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
boolean preserveAspectRatio,
boolean antialias,
ImageResizeMethod method) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
boolean preserveAspectRatio,
boolean antialias,
ImageResizeMethod method) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
CoordinateTransformationMode coorMode,
NearestMode nearestMode,
boolean preserveAspectRatio,
boolean antialias) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
CoordinateTransformationMode coorMode,
NearestMode nearestMode,
boolean preserveAspectRatio,
boolean antialias) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
double bicubicCoefficient,
boolean exclude_outside,
boolean preserveAspectRatio) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
double bicubicCoefficient,
boolean exclude_outside,
boolean preserveAspectRatio) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
double bicubicCoefficient,
CoordinateTransformationMode coorMode,
boolean exclude_outside,
boolean preserveAspectRatio) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
double bicubicCoefficient,
CoordinateTransformationMode coorMode,
boolean exclude_outside,
boolean preserveAspectRatio) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
NearestMode nearestMode,
boolean preserveAspectRatio,
boolean antialias) |
ImageResize(@NonNull INDArray in,
@NonNull INDArray size,
NearestMode nearestMode,
boolean preserveAspectRatio,
boolean antialias) |
NonMaxSuppression(INDArray boxes,
INDArray scores,
int maxOutSize,
double iouThreshold,
double scoreThreshold) |
NonMaxSuppressionWithOverlaps(INDArray boxes,
INDArray scores,
int maxOutSize,
double iouThreshold,
double scoreThreshold) |
ResizeArea(@NonNull INDArray x,
INDArray z,
int height,
int width,
boolean alignCorners) |
ResizeArea(@NonNull INDArray x,
INDArray z,
int height,
int width,
boolean alignCorners) |
ResizeBicubic(@NonNull INDArray image,
INDArray size,
boolean alignCorners,
boolean alignPixelCenters) |
ResizeBicubic(@NonNull INDArray image,
INDArray size,
boolean alignCorners,
boolean alignPixelCenters) |
ResizeBilinear(@NonNull INDArray x,
INDArray z,
int height,
int width,
boolean alignCorners,
boolean halfPixelCenters) |
ResizeBilinear(@NonNull INDArray x,
INDArray z,
int height,
int width,
boolean alignCorners,
boolean halfPixelCenters) |
ResizeBilinear(INDArray input,
int height,
int width,
boolean alignCorners,
boolean halfPixelCenters) |
| Constructor and Description |
|---|
FirstIndex(INDArray x,
boolean keepDims,
@NonNull Condition condition,
int... dimension) |
FirstIndex(INDArray x,
@NonNull Condition condition,
boolean keepDims,
int... dimension) |
FirstIndex(INDArray x,
@NonNull Condition condition,
double eps,
int... dimension) |
FirstIndex(INDArray x,
@NonNull Condition condition,
int... dimension) |
LastIndex(INDArray in,
boolean keepDim,
Condition condition,
int... dimensions) |
LastIndex(INDArray x,
@NonNull Condition condition,
boolean keepDim,
int... dimensions) |
LastIndex(INDArray x,
@NonNull Condition condition,
double eps,
int... dimensions) |
LastIndex(INDArray x,
@NonNull Condition condition,
int... dimensions) |
| Constructor and Description |
|---|
ArgAmax(INDArray[] inputs) |
ArgAmax(INDArray[] inputs,
INDArray[] outputs) |
ArgAmax(INDArray[] inputs,
INDArray[] outputs) |
ArgAmax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgAmax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgAmax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgAmax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgAmax(INDArray[] inputs,
int[] dim) |
ArgAmin(INDArray[] inputs) |
ArgAmin(INDArray[] inputs,
INDArray[] outputs) |
ArgAmin(INDArray[] inputs,
INDArray[] outputs) |
ArgAmin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgAmin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgAmin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgAmin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgAmin(INDArray[] inputs,
int[] dim) |
ArgMax(INDArray arr) |
ArgMax(INDArray[] inputs) |
ArgMax(INDArray[] inputs,
boolean keepDims,
int[] dimensions) |
ArgMax(INDArray[] inputs,
INDArray[] outputs) |
ArgMax(INDArray[] inputs,
INDArray[] outputs) |
ArgMax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgMax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgMax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
ArgMax(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
ArgMax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgMax(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgMax(INDArray in,
boolean keepDims,
int[] dimensions) |
ArgMax(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMax(String opName,
INDArray input,
INDArray output,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(INDArray arr) |
ArgMin(INDArray[] inputs) |
ArgMin(INDArray[] inputs,
boolean keepDims,
int[] dimensions) |
ArgMin(INDArray[] inputs,
INDArray[] outputs) |
ArgMin(INDArray[] inputs,
INDArray[] outputs) |
ArgMin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgMin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
ArgMin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
ArgMin(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
ArgMin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgMin(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
ArgMin(INDArray in,
boolean keepDims,
int[] dimensions) |
ArgMin(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
ArgMin(String opName,
INDArray input,
INDArray output,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
| Constructor and Description |
|---|
ExternalErrorsFunction(SameDiff sd,
List<SDVariable> inputs,
Map<String,INDArray> gradients) |
| Constructor and Description |
|---|
AvgPooling2D(@NonNull INDArray input,
INDArray output,
@NonNull Pooling2DConfig config) |
AvgPooling2D(@NonNull INDArray input,
INDArray output,
@NonNull Pooling2DConfig config) |
AvgPooling2D(@NonNull INDArray input,
Pooling2DConfig config) |
AvgPooling3D(@NonNull INDArray input,
Pooling3DConfig pooling3DConfig) |
BatchNorm(INDArray input,
INDArray mean,
INDArray variance,
INDArray gamma,
INDArray beta,
double epsilon,
int... axis) |
BatchNorm(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputArrays,
boolean inPlace,
boolean applyGamma,
boolean applyBeta,
double epsilon,
int[] axis) |
BatchNorm(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputArrays,
boolean inPlace,
boolean applyGamma,
boolean applyBeta,
double epsilon,
int[] axis) |
BatchNormDerivative(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputArrays,
boolean inPlace,
boolean applyGamma,
boolean applyBeta,
double epsilon,
int[] axis) |
BatchNormDerivative(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputArrays,
boolean inPlace,
boolean applyGamma,
boolean applyBeta,
double epsilon,
int[] axis) |
Col2Im(@NonNull INDArray in,
@NonNull Conv2DConfig conv2DConfig) |
Col2Im(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputs,
Conv2DConfig conv2DConfig) |
Col2Im(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputs,
Conv2DConfig conv2DConfig) |
Conv1D(INDArray[] inputs,
INDArray[] outputs,
Conv1DConfig config) |
Conv1D(INDArray[] inputs,
INDArray[] outputs,
Conv1DConfig config) |
Conv1D(INDArray input,
INDArray weights,
INDArray bias,
Conv1DConfig config) |
Conv1D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv1DConfig config) |
Conv1D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv1DConfig config) |
Conv1D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv1DConfig config) |
Conv1DDerivative(INDArray[] inputs,
INDArray[] outputs,
Conv1DConfig config) |
Conv1DDerivative(INDArray[] inputs,
INDArray[] outputs,
Conv1DConfig config) |
Conv1DDerivative(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
@NonNull INDArray gradOut,
INDArray output,
@NonNull Conv1DConfig config) |
Conv1DDerivative(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
@NonNull INDArray gradOut,
INDArray output,
@NonNull Conv1DConfig config) |
Conv1DDerivative(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
@NonNull INDArray gradOut,
INDArray output,
@NonNull Conv1DConfig config) |
Conv1DDerivative(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
@NonNull INDArray gradOut,
INDArray output,
@NonNull Conv1DConfig config) |
Conv2D(INDArray[] inputs,
INDArray[] outputs,
Conv2DConfig config) |
Conv2D(INDArray[] inputs,
INDArray[] outputs,
Conv2DConfig config) |
Conv2D(INDArray layerInput,
INDArray weights,
INDArray bias,
Conv2DConfig config) |
Conv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv2DConfig config) |
Conv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv2DConfig config) |
Conv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv2DConfig config) |
Conv3D(INDArray[] inputs,
INDArray[] outputs,
Conv3DConfig config) |
Conv3D(INDArray[] inputs,
INDArray[] outputs,
Conv3DConfig config) |
Conv3D(INDArray input,
INDArray weights,
Conv3DConfig config) |
Conv3D(INDArray input,
INDArray weights,
INDArray bias,
Conv3DConfig config) |
Conv3D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv3DConfig config) |
Conv3D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv3DConfig config) |
Conv3D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv3DConfig config) |
DeConv2D(INDArray[] inputs,
INDArray[] outputs,
DeConv2DConfig config) |
DeConv2D(INDArray[] inputs,
INDArray[] outputs,
DeConv2DConfig config) |
DeConv2D(INDArray layerInput,
INDArray weights,
INDArray bias,
DeConv2DConfig config) |
DeConv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull DeConv2DConfig config) |
DeConv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull DeConv2DConfig config) |
DeConv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull DeConv2DConfig config) |
DeConv2DTF(INDArray[] inputs,
INDArray[] outputs,
DeConv2DConfig config) |
DeConv2DTF(INDArray[] inputs,
INDArray[] outputs,
DeConv2DConfig config) |
DeConv3D(INDArray[] inputs,
INDArray[] outputs,
DeConv3DConfig config) |
DeConv3D(INDArray[] inputs,
INDArray[] outputs,
DeConv3DConfig config) |
DeConv3D(INDArray input,
INDArray weights,
INDArray bias,
DeConv3DConfig config) |
DeConv3D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull DeConv3DConfig config) |
DeConv3D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull DeConv3DConfig config) |
DeConv3D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull DeConv3DConfig config) |
DepthToSpace(INDArray in,
INDArray out,
int blockSize,
DataFormat dataFormat) |
DepthToSpace(INDArray in,
int blockSize,
DataFormat dataFormat) |
DepthwiseConv2D(INDArray[] inputs,
INDArray[] outputs,
Conv2DConfig config) |
DepthwiseConv2D(INDArray[] inputs,
INDArray[] outputs,
Conv2DConfig config) |
DepthwiseConv2D(INDArray layerInput,
INDArray depthWeights,
INDArray bias,
Conv2DConfig config) |
DepthwiseConv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv2DConfig config) |
DepthwiseConv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv2DConfig config) |
DepthwiseConv2D(@NonNull INDArray input,
@NonNull INDArray weights,
INDArray bias,
INDArray output,
@NonNull Conv2DConfig config) |
Im2col(INDArray in,
Conv2DConfig conv2DConfig) |
Im2col(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputs,
Conv2DConfig conv2DConfig) |
Im2col(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputs,
Conv2DConfig conv2DConfig) |
LocalResponseNormalization(@NonNull INDArray input,
INDArray output,
@NonNull LocalResponseNormalizationConfig config) |
LocalResponseNormalization(@NonNull INDArray input,
INDArray output,
@NonNull LocalResponseNormalizationConfig config) |
LocalResponseNormalization(@NonNull INDArray input,
@NonNull LocalResponseNormalizationConfig LocalResponseNormalizationConfig) |
MaxPooling2D(INDArray input,
INDArray output,
@NonNull Pooling2DConfig config) |
MaxPooling2D(INDArray input,
@NonNull Pooling2DConfig config) |
MaxPooling3D(INDArray arrayInput,
INDArray arrayOutput,
Pooling3DConfig config) |
MaxPooling3D(INDArray input,
Pooling3DConfig pooling3DConfig) |
MaxPoolWithArgmax(@NonNull INDArray input,
INDArray output,
INDArray outArgMax,
@NonNull Pooling2DConfig config) |
MaxPoolWithArgmax(@NonNull INDArray input,
INDArray output,
INDArray outArgMax,
@NonNull Pooling2DConfig config) |
MaxPoolWithArgmax(@NonNull INDArray input,
@NonNull Pooling2DConfig config) |
Pooling2D(@NonNull INDArray[] inputs,
INDArray[] outputs,
@NonNull Pooling2DConfig config) |
Pooling2D(@NonNull INDArray[] inputs,
INDArray[] outputs,
@NonNull Pooling2DConfig config) |
Pooling2D(@NonNull INDArray input,
INDArray output,
@NonNull Pooling2DConfig config) |
Pooling2D(@NonNull INDArray input,
INDArray output,
@NonNull Pooling2DConfig config) |
Pooling2DDerivative(@NonNull INDArray input,
@NonNull INDArray grad,
INDArray output,
Pooling2DConfig config) |
Pooling2DDerivative(@NonNull INDArray input,
@NonNull INDArray grad,
INDArray output,
Pooling2DConfig config) |
Pooling2DDerivative(@NonNull INDArray input,
@NonNull INDArray grad,
INDArray output,
Pooling2DConfig config) |
Pooling3D(SameDiff sameDiff,
SDVariable[] inputs,
INDArray[] inputArrays,
INDArray[] outputs,
boolean inPlace,
Pooling3DConfig pooling3DConfig,
Pooling3D.Pooling3DType type) |
Pooling3D(SameDiff sameDiff,
SDVariable[] inputs,
INDArray[] inputArrays,
INDArray[] outputs,
boolean inPlace,
Pooling3DConfig pooling3DConfig,
Pooling3D.Pooling3DType type) |
Pooling3DDerivative(SameDiff sameDiff,
SDVariable[] inputs,
INDArray[] inputArrays,
INDArray[] outputs,
boolean inPlace,
Pooling3DConfig pooling3DConfig,
Pooling3D.Pooling3DType type) |
Pooling3DDerivative(SameDiff sameDiff,
SDVariable[] inputs,
INDArray[] inputArrays,
INDArray[] outputs,
boolean inPlace,
Pooling3DConfig pooling3DConfig,
Pooling3D.Pooling3DType type) |
SConv2D(INDArray[] inputs,
INDArray[] outputs,
Conv2DConfig config) |
SConv2D(INDArray[] inputs,
INDArray[] outputs,
Conv2DConfig config) |
SConv2D(@NonNull INDArray layerInput,
@NonNull INDArray depthWeights,
INDArray pointWeights,
@NonNull Conv2DConfig Conv2DConfig) |
SConv2D(@NonNull INDArray layerInput,
@NonNull INDArray depthWeights,
INDArray pointWeights,
@NonNull Conv2DConfig Conv2DConfig) |
SConv2D(@NonNull INDArray layerInput,
@NonNull INDArray depthWeights,
INDArray pointWeights,
@NonNull Conv2DConfig Conv2DConfig) |
SConv2D(INDArray layerInput,
INDArray depthWeights,
INDArray pointWeights,
INDArray bias,
Conv2DConfig config) |
SpaceToDepth(INDArray in,
INDArray out,
int blockSize,
DataFormat dataFormat) |
SpaceToDepth(INDArray x,
int blockSize,
DataFormat dataFormat) |
Upsampling2d(INDArray input,
int scale) |
Upsampling2d(INDArray input,
int scaleH,
int scaleW,
boolean nchw) |
Upsampling3d(INDArray input,
boolean ncdhw,
int scaleH,
int scaleW,
int scaleD) |
| Constructor and Description |
|---|
GRU(@NonNull INDArray x,
@NonNull INDArray hI,
@NonNull INDArray Wx,
@NonNull INDArray Wh,
@NonNull INDArray biases) |
GRU(@NonNull INDArray x,
@NonNull INDArray hI,
@NonNull INDArray Wx,
@NonNull INDArray Wh,
@NonNull INDArray biases) |
GRU(@NonNull INDArray x,
@NonNull INDArray hI,
@NonNull INDArray Wx,
@NonNull INDArray Wh,
@NonNull INDArray biases) |
GRU(@NonNull INDArray x,
@NonNull INDArray hI,
@NonNull INDArray Wx,
@NonNull INDArray Wh,
@NonNull INDArray biases) |
GRU(@NonNull INDArray x,
@NonNull INDArray hI,
@NonNull INDArray Wx,
@NonNull INDArray Wh,
@NonNull INDArray biases) |
GRUCell(INDArray x,
INDArray hLast,
GRUWeights gruWeights) |
LSTMBlock(INDArray x,
INDArray cLast,
INDArray yLast,
INDArray maxTSLength,
LSTMWeights lstmWeights,
LSTMConfiguration lstmConfiguration) |
LSTMBlockCell(INDArray x,
INDArray cLast,
INDArray yLast,
LSTMWeights lstmWeights,
LSTMConfiguration lstmConfiguration) |
LSTMLayer(INDArray x,
INDArray cLast,
INDArray yLast,
INDArray maxTSLength,
LSTMLayerWeights lstmWeights,
LSTMLayerConfig LSTMLayerConfig) |
SRU(INDArray x,
INDArray initialC,
INDArray mask,
SRUWeights sruWeights) |
SRU(INDArray x,
INDArray initialC,
SRUWeights sruWeights) |
SRUCell(INDArray x,
INDArray cLast,
SRUWeights sruWeights) |
| Modifier and Type | Method and Description |
|---|---|
INDArray[] |
LSTMLayerWeights.argsWithInputs(INDArray... inputs) |
INDArray[] |
RNNWeights.argsWithInputs(INDArray... inputs) |
INDArray[] |
GRUWeights.arrayArgs() |
INDArray[] |
SRUWeights.arrayArgs() |
INDArray[] |
LSTMLayerWeights.arrayArgs() |
INDArray[] |
LSTMWeights.arrayArgs() |
abstract INDArray[] |
RNNWeights.arrayArgs() |
| Modifier and Type | Method and Description |
|---|---|
INDArray[] |
LSTMLayerWeights.argsWithInputs(INDArray... inputs) |
INDArray[] |
RNNWeights.argsWithInputs(INDArray... inputs) |
| Modifier and Type | Method and Description |
|---|---|
protected static INDArray |
BaseLoss.getWeights(INDArray weights,
INDArray predictions) |
| Modifier and Type | Method and Description |
|---|---|
protected static INDArray |
BaseLoss.getWeights(INDArray weights,
INDArray predictions) |
| Constructor and Description |
|---|
AbsoluteDifferenceLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce) |
BaseLoss(@NonNull LossReduce lossReduce,
@NonNull INDArray predictions,
INDArray weights,
@NonNull INDArray labels) |
BaseLoss(@NonNull LossReduce lossReduce,
@NonNull INDArray predictions,
INDArray weights,
@NonNull INDArray labels) |
BaseLoss(@NonNull LossReduce lossReduce,
@NonNull INDArray predictions,
INDArray weights,
@NonNull INDArray labels) |
CosineDistanceLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
int dimension) |
CtcLoss(INDArray targetLabels,
INDArray logitInputs,
INDArray targetLabelLengths,
INDArray logitInputLengths) |
HingeLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce) |
HuberLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double delta) |
L2Loss(INDArray var) |
LogLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double epsilon) |
LogPoissonLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
boolean full) |
MeanPairwiseSquaredErrorLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce) |
MeanSquaredErrorLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce) |
SigmoidCrossEntropyLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double labelSmoothing) |
SoftmaxCrossEntropyLoss(INDArray labels,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double labelSmoothing) |
SparseSoftmaxCrossEntropyLossWithLogits(@NonNull INDArray logits,
@NonNull INDArray labels) |
SparseSoftmaxCrossEntropyLossWithLogits(@NonNull INDArray logits,
@NonNull INDArray labels) |
WeightedCrossEntropyLoss(INDArray targets,
INDArray inputs,
INDArray weights) |
WeightedCrossEntropyLoss(@NonNull LossReduce lossReduce,
@NonNull INDArray predictions,
INDArray weights,
@NonNull INDArray labels) |
WeightedCrossEntropyLoss(@NonNull LossReduce lossReduce,
@NonNull INDArray predictions,
INDArray weights,
@NonNull INDArray labels) |
WeightedCrossEntropyLoss(@NonNull LossReduce lossReduce,
@NonNull INDArray predictions,
INDArray weights,
@NonNull INDArray labels) |
| Constructor and Description |
|---|
BaseMetaOp(INDArray x,
INDArray y) |
InvertedPredicateMetaOp(INDArray x,
INDArray y) |
PostulateMetaOp(INDArray x,
INDArray y) |
PredicateMetaOp(INDArray x,
INDArray y) |
ReduceMetaOp(INDArray x,
INDArray y) |
| Constructor and Description |
|---|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(@NonNull INDArray target,
@NonNull INDArray context,
@NonNull INDArray lockedWords,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
@NonNull INDArray indices,
@NonNull INDArray codes,
int nsRounds,
@NonNull INDArray alpha,
@NonNull INDArray nextRandom,
@NonNull INDArray inferenceVector,
@NonNull INDArray numLabels,
boolean trainWords,
int numWorkers)
full constructor
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
@NonNull int[] indices,
@NonNull byte[] codes,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
hs round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
@NonNull int[] indices,
@NonNull byte[] codes,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
hs round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
@NonNull int[] indices,
@NonNull byte[] codes,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
hs round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
@NonNull int[] indices,
@NonNull byte[] codes,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
hs round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
ns round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
ns round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
ns round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
ns round
|
CbowRound(int target,
@NonNull int[] context,
@NonNull int[] lockedWords,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long nextRandom,
@NonNull INDArray inferenceVector,
int numLabels)
ns round
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(@NonNull INDArray target,
@NonNull INDArray ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
@NonNull INDArray indices,
@NonNull INDArray codes,
@NonNull INDArray alpha,
@NonNull INDArray randomValue,
INDArray inferenceVector,
boolean preciseMode,
int numWorkers)
full constructor
|
SkipGramRound(int target,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
int[] indices,
byte[] codes,
double alpha,
long randomValue,
INDArray inferenceVector)
sg hs round
|
SkipGramRound(int target,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
int[] indices,
byte[] codes,
double alpha,
long randomValue,
INDArray inferenceVector)
sg hs round
|
SkipGramRound(int target,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
int[] indices,
byte[] codes,
double alpha,
long randomValue,
INDArray inferenceVector)
sg hs round
|
SkipGramRound(int target,
@NonNull INDArray syn0,
@NonNull INDArray syn1,
@NonNull INDArray expTable,
int[] indices,
byte[] codes,
double alpha,
long randomValue,
INDArray inferenceVector)
sg hs round
|
SkipGramRound(int target,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long randomValue,
INDArray inferenceVector)
sg ns round
|
SkipGramRound(int target,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long randomValue,
INDArray inferenceVector)
sg ns round
|
SkipGramRound(int target,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long randomValue,
INDArray inferenceVector)
sg ns round
|
SkipGramRound(int target,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long randomValue,
INDArray inferenceVector)
sg ns round
|
SkipGramRound(int target,
int ngStarter,
@NonNull INDArray syn0,
@NonNull INDArray syn1Neg,
@NonNull INDArray expTable,
@NonNull INDArray negTable,
int nsRounds,
double alpha,
long randomValue,
INDArray inferenceVector)
sg ns round
|
| Constructor and Description |
|---|
HashCode(@NonNull INDArray array) |
HashCode(@NonNull INDArray array,
@NonNull INDArray result) |
HashCode(@NonNull INDArray array,
@NonNull INDArray result) |
Mmul(INDArray x,
INDArray y) |
Mmul(INDArray x,
INDArray y,
boolean transposeX,
boolean transposeY,
boolean transposeZ) |
Mmul(INDArray x,
INDArray y,
double alpha,
double beta) |
Mmul(INDArray x,
INDArray y,
double alpha,
double beta,
boolean transposeX,
boolean transposeY,
boolean transposeZ) |
Mmul(INDArray x,
INDArray y,
INDArray z,
double alpha,
double beta,
MMulTranspose mt) |
Mmul(INDArray x,
INDArray y,
INDArray z,
MMulTranspose mt) |
MmulBp(INDArray x,
INDArray y,
INDArray eps,
INDArray dldx,
INDArray dldy,
MMulTranspose mt) |
Moments(@NonNull INDArray input,
boolean keepDims,
int... dimensions) |
Moments(INDArray input,
INDArray axes,
boolean keepDims) |
Moments(INDArray in,
INDArray outMean,
INDArray outStd,
int... axes) |
Moments(@NonNull INDArray input,
int... dimensions) |
Moments(INDArray input,
int[] axes,
boolean keepDims) |
NormalizeMoments(INDArray counts,
INDArray means,
INDArray variances,
double shift) |
NormalizeMoments(INDArray counts,
INDArray ssSum,
INDArray ssSqSum,
INDArray outMean,
INDArray outVar) |
SufficientStatistics(@NonNull INDArray x,
@NonNull INDArray axes) |
SufficientStatistics(@NonNull INDArray x,
@NonNull INDArray axes) |
SufficientStatistics(@NonNull INDArray x,
@NonNull INDArray axes,
INDArray shift) |
SufficientStatistics(@NonNull INDArray x,
@NonNull INDArray axes,
INDArray shift) |
SufficientStatistics(@NonNull INDArray x,
@NonNull INDArray axes,
INDArray shift) |
TensorMmul(INDArray x,
INDArray y,
INDArray z,
int[][] axes)
Initialize with the given
input, pairwise transform, result, and number
of elements
|
TensorMmul(INDArray x,
INDArray y,
int[][] axes) |
TensorMmul(INDArray x,
INDArray y,
int[] dimensionsX,
int[] dimensionsY,
boolean transposeX,
boolean transposeY,
boolean transposeZ) |
TensorMmulBp(INDArray x,
INDArray y,
INDArray gradAtOutput,
INDArray dldx,
INDArray dldy,
int[][] axes) |
TensorMmulBp(INDArray x,
INDArray y,
INDArray gradAtOutput,
INDArray dldx,
INDArray dldy,
int[] axesX,
int[] axesY) |
TensorMmulBp(INDArray x,
INDArray y,
INDArray gradAtOutput,
int[][] axes) |
TensorMmulBp(INDArray x,
INDArray y,
INDArray gradAtOutput,
int[] axesX,
int[] axesY) |
ZeroFraction(@NonNull INDArray input) |
| Constructor and Description |
|---|
All(INDArray x) |
All(INDArray x,
boolean keepDims,
int... dimensions) |
All(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
All(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
All(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
All(INDArray x,
INDArray z,
int... dimensions) |
All(INDArray x,
int... axis) |
Any(INDArray x) |
Any(INDArray x,
boolean keepDims,
int... dimensions) |
Any(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
Any(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
Any(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Any(INDArray x,
INDArray z,
int... dimensions) |
Any(INDArray x,
int... dimensions) |
IsInf(INDArray x) |
IsInf(INDArray x,
boolean keepDims,
int... dimensions) |
IsInf(INDArray x,
INDArray z) |
IsInf(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
IsInf(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
IsInf(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
IsInf(INDArray x,
INDArray z,
int... dimensions) |
IsInf(INDArray x,
int... dimensions) |
IsNaN(INDArray x) |
IsNaN(INDArray x,
boolean keepDims,
int... dimensions) |
IsNaN(INDArray x,
INDArray z) |
IsNaN(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
IsNaN(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
IsNaN(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
IsNaN(INDArray x,
INDArray z,
int... dimensions) |
IsNaN(INDArray x,
int... dimensions) |
| Constructor and Description |
|---|
BaseReductionBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
BaseReductionBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
BaseReductionBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
BaseReductionBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
CumProdBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
boolean exclusive,
boolean reverse,
int... dimensions) |
CumProdBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean exclusive,
boolean reverse,
int... axis) |
CumProdBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions,
boolean exclusive,
boolean reverse) |
CumProdBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
boolean exclusive,
boolean reverse,
int... dimensions) |
CumProdBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
boolean exclusive,
boolean reverse,
int... dimensions) |
CumSumBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
boolean exclusive,
boolean reverse,
int... dimensions) |
CumSumBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean exclusive,
boolean reverse,
int... axis) |
CumSumBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions,
boolean exclusive,
boolean reverse) |
CumSumBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
boolean exclusive,
boolean reverse,
int... dimensions) |
CumSumBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
boolean exclusive,
boolean reverse,
int... dimensions) |
DotBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
DotBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
DotBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
DotBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray outputX,
INDArray outputY,
boolean keepDims,
int... dimensions) |
MaxBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
MaxBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
MaxBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
MaxBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
MeanBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
MeanBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
MeanBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
MeanBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
MinBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
MinBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
MinBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
MinBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
Norm1Bp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
Norm1Bp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
Norm1Bp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
Norm1Bp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
Norm2Bp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
Norm2Bp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
Norm2Bp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
Norm2Bp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
NormMaxBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
NormMaxBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
NormMaxBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
NormMaxBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
PowBp(INDArray x,
INDArray y,
INDArray dLdz,
INDArray dLdx,
INDArray dLdy) |
ProdBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
ProdBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
ProdBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
ProdBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
SquaredNormBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
SquaredNormBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
SquaredNormBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
int... dimensions) |
SquaredNormBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
int... dimensions) |
StandardDeviationBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean biasCorrected,
boolean keepDims,
int... dimensions) |
StandardDeviationBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions,
boolean biasCorrected) |
StandardDeviationBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
boolean biasCorrected,
int... dimensions) |
StandardDeviationBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
boolean biasCorrected,
int... dimensions) |
SumBp(INDArray preReduceInput,
INDArray dLdOut,
INDArray dLdIn,
boolean keepDims) |
SumBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions) |
VarianceBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean biasCorrected,
boolean keepDims,
int... dimensions) |
VarianceBp(INDArray origInput,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
INDArray dimensions,
boolean biasCorrected) |
VarianceBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output,
boolean keepDims,
boolean biasCorrected,
int... dimensions) |
VarianceBp(INDArray origInput1,
INDArray origInput2,
INDArray gradAtOutput,
INDArray output1,
INDArray output2,
boolean keepDims,
boolean biasCorrected,
int... dimensions) |
| Constructor and Description |
|---|
BaseDynamicCustomBoolReduction(INDArray[] inputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomBoolReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomBoolReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomBoolReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomBoolReduction(String opName,
INDArray input,
INDArray output,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomIndexReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
BaseDynamicCustomIndexReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int... dimensions) |
BaseDynamicCustomIndexReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomIndexReduction(String opName,
INDArray input,
INDArray output,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomLongReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomLongReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomLongReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomLongReduction(String opName,
INDArray input,
INDArray output,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(INDArray[] inputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomReduction(INDArray[] input,
INDArray[] output,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomReduction(INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
int[] dimensions) |
BaseDynamicCustomReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BaseDynamicCustomReduction(String opName,
INDArray input,
INDArray output,
List<Double> tArguments,
int[] iArguments,
boolean keepDims,
boolean isComplex,
boolean isEmptyReduce,
int[] dimensions) |
BatchMmul(INDArray[] matricesA,
INDArray[] matricesB,
boolean transposeA,
boolean transposeB) |
BatchMmul(INDArray[] matricesA,
INDArray[] matricesB,
boolean transposeA,
boolean transposeB) |
LogSumExp(INDArray x,
boolean keepDim,
int... dimensions) |
LogSumExp(INDArray x,
INDArray z,
boolean keepDim,
int... dimensions) |
LogSumExp(INDArray x,
int... dimensions) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
Norm1.noOp() |
INDArray |
NormMax.noOp() |
INDArray |
Norm2.noOp() |
| Constructor and Description |
|---|
AMean(INDArray in,
boolean keepDims,
int[] dimensions) |
AMean(INDArray in,
INDArray dimensions,
boolean keepDims) |
AMean(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
AMean(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
AMean(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
AMean(INDArray x,
INDArray z,
int... dimensions) |
AMean(INDArray x,
int... dimensions) |
AMean(INDArray in,
int[] dimensions,
boolean keepDims) |
Entropy(INDArray in,
boolean keepDims,
int[] dimensions) |
Entropy(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
Entropy(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
Entropy(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Entropy(INDArray x,
INDArray z,
int... dimensions) |
Entropy(INDArray x,
int... dimensions) |
Entropy(INDArray in,
int[] dimensions,
boolean keepDims) |
LogEntropy(INDArray in,
boolean keepDims,
int[] dimensions) |
LogEntropy(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
LogEntropy(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
LogEntropy(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
LogEntropy(INDArray x,
INDArray z,
int... dimensions) |
LogEntropy(INDArray x,
int... dimensions) |
LogEntropy(INDArray in,
int[] dimensions,
boolean keepDims) |
Mean(INDArray x,
boolean keepDims,
int... dimensions) |
Mean(INDArray x,
INDArray z,
boolean keepDims,
int... dimensions) |
Mean(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
Mean(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Mean(INDArray x,
INDArray z,
int... dimensions) |
Mean(INDArray x,
int... dimensions) |
Mean(INDArray in,
int[] dimensions,
boolean keepDims) |
Norm1(INDArray x,
boolean keepDims,
int... dimensions) |
Norm1(INDArray in,
INDArray dimensions,
boolean keepDims) |
Norm1(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
Norm1(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
Norm1(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Norm1(INDArray x,
INDArray z,
int... dimensions) |
Norm1(INDArray x,
int... dimensions) |
Norm1(INDArray in,
int[] dimensions,
boolean keepDims) |
Norm2(INDArray x,
boolean keepDims,
int... dimensions) |
Norm2(INDArray in,
INDArray indArray,
boolean keepDims) |
Norm2(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
Norm2(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
Norm2(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Norm2(INDArray x,
INDArray z,
int... dimensions) |
Norm2(INDArray x,
int... dimensions) |
Norm2(INDArray in,
int[] dimensions,
boolean keepDims) |
NormMax(INDArray x,
boolean keepDims,
int... dimensions) |
NormMax(INDArray in,
INDArray indArray,
boolean keepDims) |
NormMax(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
NormMax(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
NormMax(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
NormMax(INDArray x,
INDArray z,
int... dimensions) |
NormMax(INDArray x,
int... dimensions) |
NormMax(INDArray in,
int[] dimensions,
boolean keepDims) |
ShannonEntropy(INDArray in,
boolean keepDims,
int[] dimensions) |
ShannonEntropy(INDArray in,
INDArray dimensions,
boolean keepDims) |
ShannonEntropy(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
ShannonEntropy(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
ShannonEntropy(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
ShannonEntropy(INDArray x,
INDArray z,
int... dimensions) |
ShannonEntropy(INDArray x,
int... dimensions) |
ShannonEntropy(INDArray in,
int[] dimensions,
boolean keepDims) |
SquaredNorm(INDArray input,
boolean keepDims,
int... dimensions) |
SquaredNorm(INDArray input,
INDArray output,
boolean keepDims,
int... dimensions) |
SquaredNorm(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
SquaredNorm(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
SquaredNorm(INDArray x,
INDArray z,
int... dimensions) |
SquaredNorm(INDArray x,
int... dimensions) |
SquaredNorm(INDArray in,
int[] dimensions,
boolean keepDims) |
| Constructor and Description |
|---|
CountNonZero(INDArray in,
boolean keepDims,
int[] dimensions) |
CountNonZero(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
CountNonZero(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
CountNonZero(INDArray x,
INDArray z,
int... dimensions) |
CountNonZero(INDArray x,
int... dimensions) |
CountNonZero(INDArray in,
int[] dimensions,
boolean keepDims) |
CountZero(INDArray x,
boolean keepDims,
int... dimensions) |
CountZero(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
CountZero(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
CountZero(INDArray x,
INDArray z,
int... dimensions) |
CountZero(INDArray x,
int... dimensions) |
CountZero(INDArray in,
int[] dimensions,
boolean keepDims) |
MatchCondition(INDArray x,
boolean keepDims,
double compare,
double eps,
int... dimensions) |
MatchCondition(INDArray x,
boolean keepDims,
double compare,
double eps,
int mode,
int... dimensions) |
MatchCondition(INDArray x,
boolean keepDims,
double compare,
int... dimensions) |
MatchCondition(INDArray x,
Condition condition,
boolean keepDims,
int... dimensions) |
MatchCondition(INDArray x,
Condition condition,
int... dimensions) |
MatchCondition(INDArray x,
double eps,
Condition condition,
int... dimensions) |
MatchCondition(INDArray x,
double compare,
double eps,
int... dimensions) |
MatchCondition(INDArray x,
double compare,
double eps,
int mode,
int... dimensions) |
MatchCondition(INDArray x,
double compare,
int... dimensions) |
MatchCondition(INDArray x,
INDArray z,
double compare,
double eps,
int... dimensions) |
MatchCondition(INDArray x,
INDArray z,
double compare,
double eps,
int mode,
int... dimensions) |
MatchCondition(INDArray x,
INDArray z,
double compare,
int... dimensions) |
MatchCondition(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double compare) |
MatchCondition(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double compare,
double eps) |
MatchCondition(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double compare,
double eps,
int mode) |
MatchCondition(INDArray x,
INDArray y,
INDArray z,
double compare,
double eps,
int... dimensions) |
MatchCondition(INDArray x,
INDArray y,
INDArray z,
double compare,
double eps,
int mode,
int... dimensions) |
MatchCondition(INDArray x,
INDArray y,
INDArray z,
double compare,
int... dimensions) |
| Constructor and Description |
|---|
AMax(INDArray in,
boolean keepDims,
int[] dimensions) |
AMax(INDArray in,
INDArray dimensions,
boolean keepDims) |
AMax(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
AMax(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
AMax(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
AMax(INDArray x,
INDArray z,
int... dimensions) |
AMax(INDArray x,
int... dimensions) |
AMax(INDArray in,
int[] dimensions,
boolean keepDims) |
AMin(INDArray in,
boolean keepDims,
int[] dimensions) |
AMin(INDArray in,
INDArray dimensions,
boolean keepDims) |
AMin(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
AMin(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
AMin(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
AMin(INDArray x,
INDArray z,
int... dimensions) |
AMin(INDArray x,
int... dimensions) |
AMin(INDArray in,
int[] dimensions,
boolean keepDims) |
ASum(INDArray in,
boolean keepDims,
int[] dimensions) |
ASum(INDArray x,
INDArray z,
boolean keepDims,
int[] dimensions) |
ASum(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
ASum(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
ASum(INDArray x,
INDArray z,
int... dimensions) |
ASum(INDArray x,
int... dimensions) |
ASum(INDArray in,
int[] dimensions,
boolean keepDims) |
Max(INDArray x,
boolean keepDims,
int... axis) |
Max(INDArray x,
INDArray z,
boolean keepDims,
int... dimensions) |
Max(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
Max(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Max(INDArray x,
INDArray z,
int... axis) |
Max(INDArray x,
int... axis)
Initialize with the given
input, pairwise transform, result, and number
of elements
|
Max(INDArray in,
int[] dimensions,
boolean keepDims) |
Min(INDArray x,
boolean keepDims,
int... dimensions) |
Min(INDArray x,
INDArray z,
boolean keepDims,
int... dimensions) |
Min(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
Min(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Min(INDArray x,
INDArray z,
int... dimensions) |
Min(INDArray x,
int... dimensions) |
Min(INDArray in,
int[] dimensions,
boolean keepDims) |
Prod(INDArray x,
boolean keepDims,
int... dimensions) |
Prod(INDArray x,
INDArray z,
boolean keepDims,
int... dimensions) |
Prod(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
Prod(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Prod(INDArray x,
INDArray z,
int... dimensions) |
Prod(INDArray x,
int... dimensions) |
Prod(INDArray in,
int[] dimensions,
boolean keepDims) |
Sum(INDArray x,
boolean keepDims,
int... dimensions) |
Sum(INDArray x,
INDArray z,
boolean keepDims,
int... dimensions) |
Sum(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions) |
Sum(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
Sum(INDArray x,
INDArray z,
int... dimensions) |
Sum(INDArray x,
int... dimensions) |
Sum(INDArray in,
int[] dimensions,
boolean keepDims) |
| Constructor and Description |
|---|
BaseReduce3Op(INDArray x,
INDArray y,
boolean allDistances,
int... dimensions) |
BaseReduce3Op(INDArray x,
INDArray y,
INDArray z) |
BaseReduce3Op(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
BaseReduce3Op(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
BaseReduce3Op(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
BaseReduce3Op(INDArray x,
INDArray y,
int... dimensions) |
CosineDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
CosineDistance(INDArray x,
INDArray y,
boolean allDistances,
int... dimension) |
CosineDistance(INDArray x,
INDArray y,
INDArray z) |
CosineDistance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
CosineDistance(INDArray x,
INDArray y,
INDArray z,
boolean allDistances,
int... dimension) |
CosineDistance(INDArray x,
INDArray y,
INDArray z,
int... dimension) |
CosineDistance(INDArray x,
INDArray y,
int... dimension) |
CosineSimilarity(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
CosineSimilarity(INDArray x,
INDArray y,
boolean allDistances,
int... dimension) |
CosineSimilarity(INDArray x,
INDArray y,
INDArray z) |
CosineSimilarity(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
CosineSimilarity(INDArray x,
INDArray y,
INDArray z,
boolean allDistances,
int... dimension) |
CosineSimilarity(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
CosineSimilarity(INDArray x,
INDArray y,
int... dimensions) |
Dot(INDArray x,
INDArray y,
boolean allDistances,
int... dimensions) |
Dot(INDArray x,
INDArray y,
INDArray z) |
Dot(INDArray x,
INDArray y,
INDArray z,
boolean newFormat,
boolean keepDims,
int... dimensions) |
Dot(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
Dot(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Full array dot product reduction, optionally along specified dimensions.
See wikipedia for details. |
Dot(INDArray x,
INDArray y,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
boolean allDistances,
double eps,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
boolean allDistances,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
double eps,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
double eps,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
double eps,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z,
double eps) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z,
double eps,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
EqualsWithEps(INDArray x,
INDArray y,
int... dimensions) |
EuclideanDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
EuclideanDistance(INDArray x,
INDArray y,
boolean allDistances,
int... dimensions) |
EuclideanDistance(INDArray x,
INDArray y,
INDArray z) |
EuclideanDistance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
EuclideanDistance(INDArray x,
INDArray y,
INDArray z,
boolean allDistances,
int... dimensions) |
EuclideanDistance(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
EuclideanDistance(INDArray x,
INDArray y,
int... dimensions) |
HammingDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
HammingDistance(INDArray x,
INDArray y,
boolean allDistances,
int... dimensions) |
HammingDistance(INDArray x,
INDArray y,
INDArray z) |
HammingDistance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
HammingDistance(INDArray x,
INDArray y,
INDArray z,
boolean allDistances,
int... dimensions) |
HammingDistance(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
HammingDistance(INDArray x,
INDArray y,
int... dimensions) |
JaccardDistance(INDArray x,
INDArray y,
boolean allDistances) |
JaccardDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
JaccardDistance(INDArray x,
INDArray y,
boolean allDistances,
int... dimensions) |
JaccardDistance(INDArray x,
INDArray y,
INDArray z) |
JaccardDistance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
JaccardDistance(INDArray x,
INDArray y,
INDArray z,
boolean allDistances,
int... dimensions) |
JaccardDistance(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
JaccardDistance(INDArray x,
INDArray y,
int... dimensions) |
ManhattanDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int[] dimensions) |
ManhattanDistance(INDArray x,
INDArray y,
boolean allDistances,
int... dimensions) |
ManhattanDistance(INDArray x,
INDArray y,
INDArray z) |
ManhattanDistance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
boolean allDistances,
int... dimensions) |
ManhattanDistance(INDArray x,
INDArray y,
INDArray z,
boolean allDistances,
int... dimensions) |
ManhattanDistance(INDArray x,
INDArray y,
INDArray z,
int... dimensions) |
ManhattanDistance(INDArray x,
INDArray y,
int... dimensions) |
| Constructor and Description |
|---|
ScatterAdd(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterAdd(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterAdd(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterDiv(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterDiv(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterDiv(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMax(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMax(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMax(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMin(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMin(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMin(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMul(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMul(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterMul(@NonNull INDArray ref,
@NonNull INDArray indices,
@NonNull INDArray update) |
ScatterSub(INDArray ref,
INDArray indices,
INDArray update) |
ScatterUpdate(INDArray ref,
INDArray indices,
INDArray update) |
| Modifier and Type | Method and Description |
|---|---|
protected static boolean |
MergeMaxIndex.areEqualShapes(INDArray... inputs) |
| Constructor and Description |
|---|
Concat(INDArray[] arrays,
int concatDimension) |
Concat(int concatDimension,
INDArray... arrays) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
@NonNull DataType dataType) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
@NonNull DataType dataType) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights,
Integer numClasses) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights,
Integer numClasses) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights,
Integer numClasses) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights,
Integer numClasses,
@NonNull DataType dataType) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights,
Integer numClasses,
@NonNull DataType dataType) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
INDArray weights,
Integer numClasses,
@NonNull DataType dataType) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
int numClasses) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
int numClasses) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
Integer numClasses,
@NonNull DataType dataType) |
ConfusionMatrix(@NonNull INDArray labels,
@NonNull INDArray predicted,
Integer numClasses,
@NonNull DataType dataType) |
Create(INDArray shape,
boolean initialize,
DataType dataType) |
Create(@NonNull INDArray shape,
char order,
boolean initialize,
DataType dataType) |
Create(INDArray shape,
DataType dataType) |
Create(INDArray shape,
DataType dataType,
String order,
boolean initialize) |
Cross(INDArray a,
INDArray b) |
Cross(INDArray a,
INDArray b,
INDArray out) |
Diag(@NonNull INDArray input) |
Diag(@NonNull INDArray input,
@NonNull INDArray output) |
Diag(@NonNull INDArray input,
@NonNull INDArray output) |
DiagPart(INDArray in) |
DiagPart(INDArray in,
INDArray out) |
ExpandDims(INDArray[] inputs,
INDArray[] outputs) |
ExpandDims(INDArray[] inputs,
INDArray[] outputs) |
ExpandDims(INDArray x,
int axis) |
Eye(@NonNull INDArray rows) |
Eye(@NonNull INDArray rows,
@NonNull INDArray columns) |
Eye(@NonNull INDArray rows,
@NonNull INDArray columns) |
Flatten2D(INDArray in,
long axis) |
Gather(INDArray df,
INDArray indexes,
int axis) |
Gather(INDArray df,
int[] indexes,
int axis) |
GatherNd(INDArray[] inputs,
INDArray[] outputs) |
GatherNd(INDArray[] inputs,
INDArray[] outputs) |
GatherNd(INDArray df,
INDArray indices) |
Linspace(DataType dataType,
INDArray start,
INDArray stop,
INDArray number) |
Linspace(@NonNull INDArray start,
@NonNull INDArray stop,
@NonNull INDArray number,
@NonNull DataType dataType) |
Linspace(@NonNull INDArray start,
@NonNull INDArray stop,
@NonNull INDArray number,
@NonNull DataType dataType) |
Linspace(@NonNull INDArray start,
@NonNull INDArray stop,
@NonNull INDArray number,
@NonNull DataType dataType) |
MergeAvg(INDArray... inputs) |
MergeMax(INDArray... inputs) |
MergeMaxIndex(INDArray... inputs) |
MergeMaxIndex(@NonNull INDArray[] x,
@NonNull DataType dataType) |
MeshGrid(@NonNull INDArray[] inputs,
boolean cartesian) |
OneHot(INDArray indices,
INDArray output,
int depth) |
OneHot(INDArray indices,
INDArray output,
int depth,
int axis,
double on,
double off) |
OneHot(INDArray indices,
int depth) |
OneHot(INDArray indices,
int depth,
int axis,
double on,
double off) |
OneHot(INDArray indices,
int depth,
int axis,
double on,
double off,
DataType dataType) |
OnesAs(@NonNull INDArray input) |
OnesAs(@NonNull INDArray input,
DataType dataType) |
OnesLike(@NonNull INDArray input) |
OnesLike(@NonNull INDArray input,
DataType dataType) |
ParallelStack(INDArray[] inputs) |
Permute(INDArray input,
INDArray result,
int... permuteDims) |
Permute(INDArray input,
int... permuteDims) |
Rank(INDArray indArray) |
Repeat(INDArray[] inputs,
INDArray[] outputs,
int axis) |
Repeat(INDArray[] inputs,
INDArray[] outputs,
int axis) |
Repeat(INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
int axis) |
Repeat(INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments,
int axis) |
Repeat(INDArray input,
INDArray repeats,
int axis) |
Reshape(INDArray in,
INDArray shape) |
Reshape(@NonNull INDArray in,
@NonNull INDArray shape,
INDArray out) |
Reshape(@NonNull INDArray in,
@NonNull INDArray shape,
INDArray out) |
Reshape(@NonNull INDArray in,
@NonNull INDArray shape,
INDArray out) |
Reshape(INDArray in,
long... shape) |
SequenceMask(@NonNull INDArray input,
@NonNull DataType dataType) |
SequenceMask(@NonNull INDArray input,
INDArray maxLength,
@NonNull DataType dataType) |
SequenceMask(@NonNull INDArray input,
INDArray maxLength,
@NonNull DataType dataType) |
SequenceMask(@NonNull INDArray input,
int maxLen,
DataType dataType) |
SetShape(INDArray in,
INDArray shape) |
SetShape(INDArray in,
INDArray shape,
INDArray out) |
Shape(INDArray in) |
Shape(INDArray in,
INDArray out) |
Size(INDArray in) |
SizeAt(INDArray input,
INDArray output,
int dimension) |
SizeAt(INDArray input,
int dimension) |
Slice(@NonNull INDArray input,
@NonNull INDArray begin,
@NonNull INDArray end) |
Slice(@NonNull INDArray input,
@NonNull INDArray begin,
@NonNull INDArray end) |
Slice(@NonNull INDArray input,
@NonNull INDArray begin,
@NonNull INDArray end) |
Slice(INDArray input,
int[] begin,
int... size) |
Split(@NonNull INDArray in,
INDArray out) |
Split(@NonNull INDArray in,
INDArray out) |
Split(INDArray input,
INDArray numSplit,
int splitDim) |
Split(INDArray input,
int numSplit,
int splitDim) |
SplitV(INDArray input,
INDArray sizes,
int numSplit,
int splitDim) |
Squeeze(INDArray x,
int axis) |
Stack(INDArray[] inputs,
INDArray output,
int axis) |
Stack(INDArray[] inputs,
INDArray output,
int axis) |
Stack(INDArray[] input,
int axis) |
StridedSlice(INDArray in,
INDArray begin,
INDArray end,
INDArray strides) |
StridedSlice(INDArray in,
INDArray begin,
INDArray end,
INDArray strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask) |
StridedSlice(INDArray in,
int[] begin,
int[] end,
int[] strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask) |
StridedSlice(INDArray in,
long[] begin,
long[] end,
long[] strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask) |
Tile(INDArray[] inputs,
INDArray[] outputs,
int[] axis) |
Tile(INDArray[] inputs,
INDArray[] outputs,
int[] axis) |
Tile(INDArray[] inputs,
INDArray[] outputs,
int[] axis,
boolean is_static_reps) |
Tile(INDArray[] inputs,
INDArray[] outputs,
int[] axis,
boolean is_static_reps) |
Tile(INDArray x,
INDArray repeat) |
Tile(INDArray inputs,
int... axis) |
Transpose(INDArray input) |
Transpose(INDArray input,
INDArray result) |
Unstack(INDArray in,
INDArray[] out,
int axis) |
Unstack(INDArray in,
INDArray[] out,
int axis) |
Unstack(@NonNull INDArray value,
int axis,
int num) |
ZerosLike(INDArray in) |
ZerosLike(INDArray in,
INDArray out) |
ZerosLike(INDArray in,
INDArray out,
DataType dataType) |
| Constructor and Description |
|---|
EmbeddingLookup(@NonNull INDArray in,
@NonNull INDArray indices,
PartitionMode partitionMode,
INDArray output) |
EmbeddingLookup(@NonNull INDArray in,
@NonNull INDArray indices,
PartitionMode partitionMode,
INDArray output) |
EmbeddingLookup(@NonNull INDArray in,
@NonNull INDArray indices,
PartitionMode partitionMode,
INDArray output) |
EmbeddingLookup(@NonNull INDArray in,
INDArray output,
PartitionMode partitionMode,
int... indices) |
EmbeddingLookup(@NonNull INDArray in,
INDArray output,
PartitionMode partitionMode,
int... indices) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
Variance.noOp() |
| Constructor and Description |
|---|
StandardDeviation(INDArray x) |
StandardDeviation(INDArray x,
boolean biasCorrected,
boolean keepDims,
int... dimension) |
StandardDeviation(INDArray x,
boolean keepDims,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
StandardDeviation(INDArray x,
boolean keepDims,
double mean,
double bias,
int... dimensions) |
StandardDeviation(INDArray x,
boolean keepDims,
double mean,
int... dimensions) |
StandardDeviation(INDArray x,
boolean biasCorrected,
int... dimension) |
StandardDeviation(INDArray x,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
StandardDeviation(INDArray x,
double mean,
double bias,
int... dimensions) |
StandardDeviation(INDArray x,
double mean,
int... dimensions) |
StandardDeviation(INDArray x,
INDArray z,
boolean newFormat,
boolean keepDims,
int[] dimensions) |
StandardDeviation(INDArray x,
INDArray z,
boolean biasCorrected,
int... dimension) |
StandardDeviation(INDArray x,
INDArray y,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
StandardDeviation(INDArray x,
INDArray y,
double mean,
double bias,
int... dimensions) |
StandardDeviation(INDArray x,
INDArray y,
double mean,
int... dimensions) |
StandardDeviation(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double mean) |
StandardDeviation(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double mean,
double bias) |
StandardDeviation(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double mean,
double bias,
boolean biasCorrected) |
StandardDeviation(INDArray x,
INDArray y,
INDArray z,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
StandardDeviation(INDArray x,
INDArray y,
INDArray z,
double mean,
double bias,
int... dimensions) |
StandardDeviation(INDArray x,
INDArray y,
INDArray z,
double mean,
int... dimensions) |
StandardDeviation(INDArray x,
int... dimension) |
Variance(INDArray x,
boolean biasCorrected,
boolean keepDims,
int... dimensions) |
Variance(INDArray x,
boolean keepDims,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
Variance(INDArray x,
boolean keepDims,
double mean,
double bias,
int... dimensions) |
Variance(INDArray x,
boolean keepDims,
double mean,
int... dimensions) |
Variance(INDArray x,
boolean biasCorrected,
int... dimensions) |
Variance(INDArray x,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
Variance(INDArray x,
double mean,
double bias,
int... dimensions) |
Variance(INDArray x,
double mean,
int... dimensions) |
Variance(INDArray x,
INDArray z,
boolean biasCorrected,
boolean keepDims,
int... dimensions) |
Variance(INDArray x,
INDArray z,
boolean biasCorrected,
int... dimensions) |
Variance(INDArray x,
INDArray y,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
Variance(INDArray x,
INDArray y,
double mean,
double bias,
int... dimensions) |
Variance(INDArray x,
INDArray y,
double mean,
int... dimensions) |
Variance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double mean) |
Variance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double mean,
double bias) |
Variance(INDArray x,
INDArray y,
INDArray z,
boolean keepDims,
int[] dimensions,
double mean,
double bias,
boolean biasCorrected) |
Variance(INDArray x,
INDArray y,
INDArray z,
double mean,
double bias,
boolean biasCorrected,
int... dimensions) |
Variance(INDArray x,
INDArray y,
INDArray z,
double mean,
double bias,
int... dimensions) |
Variance(INDArray x,
INDArray y,
INDArray z,
double mean,
int... dimensions) |
Variance(INDArray x,
int... dimension) |
| Constructor and Description |
|---|
BaseDynamicTransformOp(INDArray[] inputs,
INDArray[] outputs) |
BaseDynamicTransformOp(INDArray[] inputs,
INDArray[] outputs) |
Cholesky(INDArray input) |
Histogram(INDArray input,
INDArray output) |
Histogram(INDArray input,
long numBins) |
MaxOut(INDArray x) |
MaxOut(INDArray x,
INDArray z) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
INDArray out,
@NonNull Pad.Mode mode,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
INDArray out,
@NonNull Pad.Mode mode,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
INDArray out,
@NonNull Pad.Mode mode,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
INDArray out,
@NonNull PadMode mode,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
INDArray out,
@NonNull PadMode mode,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
INDArray out,
@NonNull PadMode mode,
double padValue) |
Pad(INDArray input,
INDArray padding,
Mode mode,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
@NonNull PadMode mode,
double padValue) |
Pad(@NonNull INDArray in,
@NonNull INDArray padding,
@NonNull PadMode mode,
double padValue) |
ReluLayer(@NonNull INDArray input,
@NonNull INDArray weights,
@NonNull INDArray bias) |
ReluLayer(@NonNull INDArray input,
@NonNull INDArray weights,
@NonNull INDArray bias) |
ReluLayer(@NonNull INDArray input,
@NonNull INDArray weights,
@NonNull INDArray bias) |
| Constructor and Description |
|---|
Assign(INDArray x) |
Assign(INDArray x,
INDArray z) |
IsMax(INDArray x) |
IsMax(INDArray x,
INDArray z) |
IsMax(INDArray x,
INDArray z,
int... dimensions) |
IsMax(INDArray x,
int... dimensions) |
| Constructor and Description |
|---|
ClipByAvgNorm(INDArray in,
double clipValue,
int... dimensions) |
ClipByAvgNorm(INDArray in,
INDArray out,
double clipValue,
int... dimensions) |
ClipByNorm(INDArray in,
double clipValue,
int... dimensions) |
ClipByNorm(INDArray in,
INDArray out,
double clipValue,
int... dimensions) |
ClipByNorm(INDArray x,
INDArray clipValue,
INDArray dimensions) |
ClipByValue(@NonNull INDArray input,
double clipValueMin,
double clipValueMax) |
ClipByValue(INDArray x,
INDArray clipValueMin,
INDArray clipValueMax) |
| Constructor and Description |
|---|
CompareAndReplace(INDArray x,
INDArray y,
Condition condition)
With this constructor, op will check each X element against given Condition, and if condition met, element Z will be set to Y value, and X otherwise
PLEASE NOTE: X will be modified inplace.
|
CompareAndReplace(INDArray x,
INDArray y,
INDArray z,
Condition condition)
With this constructor, op will check each X element against given Condition, and if condition met, element Z will be set to Y value, and X otherwise
Pseudocode:
z[i] = condition(x[i]) ? y[i] : x[i];
|
CompareAndSet(INDArray x,
double set,
Condition condition)
With this constructor, op will check each X element against given Condition, and if condition met, element will be replaced with Set value
Pseudocode:
z[i] = condition(x[i]) ? set : x[i];
PLEASE NOTE: X will be modified inplace.
|
CompareAndSet(INDArray x,
double compare,
double set,
double eps) |
CompareAndSet(INDArray x,
double compare,
double set,
double eps,
Condition condition) |
CompareAndSet(INDArray x,
INDArray y,
Condition condition)
With this constructor, op will check each Y element against given Condition, and if condition met, element Z will be set to Y value, and X otherwise
PLEASE NOTE: X will be modified inplace.
|
CompareAndSet(INDArray x,
INDArray z,
double set,
Condition condition)
With this constructor, op will check each X element against given Condition, and if condition met, element will be replaced with Set value
Pseudocode:
z[i] = condition(x[i]) ? set : x[i];
|
CompareAndSet(INDArray x,
INDArray z,
double compare,
double set,
double eps)
This constructor is shortcut to epsEquals.
|
CompareAndSet(INDArray x,
INDArray y,
INDArray z,
Condition condition)
With this constructor, op will check each Y element against given Condition, and if condition met, element Z will be set to Y value, and X otherwise
Pseudocode:
z[i] = condition(y[i]) ? y[i] : x[i];
|
Eps(INDArray indArray) |
Eps(INDArray x,
INDArray y,
INDArray z) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
InvertPermutation.generateFake(DataType dataType,
long... shape) |
INDArray |
InvertPermutation.generateFake(long... shape) |
| Constructor and Description |
|---|
Assign(INDArray[] inputs,
INDArray[] outputs) |
Assign(INDArray[] inputs,
INDArray[] outputs) |
Assign(INDArray x,
INDArray y) |
ATan2(INDArray x,
INDArray y)
Note that the order of x and y match
Math.atan2(double, double),
and are reversed when compared to OldATan2. |
ATan2(INDArray x,
INDArray y,
INDArray z)
Note that the order of x and y match
Math.atan2(double, double),
and are reversed when compared to OldATan2. |
BatchToSpace(INDArray x,
int[] blocks,
int[] croppingTop,
int... croppingBottom) |
BitsHammingDistance(@NonNull INDArray x,
@NonNull INDArray y) |
BitsHammingDistance(@NonNull INDArray x,
@NonNull INDArray y) |
BitwiseAnd(INDArray x,
INDArray y) |
BitwiseAnd(INDArray x,
INDArray y,
INDArray output) |
BitwiseOr(INDArray x,
INDArray y) |
BitwiseOr(INDArray x,
INDArray y,
INDArray output) |
BitwiseXor(INDArray x,
INDArray y) |
BitwiseXor(INDArray x,
INDArray y,
INDArray output) |
Choose(INDArray[] inputs,
Condition condition) |
Choose(INDArray[] inputs,
List<Integer> iArgs,
List<Double> tArgs,
Condition condition)
Note that iArgs (integer arguments) and tArgs(double/float arguments)
may end up being used under the following conditions:
scalar operations (if a scalar is specified the you do not need to specify an ndarray)
otherwise, if an ndarray is needed as a second input then put it in the inputs
Usually, you only need 1 input (the equivalent of the array you're trying to do indexing on)
|
Choose(String opName,
INDArray[] inputs,
Condition condition) |
Choose(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
Choose(String opName,
INDArray[] inputs,
INDArray[] outputs,
List<Double> tArguments,
List<Integer> iArguments) |
CReLU(@NonNull INDArray input) |
CReluBp(@NonNull INDArray input,
@NonNull INDArray epsilonNext,
INDArray output) |
CReluBp(@NonNull INDArray input,
@NonNull INDArray epsilonNext,
INDArray output) |
CReluBp(@NonNull INDArray input,
@NonNull INDArray epsilonNext,
INDArray output) |
CumProd(INDArray in,
boolean exclusive,
boolean reverse,
int... axis) |
CumProd(INDArray in,
INDArray result,
boolean exclusive,
boolean reverse,
int... axis) |
CumSum(INDArray in,
boolean exclusive,
boolean reverse,
int... axis) |
CumSum(INDArray in,
INDArray result,
boolean exclusive,
boolean reverse,
int... axis) |
CyclicRShiftBits(INDArray input,
INDArray shift) |
CyclicRShiftBits(INDArray input,
INDArray shift,
INDArray output) |
CyclicShiftBits(INDArray input,
INDArray shift) |
CyclicShiftBits(INDArray input,
INDArray shift,
INDArray output) |
Dilation2D(INDArray[] inputArrays,
INDArray[] outputs) |
Dilation2D(INDArray[] inputArrays,
INDArray[] outputs) |
Dilation2D(INDArray df,
INDArray weights,
int[] strides,
int[] rates,
boolean isSameMode) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled,
boolean withWeights) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled,
boolean withWeights) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled,
boolean withWeights) |
DotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
INDArray mask,
boolean scaled,
boolean withWeights) |
DynamicPartition(INDArray x,
INDArray[] partitions,
int numPartitions) |
DynamicPartition(INDArray x,
INDArray[] partitions,
int numPartitions) |
DynamicPartition(@NonNull INDArray input,
@NonNull INDArray partitions,
int numPartitions) |
DynamicPartition(@NonNull INDArray input,
@NonNull INDArray partitions,
int numPartitions) |
DynamicStitch(@NonNull INDArray[] indices,
@NonNull INDArray[] inputs) |
DynamicStitch(@NonNull INDArray[] indices,
@NonNull INDArray[] inputs) |
EqualTo(INDArray[] inputs,
INDArray[] outputs) |
EqualTo(INDArray[] inputs,
INDArray[] outputs) |
EqualTo(INDArray x,
INDArray y) |
EqualTo(INDArray x,
INDArray y,
INDArray z) |
FakeQuantWithMinMaxArgs(INDArray x,
INDArray min,
INDArray max,
int num_bits,
boolean narrow) |
FakeQuantWithMinMaxVars(INDArray x,
INDArray min,
INDArray max,
int num_bits,
boolean narrow) |
Fill(INDArray shape,
DataType dtype,
double value) |
Fill(INDArray shape,
INDArray result,
double value) |
Fill(INDArray shape,
INDArray value,
INDArray result) |
GreaterThan(INDArray[] inputs,
INDArray[] outputs) |
GreaterThan(INDArray[] inputs,
INDArray[] outputs) |
GreaterThan(INDArray x,
INDArray y) |
GreaterThan(INDArray x,
INDArray y,
INDArray z) |
GreaterThanOrEqual(INDArray[] inputs,
INDArray[] outputs) |
GreaterThanOrEqual(INDArray[] inputs,
INDArray[] outputs) |
GreaterThanOrEqual(INDArray x,
INDArray y) |
GreaterThanOrEqual(INDArray x,
INDArray y,
INDArray z) |
InvertPermutation(INDArray input) |
IsNonDecreasing(@NonNull INDArray input) |
IsNonDecreasing(@NonNull INDArray input,
INDArray output) |
IsNonDecreasing(@NonNull INDArray input,
INDArray output) |
IsNumericTensor(INDArray inputs) |
IsNumericTensor(INDArray[] inputs,
INDArray[] outputs) |
IsNumericTensor(INDArray[] inputs,
INDArray[] outputs) |
IsStrictlyIncreasing(@NonNull INDArray input) |
IsStrictlyIncreasing(@NonNull INDArray input,
INDArray output) |
IsStrictlyIncreasing(@NonNull INDArray input,
INDArray output) |
LayerNorm(@NonNull INDArray input,
@NonNull INDArray gain,
boolean channelsFirst,
int... dimensions) |
LayerNorm(@NonNull INDArray input,
@NonNull INDArray gain,
boolean channelsFirst,
int... dimensions) |
LayerNorm(INDArray input,
INDArray gain,
INDArray result,
boolean channelsFirst,
int... dimensions) |
LayerNorm(INDArray input,
INDArray gain,
INDArray bias,
INDArray result,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(INDArray input,
INDArray gain,
INDArray grad,
INDArray dLdx,
INDArray dLdg,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(@NonNull INDArray input,
@NonNull INDArray gain,
INDArray bias,
@NonNull INDArray grad,
@NonNull INDArray dLdx,
@NonNull INDArray dLdg,
INDArray dLdb,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(@NonNull INDArray input,
@NonNull INDArray gain,
INDArray bias,
@NonNull INDArray grad,
@NonNull INDArray dLdx,
@NonNull INDArray dLdg,
INDArray dLdb,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(@NonNull INDArray input,
@NonNull INDArray gain,
INDArray bias,
@NonNull INDArray grad,
@NonNull INDArray dLdx,
@NonNull INDArray dLdg,
INDArray dLdb,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(@NonNull INDArray input,
@NonNull INDArray gain,
INDArray bias,
@NonNull INDArray grad,
@NonNull INDArray dLdx,
@NonNull INDArray dLdg,
INDArray dLdb,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(@NonNull INDArray input,
@NonNull INDArray gain,
INDArray bias,
@NonNull INDArray grad,
@NonNull INDArray dLdx,
@NonNull INDArray dLdg,
INDArray dLdb,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(@NonNull INDArray input,
@NonNull INDArray gain,
INDArray bias,
@NonNull INDArray grad,
@NonNull INDArray dLdx,
@NonNull INDArray dLdg,
INDArray dLdb,
boolean channelsFirst,
int... dimensions) |
LessThan(INDArray[] inputs,
INDArray[] outputs) |
LessThan(INDArray[] inputs,
INDArray[] outputs) |
LessThan(INDArray x,
INDArray y) |
LessThan(INDArray x,
INDArray y,
INDArray z) |
LessThanOrEqual(INDArray[] inputs,
INDArray[] outputs) |
LessThanOrEqual(INDArray[] inputs,
INDArray[] outputs) |
LessThanOrEqual(INDArray x,
INDArray y) |
LessThanOrEqual(INDArray x,
INDArray y,
INDArray z) |
ListDiff(@NonNull INDArray x,
@NonNull INDArray y) |
ListDiff(@NonNull INDArray x,
@NonNull INDArray y) |
LogSoftMax(INDArray x) |
LogSoftMax(INDArray x,
INDArray z) |
LogSoftMax(INDArray x,
int dimension) |
MatrixDeterminant(@NonNull INDArray input) |
MatrixInverse(@NonNull INDArray input) |
MatrixSetDiag(@NonNull INDArray in,
@NonNull INDArray diag) |
MatrixSetDiag(@NonNull INDArray in,
@NonNull INDArray diag) |
Max(INDArray[] inputs,
INDArray[] outputs) |
Max(INDArray[] inputs,
INDArray[] outputs) |
Max(INDArray first,
INDArray second) |
Max(INDArray first,
INDArray second,
INDArray out) |
Min(INDArray[] inputs,
INDArray[] outputs) |
Min(INDArray[] inputs,
INDArray[] outputs) |
Min(INDArray first,
INDArray second) |
Min(INDArray first,
INDArray second,
INDArray out) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttention(@NonNull INDArray queries,
@NonNull INDArray keys,
@NonNull INDArray values,
@NonNull INDArray Wq,
@NonNull INDArray Wk,
@NonNull INDArray Wv,
@NonNull INDArray Wo,
INDArray mask,
boolean scaled,
boolean withWeights) |
NotEqualTo(INDArray[] inputs,
INDArray[] outputs) |
NotEqualTo(INDArray[] inputs,
INDArray[] outputs) |
NotEqualTo(INDArray x,
INDArray y) |
NotEqualTo(INDArray x,
INDArray y,
INDArray z) |
Pow(@NonNull INDArray x,
@NonNull INDArray y) |
Pow(@NonNull INDArray x,
@NonNull INDArray y) |
Qr(INDArray input) |
Qr(INDArray input,
boolean fullMatrices) |
Reverse(INDArray x)
Inplace reverse.
|
Reverse(INDArray x,
INDArray z)
Reverses whole array for compatibility with OldReverse.
|
Reverse(INDArray x,
INDArray z,
int... axis)
This constructor allows to specify axis for Reverse operation
|
Reverse(INDArray x,
int... axis)
This constructor allows to specify axis for Reverse operation
|
ReverseSequence(INDArray x,
INDArray seq_lengths) |
ReverseSequence(INDArray x,
INDArray seq_lengths,
int seqDim,
int batchDim) |
RShiftBits(INDArray input,
INDArray shift) |
RShiftBits(INDArray input,
INDArray shift,
INDArray output) |
ShiftBits(INDArray x,
INDArray y) |
ShiftBits(INDArray x,
INDArray y,
INDArray output) |
SoftMax(INDArray input) |
SoftMax(INDArray input,
INDArray result) |
SoftMax(INDArray input,
INDArray result,
int dimension) |
SoftMax(@NonNull INDArray input,
int dimension) |
SpaceToBatch(INDArray x,
int[] blocks,
int[] paddingTop,
int... paddingBottom) |
Standardize(INDArray input,
INDArray result,
int... dimensions) |
Standardize(INDArray input,
int... dimensions) |
StandardizeBp(INDArray input,
INDArray eps,
INDArray result,
int... dimensions) |
Svd(INDArray input,
boolean fullUV,
boolean computeUV,
int switchNum) |
Svd(INDArray input,
boolean full_matrices,
INDArray s,
INDArray u,
INDArray v) |
ThresholdRelu(@NonNull INDArray input,
INDArray output,
double cutoff) |
ThresholdRelu(@NonNull INDArray input,
INDArray output,
double cutoff) |
TopK(INDArray input,
double k,
boolean sorted) |
Trace(@NonNull INDArray in) |
XwPlusB(INDArray[] inputs,
INDArray output) |
XwPlusB(INDArray[] inputs,
INDArray output) |
XwPlusB(INDArray input,
INDArray weights,
INDArray bias) |
| Constructor and Description |
|---|
SegmentMax(INDArray data,
INDArray segmentIds) |
SegmentMean(INDArray data,
INDArray segmentIds) |
SegmentMin(INDArray data,
INDArray segmentIds) |
SegmentProd(INDArray data,
INDArray segmentIds) |
SegmentSum(INDArray data,
INDArray segmentIds) |
| Constructor and Description |
|---|
Cast(@NonNull INDArray arg,
@NonNull DataType dataType) |
| Constructor and Description |
|---|
RSqrt(INDArray x) |
RSqrt(INDArray x,
INDArray z) |
Sqrt(INDArray x) |
Sqrt(INDArray x,
INDArray z) |
| Constructor and Description |
|---|
CubeBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
CubeBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
CubeBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
CubeDerivative(INDArray x)
Deprecated.
|
CubeDerivative(INDArray x,
INDArray z)
Deprecated.
|
EluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
EluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
EluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
EluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double alpha) |
EluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double alpha) |
EluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double alpha) |
HardSigmoidBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
HardSigmoidBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
HardSigmoidBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
HardSigmoidDerivative(INDArray x)
Deprecated.
|
HardSigmoidDerivative(INDArray x,
INDArray z)
Deprecated.
|
HardTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
HardTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
HardTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
HardTanhDerivative(INDArray x)
Deprecated.
|
HardTanhDerivative(INDArray x,
INDArray z)
Deprecated.
|
LeakyReLUBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double alpha) |
LeakyReLUBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double alpha) |
LeakyReLUBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double alpha) |
LeakyReLUDerivative(INDArray x) |
LeakyReLUDerivative(INDArray x,
double alpha) |
LeakyReLUDerivative(INDArray x,
INDArray z) |
LeakyReLUDerivative(INDArray x,
INDArray z,
double alpha) |
LogSoftMaxDerivative(INDArray in,
INDArray gradO,
INDArray out) |
PReluBp(@NonNull INDArray input,
@NonNull INDArray alpha,
@NonNull INDArray gradient,
INDArray dLdI,
INDArray dLdA,
int... sharedAxes) |
PReluBp(@NonNull INDArray input,
@NonNull INDArray alpha,
@NonNull INDArray gradient,
INDArray dLdI,
INDArray dLdA,
int... sharedAxes) |
PReluBp(@NonNull INDArray input,
@NonNull INDArray alpha,
@NonNull INDArray gradient,
INDArray dLdI,
INDArray dLdA,
int... sharedAxes) |
PReluBp(@NonNull INDArray input,
@NonNull INDArray alpha,
@NonNull INDArray gradient,
INDArray dLdI,
INDArray dLdA,
int... sharedAxes) |
RationalTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
RationalTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
RationalTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
RationalTanhDerivative(INDArray x)
Deprecated.
|
RationalTanhDerivative(INDArray x,
INDArray z)
Deprecated.
|
RectifiedTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
RectifiedTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
RectifiedTanhBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
RectifiedTanhDerivative(INDArray x)
Deprecated.
|
RectifiedTanhDerivative(INDArray x,
INDArray z)
Deprecated.
|
Relu6Derivative(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
Relu6Derivative(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
Relu6Derivative(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
Relu6Derivative(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double cutoff) |
Relu6Derivative(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double cutoff) |
Relu6Derivative(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double cutoff) |
SeluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SeluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SeluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SELUDerivative(INDArray x)
Deprecated.
|
SELUDerivative(INDArray x,
INDArray z)
Deprecated.
|
SigmoidDerivative(@NonNull INDArray x,
@NonNull INDArray y) |
SigmoidDerivative(@NonNull INDArray x,
@NonNull INDArray y) |
SigmoidDerivative(INDArray x,
INDArray y,
INDArray z) |
SoftmaxBp(@NonNull INDArray input,
@NonNull INDArray grad,
INDArray output,
Integer dimension) |
SoftmaxBp(@NonNull INDArray input,
@NonNull INDArray grad,
INDArray output,
Integer dimension) |
SoftmaxBp(@NonNull INDArray input,
@NonNull INDArray grad,
INDArray output,
Integer dimension) |
SoftmaxBp(@NonNull INDArray input,
@NonNull INDArray grad,
Integer dimension) |
SoftmaxBp(@NonNull INDArray input,
@NonNull INDArray grad,
Integer dimension) |
SoftPlusBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SoftPlusBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SoftPlusBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SoftSignBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SoftSignBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SoftSignBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output) |
SoftSignDerivative(INDArray x)
Deprecated.
|
SoftSignDerivative(INDArray x,
INDArray z)
Deprecated.
|
TanhDerivative(INDArray x,
INDArray y) |
TanhDerivative(INDArray x,
INDArray y,
INDArray z) |
ThresholdReluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double cutoff) |
ThresholdReluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double cutoff) |
ThresholdReluBp(@NonNull INDArray input,
@NonNull INDArray gradient,
INDArray output,
double cutoff) |
| Constructor and Description |
|---|
BinaryMinimalRelativeError(INDArray x,
INDArray y,
double thresholdRelative,
double thresholdAbsolute) |
BinaryMinimalRelativeError(INDArray x,
INDArray y,
INDArray z,
double thresholdRelative,
double thresholdAbsolute) |
BinaryRelativeError(INDArray x,
INDArray y,
double threshold) |
BinaryRelativeError(INDArray x,
INDArray y,
INDArray z,
double threshold) |
RelativeError(INDArray x,
INDArray y) |
RelativeError(INDArray x,
INDArray y,
INDArray z) |
Set(INDArray x) |
Set(INDArray x,
INDArray z) |
Set(INDArray x,
INDArray y,
INDArray z) |
| Constructor and Description |
|---|
And(@NonNull INDArray x,
@NonNull INDArray y) |
And(@NonNull INDArray x,
@NonNull INDArray y) |
And(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
And(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
And(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
And(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
And(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
And(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
And(@NonNull INDArray x,
@NonNull INDArray y,
Number comparable) |
And(@NonNull INDArray x,
@NonNull INDArray y,
Number comparable) |
Not(@NonNull INDArray x,
INDArray y,
INDArray z,
Number comparable) |
Not(@NonNull INDArray x,
INDArray y,
INDArray z,
Number comparable) |
Or(@NonNull INDArray x,
@NonNull INDArray y) |
Or(@NonNull INDArray x,
@NonNull INDArray y) |
Or(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
Or(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
Or(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
Or(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
Or(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
Or(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
Xor(@NonNull INDArray x,
@NonNull INDArray y) |
Xor(@NonNull INDArray x,
@NonNull INDArray y) |
Xor(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
Xor(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
Xor(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z) |
Xor(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
Xor(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
Xor(@NonNull INDArray x,
@NonNull INDArray y,
INDArray z,
Number comparable) |
| Constructor and Description |
|---|
Abs(INDArray x) |
Abs(INDArray x,
INDArray z) |
AMax(INDArray x,
INDArray y,
INDArray z) |
AMin(INDArray x,
INDArray y,
INDArray z) |
Ceil(INDArray x) |
Ceil(INDArray x,
INDArray z) |
Cube(INDArray x) |
Cube(INDArray x,
INDArray z) |
Floor(INDArray x) |
Floor(INDArray x,
INDArray z) |
Identity(INDArray x) |
Identity(INDArray x,
INDArray z) |
Max(INDArray x,
INDArray y,
INDArray z) |
Min(INDArray x,
INDArray y,
INDArray z) |
Negative(INDArray x) |
Negative(INDArray x,
INDArray z) |
OneMinus(INDArray x) |
OneMinus(INDArray x,
INDArray z) |
Reciprocal(INDArray x) |
Reciprocal(INDArray x,
INDArray z) |
Round(INDArray x) |
Round(INDArray x,
INDArray z) |
Sign(INDArray x) |
Sign(INDArray x,
INDArray z) |
Square(INDArray x) |
Square(INDArray x,
INDArray z) |
TimesOneMinus(INDArray x) |
TimesOneMinus(INDArray x,
INDArray z) |
| Constructor and Description |
|---|
UnsortedSegmentMax(INDArray data,
INDArray segmentIds,
int numSegments) |
UnsortedSegmentMean(INDArray data,
INDArray segmentIds,
int numSegments) |
UnsortedSegmentMin(INDArray data,
INDArray segmentIds,
int numSegments) |
UnsortedSegmentProd(INDArray data,
INDArray segmentIds,
int numSegments) |
UnsortedSegmentSqrtN(INDArray data,
INDArray segmentIds,
int numSegments) |
UnsortedSegmentSum(INDArray data,
INDArray segmentIds,
int numSegments) |
| Constructor and Description |
|---|
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaBeliefUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
@NonNull INDArray updates,
@NonNull INDArray updatedStateMsg,
@NonNull INDArray updatedStateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
@NonNull INDArray updates,
@NonNull INDArray updatedStateMsg,
@NonNull INDArray updatedStateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
@NonNull INDArray updates,
@NonNull INDArray updatedStateMsg,
@NonNull INDArray updatedStateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
@NonNull INDArray updates,
@NonNull INDArray updatedStateMsg,
@NonNull INDArray updatedStateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
@NonNull INDArray updates,
@NonNull INDArray updatedStateMsg,
@NonNull INDArray updatedStateMsdx,
double rho,
double epsilon) |
AdaDeltaUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateMsg,
@NonNull INDArray stateMsdx,
@NonNull INDArray updates,
@NonNull INDArray updatedStateMsg,
@NonNull INDArray updatedStateMsdx,
double rho,
double epsilon) |
AdaGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
double lr,
double epsilon) |
AdaGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
double lr,
double epsilon) |
AdaGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double epsilon) |
AdaGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double epsilon) |
AdaGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double epsilon) |
AdaGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double epsilon) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdaMaxUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AdamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateU,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateU,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
AmsGradUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray stateH,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
@NonNull INDArray updatedStateH,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NadamUpdater(@NonNull INDArray gradients,
@NonNull INDArray stateV,
@NonNull INDArray stateM,
@NonNull INDArray updates,
@NonNull INDArray updatedStateV,
@NonNull INDArray updatedStateM,
double lr,
double beta1,
double beta2,
double epsilon,
int iteration) |
NesterovsUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
double lr,
double momentum) |
NesterovsUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
double lr,
double momentum) |
NesterovsUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double momentum) |
NesterovsUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double momentum) |
NesterovsUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double momentum) |
NesterovsUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double momentum) |
RmsPropUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
double lr,
double decay,
double epsilon) |
RmsPropUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
double lr,
double decay,
double epsilon) |
RmsPropUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double decay,
double epsilon) |
RmsPropUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double decay,
double epsilon) |
RmsPropUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double decay,
double epsilon) |
RmsPropUpdater(@NonNull INDArray gradients,
@NonNull INDArray state,
@NonNull INDArray updates,
@NonNull INDArray updatedState,
double lr,
double decay,
double epsilon) |
SgdUpdater(@NonNull INDArray input,
double lr) |
SgdUpdater(@NonNull INDArray input,
@NonNull INDArray output,
double lr) |
SgdUpdater(@NonNull INDArray input,
@NonNull INDArray output,
double lr) |
| Constructor and Description |
|---|
BaseRandomOp(INDArray x,
INDArray y,
INDArray z) |
| Constructor and Description |
|---|
RandomStandardNormal(INDArray shape) |
RandomStandardNormal(INDArray shape,
INDArray output) |
| Constructor and Description |
|---|
DistributionUniform(INDArray shape,
INDArray out,
double min,
double max) |
DistributionUniform(INDArray shape,
INDArray out,
double min,
double max,
DataType dataType) |
RandomBernoulli(INDArray shape,
INDArray out,
double p) |
RandomExponential(INDArray shape,
INDArray out,
double lambda) |
RandomGamma(@NonNull INDArray shape,
@NonNull INDArray alpha,
INDArray beta) |
RandomGamma(@NonNull INDArray shape,
@NonNull INDArray alpha,
INDArray beta) |
RandomGamma(@NonNull INDArray shape,
@NonNull INDArray alpha,
INDArray beta) |
RandomGamma(@NonNull INDArray shape,
@NonNull INDArray alpha,
INDArray beta,
int... seeds) |
RandomGamma(@NonNull INDArray shape,
@NonNull INDArray alpha,
INDArray beta,
int... seeds) |
RandomGamma(@NonNull INDArray shape,
@NonNull INDArray alpha,
INDArray beta,
int... seeds) |
RandomPoisson(@NonNull INDArray shape,
@NonNull INDArray rate) |
RandomPoisson(@NonNull INDArray shape,
@NonNull INDArray rate) |
RandomPoisson(@NonNull INDArray shape,
@NonNull INDArray rate,
int... seeds) |
RandomPoisson(@NonNull INDArray shape,
@NonNull INDArray rate,
int... seeds) |
RandomShuffle(@NonNull INDArray value) |
RandomShuffle(@NonNull INDArray value,
int... seeds) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
Linspace.x() |
INDArray |
Linspace.y() |
| Modifier and Type | Method and Description |
|---|---|
void |
Linspace.setX(INDArray x) |
void |
Linspace.setY(INDArray y) |
void |
TruncatedNormalDistribution.setZ(INDArray z) |
void |
BinomialDistribution.setZ(INDArray z) |
void |
LogNormalDistribution.setZ(INDArray z) |
| Constructor and Description |
|---|
AlphaDropOut(@NonNull INDArray x,
double p,
double alpha,
double alphaPrime,
double beta) |
AlphaDropOut(@NonNull INDArray x,
@NonNull INDArray z,
double p,
double alpha,
double alphaPrime,
double beta) |
AlphaDropOut(@NonNull INDArray x,
@NonNull INDArray z,
double p,
double alpha,
double alphaPrime,
double beta) |
BernoulliDistribution(@NonNull INDArray z,
double prob)
This op fills Z with bernoulli trial results, so 0, or 1, depending by common probability
|
BernoulliDistribution(@NonNull INDArray z,
@NonNull INDArray prob)
This op fills Z with bernoulli trial results, so 0, or 1, each element will have it's own success probability defined in prob array
|
BernoulliDistribution(@NonNull INDArray z,
@NonNull INDArray prob)
This op fills Z with bernoulli trial results, so 0, or 1, each element will have it's own success probability defined in prob array
|
BinomialDistribution(@NonNull INDArray z,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
BinomialDistribution(@NonNull INDArray z,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
BinomialDistribution(@NonNull INDArray z,
int trials,
double probability)
This op fills Z with binomial distribution over given trials with single given probability for all trials
|
BinomialDistribution(@NonNull INDArray z,
int trials,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
BinomialDistribution(@NonNull INDArray z,
int trials,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
BinomialDistributionEx(@NonNull INDArray z,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
BinomialDistributionEx(@NonNull INDArray z,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
BinomialDistributionEx(@NonNull INDArray z,
long trials,
double probability)
This op fills Z with binomial distribution over given trials with single given probability for all trials
|
BinomialDistributionEx(@NonNull INDArray z,
long trials,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
BinomialDistributionEx(@NonNull INDArray z,
long trials,
@NonNull INDArray probabilities)
This op fills Z with binomial distribution over given trials with probability for each trial given as probabilities INDArray
|
Choice(@NonNull INDArray source,
@NonNull INDArray probabilities,
@NonNull INDArray z) |
Choice(@NonNull INDArray source,
@NonNull INDArray probabilities,
@NonNull INDArray z) |
Choice(@NonNull INDArray source,
@NonNull INDArray probabilities,
@NonNull INDArray z) |
DropOut(@NonNull INDArray x,
double p) |
DropOut(@NonNull INDArray x,
@NonNull INDArray z,
double p) |
DropOut(@NonNull INDArray x,
@NonNull INDArray z,
double p) |
DropOutInverted(@NonNull INDArray x,
double p) |
DropOutInverted(@NonNull INDArray x,
@NonNull INDArray z,
double p) |
DropOutInverted(@NonNull INDArray x,
@NonNull INDArray z,
double p) |
GaussianDistribution(@NonNull INDArray z)
This op fills Z with random values within -1.0..0..1.0
|
GaussianDistribution(@NonNull INDArray z,
double stddev)
This op fills Z with random values within stddev..0..stddev
|
GaussianDistribution(@NonNull INDArray z,
double mean,
double stddev)
This op fills Z with random values within stddev..mean..stddev boundaries
|
GaussianDistribution(@NonNull INDArray z,
@NonNull INDArray means,
double stddev) |
GaussianDistribution(@NonNull INDArray z,
@NonNull INDArray means,
double stddev) |
Linspace(@NonNull INDArray z,
double from,
double to) |
Linspace(@NonNull INDArray z,
double from,
double to,
double step) |
LogNormalDistribution(@NonNull INDArray z)
This op fills Z with random values within -1.0..0..1.0
|
LogNormalDistribution(@NonNull INDArray z,
double stddev)
This op fills Z with random values within stddev..0..stddev
|
LogNormalDistribution(@NonNull INDArray z,
double mean,
double stddev)
This op fills Z with random values within stddev..mean..stddev boundaries
|
LogNormalDistribution(@NonNull INDArray z,
@NonNull INDArray means,
double stddev) |
LogNormalDistribution(@NonNull INDArray z,
@NonNull INDArray means,
double stddev) |
ProbablisticMerge(@NonNull INDArray x,
@NonNull INDArray y,
double probability) |
ProbablisticMerge(@NonNull INDArray x,
@NonNull INDArray y,
double probability) |
ProbablisticMerge(@NonNull INDArray x,
@NonNull INDArray y,
@NonNull INDArray z,
double probability) |
ProbablisticMerge(@NonNull INDArray x,
@NonNull INDArray y,
@NonNull INDArray z,
double probability) |
ProbablisticMerge(@NonNull INDArray x,
@NonNull INDArray y,
@NonNull INDArray z,
double probability) |
Range(INDArray from,
INDArray to,
INDArray step,
DataType dataType) |
TruncatedNormalDistribution(@NonNull INDArray z)
This op fills Z with random values within -1.0..0..1.0
|
TruncatedNormalDistribution(@NonNull INDArray z,
double stddev)
This op fills Z with random values within stddev..0..stddev
|
TruncatedNormalDistribution(@NonNull INDArray z,
double mean,
double stddev)
This op fills Z with random values within stddev..mean..stddev boundaries
|
TruncatedNormalDistribution(@NonNull INDArray z,
@NonNull INDArray means,
double stddev) |
TruncatedNormalDistribution(@NonNull INDArray z,
@NonNull INDArray means,
double stddev) |
UniformDistribution(@NonNull INDArray z)
This op fills Z with random values within 0...1
|
UniformDistribution(@NonNull INDArray z,
double to)
This op fills Z with random values within 0...to
|
UniformDistribution(@NonNull INDArray z,
double from,
double to)
This op fills Z with random values within from...to boundaries
|
| Constructor and Description |
|---|
PrintAffinity(INDArray array) |
PrintVariable(INDArray array) |
PrintVariable(INDArray array,
boolean printSpecial) |
PrintVariable(INDArray array,
INDArray message) |
PrintVariable(INDArray array,
INDArray message,
boolean printSpecial) |
PrintVariable(INDArray array,
String message) |
PrintVariable(INDArray array,
String message,
boolean printSpecial) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
DefaultRandom.nextDouble(char order,
int[] shape) |
INDArray |
Random.nextDouble(char order,
int[] shape)
Generate a uniform number ndarray
of the specified shape and order
|
INDArray |
DefaultRandom.nextDouble(char order,
long[] shape) |
INDArray |
Random.nextDouble(char order,
long[] shape) |
INDArray |
DefaultRandom.nextDouble(int[] shape) |
INDArray |
Random.nextDouble(int[] shape)
Generate a uniform number ndarray
of the specified shape
|
INDArray |
DefaultRandom.nextDouble(long[] shape) |
INDArray |
Random.nextDouble(long[] shape) |
INDArray |
DefaultRandom.nextFloat(char order,
int[] shape) |
INDArray |
Random.nextFloat(char order,
int[] shape)
Generate a uniform number ndarray
of the specified shape
|
INDArray |
DefaultRandom.nextFloat(char order,
long[] shape) |
INDArray |
Random.nextFloat(char order,
long[] shape) |
INDArray |
DefaultRandom.nextFloat(int[] shape) |
INDArray |
Random.nextFloat(int[] shape)
Generate a uniform number ndarray
of the specified shape
|
INDArray |
DefaultRandom.nextFloat(long[] shape) |
INDArray |
Random.nextFloat(long[] shape) |
INDArray |
DefaultRandom.nextGaussian(char order,
int[] shape) |
INDArray |
Random.nextGaussian(char order,
int[] shape)
Generate a gaussian number ndarray
of the specified shape and order
|
INDArray |
DefaultRandom.nextGaussian(char order,
long[] shape) |
INDArray |
Random.nextGaussian(char order,
long[] shape) |
INDArray |
DefaultRandom.nextGaussian(int[] shape) |
INDArray |
Random.nextGaussian(int[] shape)
Generate a gaussian number ndarray
of the specified shape
|
INDArray |
DefaultRandom.nextGaussian(long[] shape) |
INDArray |
Random.nextGaussian(long[] shape) |
INDArray |
DefaultRandom.nextInt(int[] shape) |
INDArray |
Random.nextInt(int[] shape)
Generate a random set of integers
of the specified shape.
|
INDArray |
DefaultRandom.nextInt(int n,
int[] shape) |
INDArray |
Random.nextInt(int n,
int[] shape)
Generate a random set of integers
of the specified shape.
|
INDArray |
DefaultRandom.nextInt(int n,
long[] shape) |
INDArray |
Random.nextInt(int n,
long[] shape) |
INDArray |
DefaultRandom.nextInt(long[] shape) |
INDArray |
Random.nextInt(long[] shape) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
Distribution.sample(INDArray target)
Fill the target array by sampling from the distribution
|
INDArray |
BaseDistribution.sample(INDArray target) |
INDArray |
Distribution.sample(int[] shape)
Sample the given shape
|
INDArray |
BaseDistribution.sample(int[] shape) |
INDArray |
Distribution.sample(long[] shape) |
INDArray |
BaseDistribution.sample(long[] shape) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
Distribution.sample(INDArray target)
Fill the target array by sampling from the distribution
|
INDArray |
BaseDistribution.sample(INDArray target) |
| Modifier and Type | Method and Description |
|---|---|
Distribution |
DistributionFactory.createBinomial(int n,
INDArray p)
Create a distribution
|
Distribution |
DefaultDistributionFactory.createBinomial(int n,
INDArray p) |
Distribution |
DistributionFactory.createNormal(INDArray mean,
double std)
Create a normal distribution
with the given mean and std
|
Distribution |
DefaultDistributionFactory.createNormal(INDArray mean,
double std) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
TruncatedNormalDistribution.sample(INDArray ret) |
INDArray |
ConstantDistribution.sample(INDArray target) |
INDArray |
BinomialDistribution.sample(INDArray ret) |
INDArray |
OrthogonalDistribution.sample(INDArray target) |
INDArray |
NormalDistribution.sample(INDArray ret) |
INDArray |
LogNormalDistribution.sample(INDArray ret) |
INDArray |
UniformDistribution.sample(INDArray ret) |
INDArray |
TruncatedNormalDistribution.sample(int[] shape) |
INDArray |
ConstantDistribution.sample(int[] shape) |
INDArray |
BinomialDistribution.sample(int[] shape) |
INDArray |
OrthogonalDistribution.sample(int[] shape) |
INDArray |
NormalDistribution.sample(int[] shape) |
INDArray |
LogNormalDistribution.sample(int[] shape) |
INDArray |
UniformDistribution.sample(int[] shape) |
INDArray |
OrthogonalDistribution.sample(long[] shape) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
TruncatedNormalDistribution.sample(INDArray ret) |
INDArray |
ConstantDistribution.sample(INDArray target) |
INDArray |
BinomialDistribution.sample(INDArray ret) |
INDArray |
OrthogonalDistribution.sample(INDArray target) |
INDArray |
NormalDistribution.sample(INDArray ret) |
INDArray |
LogNormalDistribution.sample(INDArray ret) |
INDArray |
UniformDistribution.sample(INDArray ret) |
| Constructor and Description |
|---|
BinomialDistribution(int n,
INDArray p) |
LogNormalDistribution(double standardDeviation,
INDArray means) |
LogNormalDistribution(INDArray mean,
double std) |
LogNormalDistribution(Random rng,
double standardDeviation,
INDArray means) |
NormalDistribution(double standardDeviation,
INDArray means) |
NormalDistribution(INDArray mean,
double std) |
NormalDistribution(Random rng,
double standardDeviation,
INDArray means) |
TruncatedNormalDistribution(double standardDeviation,
INDArray means) |
TruncatedNormalDistribution(INDArray mean,
double std) |
TruncatedNormalDistribution(Random rng,
double standardDeviation,
INDArray means) |
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Shape.ndArrayDimFromInt(int... dimensions)
Create an INDArray to represent the (possibly null) int[] dimensions.
|
static INDArray |
Shape.newShapeNoCopy(INDArray arr,
int[] newShape,
boolean isFOrder) |
static INDArray |
Shape.newShapeNoCopy(INDArray arr,
long[] newShape,
boolean isFOrder)
A port of numpy's reshaping algorithm that leverages
no copy where possible and returns
null if the reshape
couldn't happen without copying
|
static INDArray |
Shape.toMmulCompatible(INDArray input)
This method is used in DL4J LSTM implementation
|
static INDArray |
Shape.toOffsetZero(INDArray arr)
Create a copy of the matrix
where the new offset is zero
|
static INDArray |
Shape.toOffsetZeroCopy(INDArray arr)
Create a copy of the ndarray where the new offset is zero
|
static INDArray |
Shape.toOffsetZeroCopy(INDArray arr,
char order)
Create a copy of the ndarray where the new offset is zero, and has specified order
|
static INDArray |
Shape.toOffsetZeroCopyAnyOrder(INDArray arr)
Create a copy of the ndarray where the new offset is zero.
|
| Modifier and Type | Method and Description |
|---|---|
static void |
Shape.assertBroadcastable(@NonNull INDArray x,
@NonNull INDArray y) |
static void |
Shape.assertBroadcastable(@NonNull INDArray x,
@NonNull INDArray y) |
static void |
Shape.assertBroadcastable(String op,
INDArray first,
INDArray second,
INDArray result)
Assert that the broadcast operation
result = first.op(second) is valid, given the
shapes of first, second, and result.Throws an exception otherwise |
static double |
Shape.getDouble(INDArray arr,
int[] indices)
Get a double based on the array and given indices
|
static double |
Shape.getDouble(INDArray arr,
long... indices) |
static long |
Shape.getLong(INDArray arr,
long... indices) |
static long[] |
Shape.getMaxShape(INDArray... inputs)
Return the shape of the largest length array
based on the input
|
static char |
Shape.getOrder(INDArray arr)
Infer the order for the ndarray based on the
array's strides
|
static boolean |
Shape.hasDefaultStridesForShape(INDArray input) |
static long[] |
Shape.ind2sub(INDArray arr,
long index)
Convert a linear index to
the equivalent nd index based on the shape of the specified ndarray.
|
static long[] |
Shape.ind2subC(INDArray arr,
long index)
Convert a linear index to
the equivalent nd index based on the shape of the specified ndarray.
|
static boolean |
Shape.isContiguousInBuffer(INDArray in)
Are the elements in the buffer contiguous for this NDArray?
|
static void |
Shape.iterate(INDArray arr,
CoordinateFunction coordinateFunction)
Iterate over 2
coordinate spaces given 2 arrays
|
static void |
Shape.iterate(INDArray arr,
INDArray arr2,
CoordinateFunction coordinateFunction)
Iterate over 2
coordinate spaces given 2 arrays
|
static INDArray |
Shape.newShapeNoCopy(INDArray arr,
int[] newShape,
boolean isFOrder) |
static INDArray |
Shape.newShapeNoCopy(INDArray arr,
long[] newShape,
boolean isFOrder)
A port of numpy's reshaping algorithm that leverages
no copy where possible and returns
null if the reshape
couldn't happen without copying
|
static long[] |
Shape.reductionShape(INDArray x,
int[] dimension,
boolean newFormat,
boolean keepDims)
Calculate the shape of the returned array, for a reduction along dimension
|
static String |
Shape.shapeToString(INDArray arr)
Prints the shape
for this shape information
|
static String |
Shape.shapeToStringShort(INDArray arr) |
static boolean |
Shape.strideDescendingCAscendingF(INDArray array)
Check if strides are in order suitable for non-strided mmul etc.
|
static INDArray |
Shape.toMmulCompatible(INDArray input)
This method is used in DL4J LSTM implementation
|
static INDArray |
Shape.toOffsetZero(INDArray arr)
Create a copy of the matrix
where the new offset is zero
|
static INDArray |
Shape.toOffsetZeroCopy(INDArray arr)
Create a copy of the ndarray where the new offset is zero
|
static INDArray |
Shape.toOffsetZeroCopy(INDArray arr,
char order)
Create a copy of the ndarray where the new offset is zero, and has specified order
|
static INDArray |
Shape.toOffsetZeroCopyAnyOrder(INDArray arr)
Create a copy of the ndarray where the new offset is zero.
|
| Modifier and Type | Method and Description |
|---|---|
Pair<DataBuffer,DataBuffer> |
TADManager.getTADOnlyShapeInfo(INDArray array,
int... dimension)
This method returns TAD shapeInfo and all offsets
for specified tensor and dimensions.
|
| Constructor and Description |
|---|
TadDescriptor(INDArray array,
int[] dimension)
Pass in an ndarray to get the databuffer
and the appropriate dimensions
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
CheckUtil.convertFromApacheMatrix(org.apache.commons.math3.linear.RealMatrix matrix,
DataType dataType) |
| Modifier and Type | Method and Description |
|---|---|
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dPermutedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dPermutedWithShape(long seed,
long[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dReshapedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dReshapedWithShape(long seed,
long[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dSubArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dSubArraysWithShape(long seed,
long[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dTensorAlongDimensionWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get3dTensorAlongDimensionWithShape(long seed,
long[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get4dPermutedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get4dReshapedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get4dSubArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get4dTensorAlongDimensionWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get5dPermutedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get5dReshapedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get5dSubArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get5dTensorAlongDimensionWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get6dPermutedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get6dReshapedWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.get6dSubArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAll3dTestArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAll3dTestArraysWithShape(long seed,
long[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAll4dTestArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAll4dTestArraysWithShape(int seed,
long[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAll5dTestArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAll6dTestArraysWithShape(int seed,
int[] shape,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAllTestMatricesWithShape(char ordering,
int rows,
int cols,
int seed,
DataType dataType)
Get an array of INDArrays (2d) all with the specified shape.
|
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getAllTestMatricesWithShape(long rows,
long cols,
long seed,
DataType dataType)
Get an array of INDArrays (2d) all with the specified shape.
|
static Pair<INDArray,String> |
NDArrayCreationUtil.getPermutedWithShape(char ordering,
long rows,
long cols,
long seed,
DataType dataType) |
static Pair<INDArray,String> |
NDArrayCreationUtil.getPermutedWithShape(long rows,
long cols,
long seed,
DataType dataType) |
static Pair<INDArray,String> |
NDArrayCreationUtil.getReshapedWithShape(char ordering,
long rows,
long cols,
long seed,
DataType dataType) |
static Pair<INDArray,String> |
NDArrayCreationUtil.getReshapedWithShape(long rows,
long cols,
long seed,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getSubMatricesWithShape(char ordering,
long rows,
long cols,
long seed,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getSubMatricesWithShape(long rows,
long cols,
long seed,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getTensorAlongDimensionMatricesWithShape(char ordering,
long rows,
long cols,
long seed,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getTensorAlongDimensionMatricesWithShape(long rows,
long cols,
long seed,
DataType dataType) |
static List<Pair<INDArray,String>> |
NDArrayCreationUtil.getTestMatricesWithVaryingShapes(int rank,
char order,
DataType dataType)
Test utility to sweep shapes given a rank
Given a rank will generate random test matrices that will cover all cases of a shape with a '1' anywhere in the shape
as well a shape with random ints that are not 0 or 1
eg.
|
static Pair<INDArray,String> |
NDArrayCreationUtil.getTransposedMatrixWithShape(char ordering,
int rows,
int cols,
int seed,
DataType dataType) |
static Pair<INDArray,String> |
NDArrayCreationUtil.getTransposedMatrixWithShape(long rows,
long cols,
long seed,
DataType dataType) |
| Modifier and Type | Method and Description |
|---|---|
static boolean |
CheckUtil.checkAdd(INDArray first,
INDArray second,
double maxRelativeDifference,
double minAbsDifference)
Same as checkMmul, but for matrix addition
|
static boolean |
CheckUtil.checkDivManually(INDArray first,
INDArray second,
double maxRelativeDifference,
double minAbsDifference) |
static boolean |
CheckUtil.checkEntries(INDArray expected,
INDArray actual,
double maxRelativeDifference,
double minAbsDifference) |
static boolean |
CheckUtil.checkEntries(org.apache.commons.math3.linear.RealMatrix rmResult,
INDArray result,
double maxRelativeDifference,
double minAbsDifference) |
static boolean |
CheckUtil.checkGemm(INDArray a,
INDArray b,
INDArray c,
boolean transposeA,
boolean transposeB,
double alpha,
double beta,
double maxRelativeDifference,
double minAbsDifference) |
static boolean |
CheckUtil.checkMmul(INDArray first,
INDArray second,
double maxRelativeDifference,
double minAbsDifference)
Check first.mmul(second) using Apache commons math mmul.
|
static boolean |
CheckUtil.checkMulManually(INDArray first,
INDArray second,
double maxRelativeDifference,
double minAbsDifference) |
static boolean |
CheckUtil.checkSubtract(INDArray first,
INDArray second,
double maxRelativeDifference,
double minAbsDifference)
Same as checkMmul, but for matrix subtraction
|
static org.apache.commons.math3.linear.RealMatrix |
CheckUtil.convertToApacheMatrix(INDArray matrix) |
static void |
CheckUtil.printFailureDetails(INDArray first,
INDArray second,
INDArray expected,
INDArray actual,
INDArray onCopies,
String op) |
static void |
CheckUtil.printFailureDetails(INDArray first,
INDArray second,
org.apache.commons.math3.linear.RealMatrix expected,
INDArray actual,
INDArray onCopies,
String op) |
static void |
CheckUtil.printGemmFailureDetails(INDArray a,
INDArray b,
INDArray c,
boolean transposeA,
boolean transposeB,
double alpha,
double beta,
org.apache.commons.math3.linear.RealMatrix expected,
INDArray actual,
INDArray onCopies) |
static void |
CheckUtil.printMatrixFullPrecision(INDArray matrix) |
static void |
CheckUtil.printNDArrayHeader(INDArray array) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BasicNDArrayCompressor.compress(double[] array)
This method returns compressed INDArray instance which contains JVM array passed in
|
INDArray |
NDArrayCompressor.compress(double[] data)
This method creates compressed INDArray from Java double array, skipping usual INDArray instantiation routines
Please note: This method compresses input data as vector
|
INDArray |
NDArrayCompressor.compress(double[] data,
int[] shape,
char order)
This method creates compressed INDArray from Java double array, skipping usual INDArray instantiation routines
|
INDArray |
BasicNDArrayCompressor.compress(float[] array)
This method returns compressed INDArray instance which contains JVM array passed in
|
INDArray |
NDArrayCompressor.compress(float[] data)
This method creates compressed INDArray from Java float array, skipping usual INDArray instantiation routines
Please note: This method compresses input data as vector
|
INDArray |
NDArrayCompressor.compress(float[] data,
int[] shape,
char order)
This method creates compressed INDArray from Java float array, skipping usual INDArray instantiation routines
|
INDArray |
BasicNDArrayCompressor.compress(INDArray array) |
INDArray |
NDArrayCompressor.compress(INDArray array)
This method returns compressed copy of referenced array
|
INDArray |
BasicNDArrayCompressor.compress(INDArray array,
String algorithm)
Returns a compressed version of the
given ndarray
|
INDArray |
BasicNDArrayCompressor.decompress(INDArray array) |
INDArray |
NDArrayCompressor.decompress(INDArray array)
This method returns
decompressed copy of referenced array
|
INDArray |
AbstractStorage.get(T key)
Get object from the storage, by key
|
| Modifier and Type | Method and Description |
|---|---|
void |
BasicNDArrayCompressor.autoDecompress(INDArray... arrays)
Decompress several ndarrays
|
void |
BasicNDArrayCompressor.autoDecompress(INDArray array) |
INDArray |
BasicNDArrayCompressor.compress(INDArray array) |
INDArray |
NDArrayCompressor.compress(INDArray array)
This method returns compressed copy of referenced array
|
INDArray |
BasicNDArrayCompressor.compress(INDArray array,
String algorithm)
Returns a compressed version of the
given ndarray
|
void |
BasicNDArrayCompressor.compressi(INDArray array)
In place compression of the passed in ndarray
with the default compression algorithm
|
void |
NDArrayCompressor.compressi(INDArray array)
Inplace compression of INDArray
|
void |
BasicNDArrayCompressor.compressi(INDArray array,
String algorithm)
In place Compress the given ndarray
with the given algorithm
|
INDArray |
BasicNDArrayCompressor.decompress(INDArray array) |
INDArray |
NDArrayCompressor.decompress(INDArray array)
This method returns
decompressed copy of referenced array
|
void |
BasicNDArrayCompressor.decompressi(INDArray array)
in place decompression of the given
ndarray.
|
void |
NDArrayCompressor.decompressi(INDArray array)
Inplace decompression of INDArray
|
void |
AbstractStorage.store(T key,
INDArray object)
Store object into storage
|
boolean |
AbstractStorage.storeIfAbsent(T key,
INDArray object)
Store object into storage, if it doesn't exist
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Convolution.col2im(INDArray col,
INDArray z,
int sH,
int sW,
int pH,
int pW,
int kH,
int kW,
int dH,
int dW) |
static INDArray |
OldConvolution.col2im(INDArray col,
int[] stride,
int[] padding,
int height,
int width) |
static INDArray |
Convolution.col2im(INDArray col,
int[] stride,
int[] padding,
int height,
int width) |
static INDArray |
OldConvolution.col2im(INDArray col,
int sy,
int sx,
int ph,
int pw,
int h,
int w)
Rearrange matrix
columns into blocks
|
static INDArray |
Convolution.col2im(INDArray col,
int sH,
int sW,
int ph,
int pW,
int kH,
int kW)
Rearrange matrix
columns into blocks
|
INDArray |
ConvolutionInstance.conv2d(INDArray input,
INDArray kernel,
Convolution.Type type)
2d convolution (aka the last 2 dimensions
|
INDArray |
BaseConvolution.conv2d(INDArray input,
INDArray kernel,
Convolution.Type type)
2d convolution (aka the last 2 dimensions
|
static INDArray |
Convolution.conv2d(INDArray input,
INDArray kernel,
Convolution.Type type)
2d convolution (aka the last 2 dimensions
|
INDArray |
ConvolutionInstance.convn(INDArray input,
INDArray kernel,
Convolution.Type type)
ND Convolution
|
INDArray |
BaseConvolution.convn(INDArray input,
INDArray kernel,
Convolution.Type type)
ND Convolution
|
static INDArray |
Convolution.convn(INDArray input,
INDArray kernel,
Convolution.Type type)
ND Convolution
|
INDArray |
DefaultConvolutionInstance.convn(INDArray input,
INDArray kernel,
Convolution.Type type,
int[] axes)
ND Convolution
|
INDArray |
ConvolutionInstance.convn(INDArray input,
INDArray kernel,
Convolution.Type type,
int[] axes)
ND Convolution
|
static INDArray |
Convolution.convn(INDArray input,
INDArray kernel,
Convolution.Type type,
int[] axes)
ND Convolution
|
static INDArray |
OldConvolution.im2col(INDArray img,
int[] kernel,
int[] stride,
int[] padding) |
static INDArray |
Convolution.im2col(INDArray img,
int[] kernel,
int[] stride,
int[] padding) |
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
boolean isSameMode)
Implement column formatted images
|
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
boolean isSameMode,
INDArray out) |
static INDArray |
OldConvolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int pval,
boolean coverAll)
Implement column formatted images
|
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int pval,
boolean isSameMode)
Implement column formatted images
|
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int dh,
int dw,
boolean isSameMode) |
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int dH,
int dW,
boolean isSameMode,
INDArray out)
Execute im2col.
|
static INDArray |
Convolution.pooling2D(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int dh,
int dw,
boolean isSameMode,
Pooling2D.Pooling2DType type,
Pooling2D.Divisor divisor,
double extra,
int virtualHeight,
int virtualWidth,
INDArray out)
Pooling 2d implementation
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Convolution.col2im(INDArray col,
INDArray z,
int sH,
int sW,
int pH,
int pW,
int kH,
int kW,
int dH,
int dW) |
static INDArray |
OldConvolution.col2im(INDArray col,
int[] stride,
int[] padding,
int height,
int width) |
static INDArray |
Convolution.col2im(INDArray col,
int[] stride,
int[] padding,
int height,
int width) |
static INDArray |
OldConvolution.col2im(INDArray col,
int sy,
int sx,
int ph,
int pw,
int h,
int w)
Rearrange matrix
columns into blocks
|
static INDArray |
Convolution.col2im(INDArray col,
int sH,
int sW,
int ph,
int pW,
int kH,
int kW)
Rearrange matrix
columns into blocks
|
INDArray |
ConvolutionInstance.conv2d(INDArray input,
INDArray kernel,
Convolution.Type type)
2d convolution (aka the last 2 dimensions
|
INDArray |
BaseConvolution.conv2d(INDArray input,
INDArray kernel,
Convolution.Type type)
2d convolution (aka the last 2 dimensions
|
static INDArray |
Convolution.conv2d(INDArray input,
INDArray kernel,
Convolution.Type type)
2d convolution (aka the last 2 dimensions
|
INDArray |
ConvolutionInstance.convn(INDArray input,
INDArray kernel,
Convolution.Type type)
ND Convolution
|
INDArray |
BaseConvolution.convn(INDArray input,
INDArray kernel,
Convolution.Type type)
ND Convolution
|
static INDArray |
Convolution.convn(INDArray input,
INDArray kernel,
Convolution.Type type)
ND Convolution
|
INDArray |
DefaultConvolutionInstance.convn(INDArray input,
INDArray kernel,
Convolution.Type type,
int[] axes)
ND Convolution
|
INDArray |
ConvolutionInstance.convn(INDArray input,
INDArray kernel,
Convolution.Type type,
int[] axes)
ND Convolution
|
static INDArray |
Convolution.convn(INDArray input,
INDArray kernel,
Convolution.Type type,
int[] axes)
ND Convolution
|
static INDArray |
OldConvolution.im2col(INDArray img,
int[] kernel,
int[] stride,
int[] padding) |
static INDArray |
Convolution.im2col(INDArray img,
int[] kernel,
int[] stride,
int[] padding) |
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
boolean isSameMode)
Implement column formatted images
|
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
boolean isSameMode,
INDArray out) |
static INDArray |
OldConvolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int pval,
boolean coverAll)
Implement column formatted images
|
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int pval,
boolean isSameMode)
Implement column formatted images
|
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int dh,
int dw,
boolean isSameMode) |
static INDArray |
Convolution.im2col(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int dH,
int dW,
boolean isSameMode,
INDArray out)
Execute im2col.
|
static INDArray |
Convolution.pooling2D(INDArray img,
int kh,
int kw,
int sy,
int sx,
int ph,
int pw,
int dh,
int dw,
boolean isSameMode,
Pooling2D.Pooling2DType type,
Pooling2D.Divisor divisor,
double extra,
int virtualHeight,
int virtualWidth,
INDArray out)
Pooling 2d implementation
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
DataSet.exampleMaxs() |
INDArray |
DataSet.exampleMeans() |
INDArray |
DataSet.exampleSums() |
INDArray[] |
MultiDataSet.getFeatures() |
INDArray |
DataSet.getFeatures() |
INDArray |
MultiDataSet.getFeatures(int index) |
INDArray |
DataSet.getFeaturesMaskArray() |
INDArray |
MultiDataSet.getFeaturesMaskArray(int index) |
INDArray[] |
MultiDataSet.getFeaturesMaskArrays() |
INDArray[] |
MultiDataSet.getLabels() |
INDArray |
DataSet.getLabels()
Returns the labels for the dataset
|
INDArray |
MultiDataSet.getLabels(int index) |
INDArray |
DataSet.getLabelsMaskArray() |
INDArray |
MultiDataSet.getLabelsMaskArray(int index) |
INDArray[] |
MultiDataSet.getLabelsMaskArrays() |
| Modifier and Type | Method and Description |
|---|---|
void |
DataSet.addFeatureVector(INDArray toAdd)
Adds a feature for each example on to the current feature vector
|
void |
DataSet.addFeatureVector(INDArray feature,
int example)
The feature to add, and the example/row number
|
List<String> |
DataSet.getLabelNames(INDArray idxs) |
void |
DataSet.setFeatures(INDArray features) |
void |
MultiDataSet.setFeatures(INDArray[] features) |
void |
MultiDataSet.setFeatures(int idx,
INDArray features) |
void |
DataSet.setFeaturesMaskArray(INDArray featuresMask) |
void |
MultiDataSet.setFeaturesMaskArray(int idx,
INDArray maskArray) |
void |
MultiDataSet.setFeaturesMaskArrays(INDArray[] maskArrays) |
void |
DataSet.setLabels(INDArray labels) |
void |
MultiDataSet.setLabels(INDArray[] labels) |
void |
MultiDataSet.setLabels(int idx,
INDArray labels) |
void |
DataSet.setLabelsMaskArray(INDArray labelsMask) |
void |
MultiDataSet.setLabelsMaskArray(INDArray[] labelsMaskArrays) |
void |
MultiDataSet.setLabelsMaskArray(int idx,
INDArray labelsMaskArray) |
| Constructor and Description |
|---|
DataSet(INDArray first,
INDArray second)
Creates a dataset with the specified input matrix and labels
|
DataSet(INDArray features,
INDArray labels,
INDArray featuresMask,
INDArray labelsMask)
Create a dataset with the specified input INDArray and labels (output) INDArray, plus (optionally) mask arrays
for the features and labels
|
MultiDataSet(INDArray[] features,
INDArray[] labels)
MultiDataSet constructor with no mask arrays
|
MultiDataSet(INDArray[] features,
INDArray[] labels)
MultiDataSet constructor with no mask arrays
|
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays) |
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays) |
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays) |
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays) |
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays,
List<Serializable> exampleMetaData) |
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays,
List<Serializable> exampleMetaData) |
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays,
List<Serializable> exampleMetaData) |
MultiDataSet(INDArray[] features,
INDArray[] labels,
INDArray[] featuresMaskArrays,
INDArray[] labelsMaskArrays,
List<Serializable> exampleMetaData) |
MultiDataSet(INDArray features,
INDArray labels)
MultiDataSet constructor with single features/labels input, no mask arrays
|
MultiDataSet(INDArray features,
INDArray labels,
INDArray featuresMask,
INDArray labelsMask)
MultiDataSet constructor with single features/labels input, single mask arrays
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
DataSet.exampleMaxs() |
INDArray |
DataSet.exampleMeans() |
INDArray |
DataSet.exampleSums() |
INDArray[] |
MultiDataSet.getFeatures()
Get all of the input features, as an array of INDArrays
|
INDArray |
DataSet.getFeatures()
Returns the features array for the DataSet
|
INDArray |
MultiDataSet.getFeatures(int index)
Get a single feature/input array
|
INDArray |
DataSet.getFeaturesMaskArray()
Input mask array: a mask array for input, where each value is in {0,1} in order to specify whether an input is
actually present or not.
|
INDArray |
MultiDataSet.getFeaturesMaskArray(int index)
Get the specified feature mask array.
|
INDArray[] |
MultiDataSet.getFeaturesMaskArrays()
Get the feature mask arrays.
|
INDArray[] |
MultiDataSet.getLabels()
Get all of the labels, as an array of INDArrays
|
INDArray |
DataSet.getLabels() |
INDArray |
MultiDataSet.getLabels(int index)
Get a single label/output array
|
INDArray |
DataSet.getLabelsMaskArray()
Labels (output) mask array: a mask array for input, where each value is in {0,1} in order to specify whether an
output is actually present or not.
|
INDArray |
MultiDataSet.getLabelsMaskArray(int index)
Get the specified label mask array.
|
INDArray[] |
MultiDataSet.getLabelsMaskArrays()
Get the labels mask arrays.
|
static INDArray |
DataSetUtil.mergeMasks2d(long[] outShape,
INDArray[] arrays,
INDArray[] masks) |
static INDArray |
DataSetUtil.mergeMasks4d(INDArray[] featuresOrLabels,
INDArray[] masks) |
static INDArray |
DataSetUtil.mergePerOutputMasks2d(long[] outShape,
INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx) |
static INDArray |
DataSetUtil.mergePerOutputMasks2d(long[] outShape,
INDArray[] arrays,
INDArray[] masks)
Deprecated.
|
static INDArray |
DataSetUtil.tailor2d(@NonNull DataSet dataSet,
boolean areFeatures) |
static INDArray |
DataSetUtil.tailor2d(@NonNull INDArray data,
INDArray mask) |
static INDArray |
DataSetUtil.tailor3d2d(DataSet dataset,
boolean areFeatures)
Deprecated.
|
static INDArray |
DataSetUtil.tailor3d2d(@NonNull INDArray data,
INDArray mask) |
static INDArray |
DataSetUtil.tailor4d2d(DataSet dataset,
boolean areFeatures) |
static INDArray |
DataSetUtil.tailor4d2d(@NonNull INDArray data) |
| Modifier and Type | Method and Description |
|---|---|
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(@NonNull INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(@NonNull INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(INDArray[][] featuresToMerge,
INDArray[][] featureMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(INDArray[][] featuresToMerge,
INDArray[][] featureMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(@NonNull INDArray[] featuresToMerge,
INDArray[] featureMasksToMerge)
Merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(@NonNull INDArray[] featuresToMerge,
INDArray[] featureMasksToMerge)
Merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(@NonNull INDArray[][] labelsToMerge,
INDArray[][] labelMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified label and label mask arrays
(i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(@NonNull INDArray[][] labelsToMerge,
INDArray[][] labelMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified label and label mask arrays
(i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(INDArray[] labelsToMerge,
INDArray[] labelMasksToMerge)
Merge the specified labels and label mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(INDArray[] labelsToMerge,
INDArray[] labelMasksToMerge)
Merge the specified labels and label mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified time series (3d) arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified time series (3d) arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[] arrays,
INDArray[] masks)
Merge the specified time series (3d) arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[] arrays,
INDArray[] masks)
Merge the specified time series (3d) arrays and masks.
|
| Modifier and Type | Method and Description |
|---|---|
void |
DataSet.addFeatureVector(INDArray toAdd) |
void |
DataSet.addFeatureVector(INDArray feature,
int example) |
List<String> |
DataSet.getLabelNames(INDArray idxs) |
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(@NonNull INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(@NonNull INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge2d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 2d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.merge4d(INDArray[] arrays,
INDArray[] masks)
Merge the specified 4d arrays and masks.
|
static Pair<INDArray[],INDArray[]> |
DataSetUtil.mergeFeatures(@NonNull INDArray[][] featuresToMerge,
INDArray[][] featureMasksToMerge)
Merge all of the features arrays into one minibatch.
|
static Pair<INDArray[],INDArray[]> |
DataSetUtil.mergeFeatures(@NonNull INDArray[][] featuresToMerge,
INDArray[][] featureMasksToMerge)
Merge all of the features arrays into one minibatch.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(INDArray[][] featuresToMerge,
INDArray[][] featureMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(INDArray[][] featuresToMerge,
INDArray[][] featureMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(@NonNull INDArray[] featuresToMerge,
INDArray[] featureMasksToMerge)
Merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeFeatures(@NonNull INDArray[] featuresToMerge,
INDArray[] featureMasksToMerge)
Merge the specified features and mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(@NonNull INDArray[][] labelsToMerge,
INDArray[][] labelMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified label and label mask arrays
(i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(@NonNull INDArray[][] labelsToMerge,
INDArray[][] labelMasksToMerge,
int inOutIdx)
Extract out the specified column, and merge the specified label and label mask arrays
(i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(INDArray[] labelsToMerge,
INDArray[] labelMasksToMerge)
Merge the specified labels and label mask arrays (i.e., concatenate the examples)
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeLabels(INDArray[] labelsToMerge,
INDArray[] labelMasksToMerge)
Merge the specified labels and label mask arrays (i.e., concatenate the examples)
|
static INDArray |
DataSetUtil.mergeMasks2d(long[] outShape,
INDArray[] arrays,
INDArray[] masks) |
static INDArray |
DataSetUtil.mergeMasks2d(long[] outShape,
INDArray[] arrays,
INDArray[] masks) |
static INDArray |
DataSetUtil.mergeMasks4d(INDArray[] featuresOrLabels,
INDArray[] masks) |
static INDArray |
DataSetUtil.mergeMasks4d(INDArray[] featuresOrLabels,
INDArray[] masks) |
static INDArray |
DataSetUtil.mergePerOutputMasks2d(long[] outShape,
INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx) |
static INDArray |
DataSetUtil.mergePerOutputMasks2d(long[] outShape,
INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx) |
static INDArray |
DataSetUtil.mergePerOutputMasks2d(long[] outShape,
INDArray[] arrays,
INDArray[] masks)
Deprecated.
|
static INDArray |
DataSetUtil.mergePerOutputMasks2d(long[] outShape,
INDArray[] arrays,
INDArray[] masks)
Deprecated.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified time series (3d) arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[][] arrays,
INDArray[][] masks,
int inOutIdx)
Merge the specified time series (3d) arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[] arrays,
INDArray[] masks)
Merge the specified time series (3d) arrays and masks.
|
static Pair<INDArray,INDArray> |
DataSetUtil.mergeTimeSeries(INDArray[] arrays,
INDArray[] masks)
Merge the specified time series (3d) arrays and masks.
|
void |
DataSet.setFeatures(INDArray features)
Set the features array for the DataSet
|
void |
MultiDataSet.setFeatures(INDArray[] features)
Set all of the features arrays for the MultiDataSet
|
void |
MultiDataSet.setFeatures(int idx,
INDArray features)
Set a single features array (by index) for the MultiDataSet
|
void |
DataSet.setFeaturesMaskArray(INDArray inputMask)
Set the features mask array in this DataSet
|
void |
MultiDataSet.setFeaturesMaskArray(int idx,
INDArray maskArray)
Set a single feature mask array by index
|
void |
MultiDataSet.setFeaturesMaskArrays(INDArray[] maskArrays)
Set the feature mask arrays
|
void |
DataSet.setLabels(INDArray labels) |
void |
MultiDataSet.setLabels(INDArray[] labels)
Set all of the labels arrays for the MultiDataSet
|
void |
MultiDataSet.setLabels(int idx,
INDArray labels)
Set a single labels array (by index) for the MultiDataSet
|
void |
DataSet.setLabelsMaskArray(INDArray labelsMask)
Set the labels mask array in this data set
|
void |
MultiDataSet.setLabelsMaskArray(INDArray[] labels)
Set the labels mask arrays
|
void |
MultiDataSet.setLabelsMaskArray(int idx,
INDArray labelsMaskArray)
Set a single labels mask array by index
|
static void |
DataSetUtil.setMaskedValuesToZero(INDArray data,
INDArray mask) |
static INDArray |
DataSetUtil.tailor2d(@NonNull INDArray data,
INDArray mask) |
static INDArray |
DataSetUtil.tailor2d(@NonNull INDArray data,
INDArray mask) |
static INDArray |
DataSetUtil.tailor3d2d(@NonNull INDArray data,
INDArray mask) |
static INDArray |
DataSetUtil.tailor3d2d(@NonNull INDArray data,
INDArray mask) |
static INDArray |
DataSetUtil.tailor4d2d(@NonNull INDArray data) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
StandardScaler.getMean()
Deprecated.
|
INDArray |
StandardScaler.getStd()
Deprecated.
|
| Modifier and Type | Method and Description |
|---|---|
protected INDArray |
BaseDataFetcher.createInputMatrix(int numRows)
Creates a feature vector
|
protected INDArray |
BaseDataFetcher.createOutputMatrix(int numRows) |
protected INDArray |
BaseDataFetcher.createOutputVector(int outcomeLabel)
Creates an output label matrix
|
| Modifier and Type | Field and Description |
|---|---|
static INDArray |
VGG16ImagePreProcessor.VGG_MEAN_OFFSET_BGR |
static INDArray |
VGG16ImagePreProcessor.VGG_MEAN_OFFSET_BGR_FLOAT |
static INDArray |
VGG16ImagePreProcessor.VGG_MEAN_OFFSET_BGR_FLOAT16 |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
MultiNormalizerStandardize.getFeatureMean(int input) |
INDArray |
MultiNormalizerStandardize.getFeatureStd(int input) |
INDArray |
NormalizerMinMaxScaler.getLabelMax() |
INDArray |
MultiNormalizerMinMaxScaler.getLabelMax(int output) |
INDArray |
NormalizerStandardize.getLabelMean() |
INDArray |
MultiNormalizerStandardize.getLabelMean(int output) |
INDArray |
NormalizerMinMaxScaler.getLabelMin() |
INDArray |
MultiNormalizerMinMaxScaler.getLabelMin(int output) |
INDArray |
NormalizerStandardize.getLabelStd() |
INDArray |
MultiNormalizerStandardize.getLabelStd(int output) |
INDArray |
NormalizerMinMaxScaler.getMax() |
INDArray |
MultiNormalizerMinMaxScaler.getMax(int input) |
INDArray |
NormalizerStandardize.getMean() |
protected static INDArray |
VGG16ImagePreProcessor.getMeanFor(INDArray features) |
INDArray |
NormalizerMinMaxScaler.getMin() |
INDArray |
MultiNormalizerMinMaxScaler.getMin(int input) |
INDArray |
NormalizerStandardize.getStd() |
| Modifier and Type | Method and Description |
|---|---|
protected static INDArray |
VGG16ImagePreProcessor.getMeanFor(INDArray features) |
void |
ImagePreProcessingScaler.preProcess(INDArray features) |
void |
VGG16ImagePreProcessor.preProcess(INDArray features) |
void |
StandardizeStrategy.preProcess(INDArray array,
INDArray maskArray,
DistributionStats stats)
Normalize a data array
|
void |
MinMaxStrategy.preProcess(INDArray array,
INDArray maskArray,
MinMaxStats stats)
Normalize a data array
|
void |
NormalizerStrategy.preProcess(INDArray array,
INDArray maskArray,
S stats)
Normalize a data array
|
void |
StandardizeStrategy.revert(INDArray array,
INDArray maskArray,
DistributionStats stats)
Denormalize a data array
|
void |
MinMaxStrategy.revert(INDArray array,
INDArray maskArray,
MinMaxStats stats)
Denormalize a data array
|
void |
NormalizerStrategy.revert(INDArray array,
INDArray maskArray,
S stats)
Denormalize a data array
|
void |
ImagePreProcessingScaler.revertFeatures(INDArray features) |
void |
VGG16ImagePreProcessor.revertFeatures(INDArray features) |
void |
AbstractDataSetNormalizer.revertFeatures(INDArray features) |
void |
DataNormalization.revertFeatures(INDArray features)
Undo (revert) the normalization applied by this DataNormalization instance to the specified features array
|
void |
AbstractMultiDataSetNormalizer.revertFeatures(@NonNull INDArray[] features)
Undo (revert) the normalization applied by this normalizer to the features arrays
|
void |
MultiDataNormalization.revertFeatures(INDArray[] features)
Undo (revert) the normalization applied by this DataNormalization instance to the specified features array
|
void |
MultiNormalizerHybrid.revertFeatures(@NonNull INDArray[] features)
Undo (revert) the normalization applied by this DataNormalization instance to the entire inputs array
|
void |
ImageMultiPreProcessingScaler.revertFeatures(INDArray[] features) |
void |
AbstractMultiDataSetNormalizer.revertFeatures(@NonNull INDArray[] features,
INDArray[] maskArrays)
Undo (revert) the normalization applied by this normalizer to the features arrays
|
void |
AbstractMultiDataSetNormalizer.revertFeatures(@NonNull INDArray[] features,
INDArray[] maskArrays)
Undo (revert) the normalization applied by this normalizer to the features arrays
|
void |
MultiDataNormalization.revertFeatures(INDArray[] features,
INDArray[] featuresMask)
Undo (revert) the normalization applied by this DataNormalization instance to the specified features array
|
void |
MultiDataNormalization.revertFeatures(INDArray[] features,
INDArray[] featuresMask)
Undo (revert) the normalization applied by this DataNormalization instance to the specified features array
|
void |
MultiNormalizerHybrid.revertFeatures(@NonNull INDArray[] features,
INDArray[] maskArrays)
Undo (revert) the normalization applied by this DataNormalization instance to the entire inputs array
|
void |
MultiNormalizerHybrid.revertFeatures(@NonNull INDArray[] features,
INDArray[] maskArrays)
Undo (revert) the normalization applied by this DataNormalization instance to the entire inputs array
|
void |
ImageMultiPreProcessingScaler.revertFeatures(INDArray[] features,
INDArray[] featuresMask) |
void |
ImageMultiPreProcessingScaler.revertFeatures(INDArray[] features,
INDArray[] featuresMask) |
void |
MultiNormalizerHybrid.revertFeatures(@NonNull INDArray[] features,
INDArray[] maskArrays,
int input)
Undo (revert) the normalization applied by this DataNormalization instance to the features of a particular input
|
void |
MultiNormalizerHybrid.revertFeatures(@NonNull INDArray[] features,
INDArray[] maskArrays,
int input)
Undo (revert) the normalization applied by this DataNormalization instance to the features of a particular input
|
void |
ImagePreProcessingScaler.revertFeatures(INDArray features,
INDArray featuresMask) |
void |
VGG16ImagePreProcessor.revertFeatures(INDArray features,
INDArray featuresMask) |
void |
AbstractDataSetNormalizer.revertFeatures(INDArray features,
INDArray featuresMask) |
void |
DataNormalization.revertFeatures(INDArray features,
INDArray featuresMask)
Undo (revert) the normalization applied by this DataNormalization instance to the specified features array
|
void |
AbstractMultiDataSetNormalizer.revertFeatures(@NonNull INDArray features,
INDArray mask,
int input)
Undo (revert) the normalization applied by this normalizer to a specific features array.
|
void |
AbstractMultiDataSetNormalizer.revertFeatures(@NonNull INDArray features,
INDArray mask,
int input)
Undo (revert) the normalization applied by this normalizer to a specific features array.
|
void |
ImagePreProcessingScaler.revertLabels(INDArray labels) |
void |
VGG16ImagePreProcessor.revertLabels(INDArray labels) |
void |
AbstractDataSetNormalizer.revertLabels(INDArray labels)
Undo (revert) the normalization applied by this DataNormalization instance to the specified labels array.
|
void |
DataNormalization.revertLabels(INDArray labels)
Undo (revert) the normalization applied by this DataNormalization instance to the specified labels array.
|
void |
AbstractMultiDataSetNormalizer.revertLabels(INDArray[] labels)
Undo (revert) the normalization applied by this DataNormalization instance to the specified labels array.
|
void |
MultiDataNormalization.revertLabels(INDArray[] labels)
Undo (revert) the normalization applied by this DataNormalization instance to the specified labels array.
|
void |
MultiNormalizerHybrid.revertLabels(@NonNull INDArray[] labels)
Undo (revert) the normalization applied by this DataNormalization instance to the entire outputs array
|
void |
ImageMultiPreProcessingScaler.revertLabels(INDArray[] labels) |
void |
AbstractMultiDataSetNormalizer.revertLabels(@NonNull INDArray[] labels,
INDArray[] labelsMask)
Undo (revert) the normalization applied by this normalizer to the labels arrays.
|
void |
AbstractMultiDataSetNormalizer.revertLabels(@NonNull INDArray[] labels,
INDArray[] labelsMask)
Undo (revert) the normalization applied by this normalizer to the labels arrays.
|
void |
MultiDataNormalization.revertLabels(INDArray[] labels,
INDArray[] labelsMask)
Undo (revert) the normalization applied by this DataNormalization instance to the specified labels array.
|
void |
MultiDataNormalization.revertLabels(INDArray[] labels,
INDArray[] labelsMask)
Undo (revert) the normalization applied by this DataNormalization instance to the specified labels array.
|
void |
MultiNormalizerHybrid.revertLabels(@NonNull INDArray[] labels,
INDArray[] maskArrays)
Undo (revert) the normalization applied by this DataNormalization instance to the entire outputs array
|
void |
MultiNormalizerHybrid.revertLabels(@NonNull INDArray[] labels,
INDArray[] maskArrays)
Undo (revert) the normalization applied by this DataNormalization instance to the entire outputs array
|
void |
ImageMultiPreProcessingScaler.revertLabels(INDArray[] labels,
INDArray[] labelsMask) |
void |
ImageMultiPreProcessingScaler.revertLabels(INDArray[] labels,
INDArray[] labelsMask) |
void |
MultiNormalizerHybrid.revertLabels(@NonNull INDArray[] labels,
INDArray[] maskArrays,
int output)
Undo (revert) the normalization applied by this DataNormalization instance to the labels of a particular output
|
void |
MultiNormalizerHybrid.revertLabels(@NonNull INDArray[] labels,
INDArray[] maskArrays,
int output)
Undo (revert) the normalization applied by this DataNormalization instance to the labels of a particular output
|
void |
ImagePreProcessingScaler.revertLabels(INDArray labels,
INDArray labelsMask) |
void |
VGG16ImagePreProcessor.revertLabels(INDArray labels,
INDArray labelsMask) |
void |
AbstractDataSetNormalizer.revertLabels(INDArray labels,
INDArray labelsMask) |
void |
DataNormalization.revertLabels(INDArray labels,
INDArray labelsMask)
Undo (revert) the normalization applied by this DataNormalization instance to the specified labels array.
|
void |
AbstractMultiDataSetNormalizer.revertLabels(@NonNull INDArray labels,
INDArray mask,
int output)
Undo (revert) the normalization applied by this normalizer to a specific labels array.
|
void |
AbstractMultiDataSetNormalizer.revertLabels(@NonNull INDArray labels,
INDArray mask,
int output)
Undo (revert) the normalization applied by this normalizer to a specific labels array.
|
void |
NormalizerMinMaxScaler.setFeatureStats(@NonNull INDArray featureMin,
@NonNull INDArray featureMax) |
void |
NormalizerMinMaxScaler.setFeatureStats(@NonNull INDArray featureMin,
@NonNull INDArray featureMax) |
void |
NormalizerStandardize.setLabelStats(@NonNull INDArray labelMean,
@NonNull INDArray labelStd) |
void |
NormalizerStandardize.setLabelStats(@NonNull INDArray labelMean,
@NonNull INDArray labelStd) |
void |
NormalizerMinMaxScaler.setLabelStats(@NonNull INDArray labelMin,
@NonNull INDArray labelMax) |
void |
NormalizerMinMaxScaler.setLabelStats(@NonNull INDArray labelMin,
@NonNull INDArray labelMax) |
void |
ImagePreProcessingScaler.transform(INDArray features) |
void |
VGG16ImagePreProcessor.transform(INDArray features) |
void |
AbstractDataSetNormalizer.transform(INDArray features)
Transform the given INDArray
|
void |
DataNormalization.transform(INDArray features)
Transform the dataset
|
void |
ImagePreProcessingScaler.transform(INDArray features,
INDArray featuresMask) |
void |
VGG16ImagePreProcessor.transform(INDArray features,
INDArray featuresMask) |
void |
AbstractDataSetNormalizer.transform(INDArray features,
INDArray featuresMask) |
void |
DataNormalization.transform(INDArray features,
INDArray featuresMask)
Transform the features, with an optional mask array
|
void |
ImagePreProcessingScaler.transformLabel(INDArray label) |
void |
VGG16ImagePreProcessor.transformLabel(INDArray label) |
void |
AbstractDataSetNormalizer.transformLabel(INDArray label)
Transform the labels.
|
void |
DataNormalization.transformLabel(INDArray labels)
Transform the labels.
|
void |
ImagePreProcessingScaler.transformLabel(INDArray labels,
INDArray labelsMask) |
void |
VGG16ImagePreProcessor.transformLabel(INDArray labels,
INDArray labelsMask) |
void |
AbstractDataSetNormalizer.transformLabel(INDArray label,
INDArray labelsMask) |
void |
DataNormalization.transformLabel(INDArray labels,
INDArray labelsMask)
Transform the labels.
|
| Constructor and Description |
|---|
NormalizerStandardize(@NonNull INDArray featureMean,
@NonNull INDArray featureStd) |
NormalizerStandardize(@NonNull INDArray featureMean,
@NonNull INDArray featureStd) |
NormalizerStandardize(@NonNull INDArray featureMean,
@NonNull INDArray featureStd,
@NonNull INDArray labelMean,
@NonNull INDArray labelStd) |
NormalizerStandardize(@NonNull INDArray featureMean,
@NonNull INDArray featureStd,
@NonNull INDArray labelMean,
@NonNull INDArray labelStd) |
NormalizerStandardize(@NonNull INDArray featureMean,
@NonNull INDArray featureStd,
@NonNull INDArray labelMean,
@NonNull INDArray labelStd) |
NormalizerStandardize(@NonNull INDArray featureMean,
@NonNull INDArray featureStd,
@NonNull INDArray labelMean,
@NonNull INDArray labelStd) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BaseUnderSamplingPreProcessor.adjustMasks(INDArray label,
INDArray labelMask,
int minorityLabel,
double targetDist) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BaseUnderSamplingPreProcessor.adjustMasks(INDArray label,
INDArray labelMask,
int minorityLabel,
double targetDist) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
MinMaxStats.getRange()
Get the feature wise
range for the statistics.
|
| Modifier and Type | Method and Description |
|---|---|
MinMaxStats.Builder |
MinMaxStats.Builder.add(@NonNull INDArray data,
INDArray mask)
Add rows of data to the statistics
|
MinMaxStats.Builder |
MinMaxStats.Builder.add(@NonNull INDArray data,
INDArray mask)
Add rows of data to the statistics
|
NormalizerStats.Builder<S> |
NormalizerStats.Builder.add(INDArray data,
INDArray mask)
Add rows of data to the statistics
|
DistributionStats.Builder |
DistributionStats.Builder.add(@NonNull INDArray data,
INDArray mask)
Add rows of data to the statistics
|
DistributionStats.Builder |
DistributionStats.Builder.add(@NonNull INDArray data,
INDArray mask)
Add rows of data to the statistics
|
| Constructor and Description |
|---|
DistributionStats(@NonNull INDArray mean,
@NonNull INDArray std) |
DistributionStats(@NonNull INDArray mean,
@NonNull INDArray std) |
MinMaxStats(@NonNull INDArray lower,
@NonNull INDArray upper) |
MinMaxStats(@NonNull INDArray lower,
@NonNull INDArray upper) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
PCA.convertBackToFeatures(INDArray data)
Take the data that has been transformed to the principal components about the mean and
transform it back into the original feature set.
|
INDArray |
PCA.convertToComponents(INDArray data)
Takes a set of data on each row, with the same number of features as the constructing data
and returns the data in the coordinates of the basis set about the mean.
|
static INDArray[] |
PCA.covarianceMatrix(INDArray in)
Returns the covariance matrix of a data set of many records, each with N features.
|
INDArray |
PCA.generateGaussianSamples(long count)
Generates a set of count random samples with the same variance and mean and eigenvector/values
as the data set used to initialize the PCA object, with same number of features N.
|
INDArray |
PCA.getCovarianceMatrix() |
INDArray |
PCA.getEigenvalues() |
INDArray |
PCA.getEigenvectors() |
INDArray |
PCA.getMean() |
static INDArray |
PCA.pca_factor(INDArray A,
double variance,
boolean normalize)
Calculates pca vectors of a matrix, for a given variance.
|
static INDArray |
PCA.pca_factor(INDArray A,
int nDims,
boolean normalize)
Calculates pca factors of a matrix, for a flags number of reduced features
returns the factors to scale observations
The return is a factor matrix to reduce (normalized) feature sets
|
static INDArray |
PCA.pca(INDArray A,
double variance,
boolean normalize)
Calculates pca reduced value of a matrix, for a given variance.
|
static INDArray |
PCA.pca(INDArray A,
int nDims,
boolean normalize)
Calculates pca vectors of a matrix, for a flags number of reduced features
returns the reduced feature set
The return is a projection of A onto principal nDims components
To use the PCA: assume A is the original feature set
then project A onto a reduced set of features.
|
static INDArray |
PCA.pca2(INDArray in,
double variance)
This method performs a dimensionality reduction, including principal components
that cover a fraction of the total variance of the system.
|
static INDArray[] |
PCA.principalComponents(INDArray cov)
Calculates the principal component vectors and their eigenvalues (lambda) for the covariance matrix.
|
INDArray |
RandomProjection.project(INDArray data)
Create a copy random projection by using matrix product with a random matrix
|
INDArray |
RandomProjection.project(INDArray data,
INDArray result)
Create a copy random projection by using matrix product with a random matrix
|
INDArray |
RandomProjection.projecti(INDArray data)
Create an in-place random projection by using in-place matrix product with a random matrix
|
INDArray |
RandomProjection.projecti(INDArray data,
INDArray result)
Create an in-place random projection by using in-place matrix product with a random matrix
|
INDArray |
PCA.reducedBasis(double variance)
Return a reduced basis set that covers a certain fraction of the variance of the data
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
PCA.convertBackToFeatures(INDArray data)
Take the data that has been transformed to the principal components about the mean and
transform it back into the original feature set.
|
INDArray |
PCA.convertToComponents(INDArray data)
Takes a set of data on each row, with the same number of features as the constructing data
and returns the data in the coordinates of the basis set about the mean.
|
static INDArray[] |
PCA.covarianceMatrix(INDArray in)
Returns the covariance matrix of a data set of many records, each with N features.
|
double |
PCA.estimateVariance(INDArray data,
int ndims)
Estimate the variance of a single record with reduced # of dimensions.
|
static INDArray |
PCA.pca_factor(INDArray A,
double variance,
boolean normalize)
Calculates pca vectors of a matrix, for a given variance.
|
static INDArray |
PCA.pca_factor(INDArray A,
int nDims,
boolean normalize)
Calculates pca factors of a matrix, for a flags number of reduced features
returns the factors to scale observations
The return is a factor matrix to reduce (normalized) feature sets
|
static INDArray |
PCA.pca(INDArray A,
double variance,
boolean normalize)
Calculates pca reduced value of a matrix, for a given variance.
|
static INDArray |
PCA.pca(INDArray A,
int nDims,
boolean normalize)
Calculates pca vectors of a matrix, for a flags number of reduced features
returns the reduced feature set
The return is a projection of A onto principal nDims components
To use the PCA: assume A is the original feature set
then project A onto a reduced set of features.
|
static INDArray |
PCA.pca2(INDArray in,
double variance)
This method performs a dimensionality reduction, including principal components
that cover a fraction of the total variance of the system.
|
static INDArray[] |
PCA.principalComponents(INDArray cov)
Calculates the principal component vectors and their eigenvalues (lambda) for the covariance matrix.
|
INDArray |
RandomProjection.project(INDArray data)
Create a copy random projection by using matrix product with a random matrix
|
INDArray |
RandomProjection.project(INDArray data,
INDArray result)
Create a copy random projection by using matrix product with a random matrix
|
INDArray |
RandomProjection.projecti(INDArray data)
Create an in-place random projection by using in-place matrix product with a random matrix
|
INDArray |
RandomProjection.projecti(INDArray data,
INDArray result)
Create an in-place random projection by using in-place matrix product with a random matrix
|
static long[] |
RandomProjection.targetShape(INDArray X,
double eps)
Compute the target shape of a suitable projection matrix
|
protected static long[] |
RandomProjection.targetShape(INDArray X,
int targetDimension)
Compute the target shape of a suitable projection matrix
|
| Constructor and Description |
|---|
PCA(INDArray dataset)
Create a PCA instance with calculated data: covariance, mean, eigenvectors, and eigenvalues.
|
| Modifier and Type | Field and Description |
|---|---|
static INDArray |
Eigen.dummy |
| Modifier and Type | Method and Description |
|---|---|
static INDArray[] |
Eigen.eig(INDArray A)
Compute the eigenvalues and eigenvectors of a square matrix
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A)
Compute generalized eigenvalues of the problem A x = L x.
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A,
boolean calculateVectors)
Compute generalized eigenvalues of the problem A x = L x.
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A,
INDArray B)
Compute generalized eigenvalues of the problem A x = L B x.
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A,
INDArray B,
boolean calculateVectors)
Compute generalized eigenvalues of the problem A x = L B x.
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray[] |
Eigen.eig(INDArray A)
Compute the eigenvalues and eigenvectors of a square matrix
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A)
Compute generalized eigenvalues of the problem A x = L x.
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A,
boolean calculateVectors)
Compute generalized eigenvalues of the problem A x = L x.
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A,
INDArray B)
Compute generalized eigenvalues of the problem A x = L B x.
|
static INDArray |
Eigen.symmetricGeneralizedEigenvalues(INDArray A,
INDArray B,
boolean calculateVectors)
Compute generalized eigenvalues of the problem A x = L B x.
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Nd4j.accumulate(INDArray... arrays)
This method sums given arrays and stores them to a new array
|
static INDArray |
Nd4j.accumulate(INDArray target,
Collection<INDArray> arrays)
This method sums given arrays and stores them to a given target array
|
INDArray |
NDArrayFactory.accumulate(INDArray target,
INDArray... arrays)
This method sums given arrays to target
|
static INDArray |
Nd4j.accumulate(INDArray target,
INDArray[] arrays)
This method sums given arrays and stores them to a given target array
|
static INDArray |
Broadcast.add(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast add op.
|
static INDArray |
Broadcast.amax(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast absolute max op.
|
static INDArray |
Broadcast.amin(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast absolute min op.
|
static INDArray |
Nd4j.append(INDArray arr,
int padAmount,
double val,
int axis)
Append the given array with the specified value size along a particular axis.
|
INDArray |
BaseNDArrayFactory.appendBias(INDArray... vectors)
Merge the vectors and append a bias.
|
static INDArray |
Nd4j.appendBias(INDArray... vectors) |
INDArray |
NDArrayFactory.appendBias(INDArray... vectors) |
static INDArray |
Nd4j.arange(double end)
Create a 1D array of evenly spaced values between 0 (inclusive) and
end (exclusive)
with a step size of 1
See Nd4j.arange(double, double, double) with begin = 0 and step size 1. |
static INDArray |
Nd4j.arange(double begin,
double end)
Create a 1D array of evenly spaced values between
begin (inclusive) and end (exclusive)
with a step size of 1
See Nd4j.arange(double, double, double) with step size 1. |
INDArray |
BaseNDArrayFactory.arange(double begin,
double end,
double step)
Array of evenly spaced values.
|
static INDArray |
Nd4j.arange(double begin,
double end,
double step)
Create a 1D array of evenly spaced values between
begin (inclusive) and end (exclusive)
with a step size. |
INDArray |
NDArrayFactory.arange(double begin,
double end,
double step)
Array of evenly spaced values.
|
static INDArray |
Nd4j.argMax(INDArray arr,
int... dimension)
Get the maximum values for a dimension.
|
static INDArray |
Nd4j.argMin(INDArray arr,
int... dimension)
See
Nd4j.argMax(INDArray, int...) but return minimum values. |
INDArray |
NDArrayFactory.average(Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array
|
INDArray |
NDArrayFactory.average(INDArray[] arrays)
This method averages input arrays, and returns averaged array
|
INDArray |
NDArrayFactory.average(INDArray target,
Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array
|
INDArray |
NDArrayFactory.average(INDArray target,
INDArray[] arrays)
This method averages input arrays, and returns averaged array
|
static INDArray |
Nd4j.averageAndPropagate(Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array.
|
static INDArray |
Nd4j.averageAndPropagate(INDArray[] arrays)
This method averages input arrays, and returns averaged array.
|
static INDArray |
Nd4j.averageAndPropagate(INDArray target,
Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array.
|
INDArray |
BaseBlasWrapper.axpy(double da,
INDArray dx,
INDArray dy) |
INDArray |
BlasWrapper.axpy(double da,
INDArray dx,
INDArray dy)
Deprecated.
|
INDArray |
BaseBlasWrapper.axpy(float da,
INDArray dx,
INDArray dy) |
INDArray |
BlasWrapper.axpy(float da,
INDArray dx,
INDArray dy)
Deprecated.
|
INDArray |
BaseBlasWrapper.axpy(Number da,
INDArray dx,
INDArray dy) |
INDArray |
BlasWrapper.axpy(Number da,
INDArray dx,
INDArray dy)
Compute y <- y + x * alpha
|
INDArray |
BaseNDArrayFactory.bilinearProducts(INDArray curr,
INDArray in)
Returns a column vector where each entry is the nth bilinear
product of the nth slices of the two tensors.
|
INDArray |
NDArrayFactory.bilinearProducts(INDArray curr,
INDArray in)
Returns a column vector where each entry is the nth bilinear
product of the nth slices of the two tensors.
|
static INDArray |
Nd4j.choice(INDArray source,
INDArray probs,
INDArray target)
This method samples value from Source array to Target,the default random number generator.
|
static INDArray |
Nd4j.choice(@NonNull INDArray source,
@NonNull INDArray probs,
@NonNull INDArray target,
Random rng)
This method samples value from Source array to Target, with probabilites provided in Probs argument
|
static INDArray |
Nd4j.choice(INDArray source,
INDArray probs,
int numSamples)
This method returns new INDArray instance, sampled from Source array with probabilities given in Probs
using the default random number generator.
|
static INDArray |
Nd4j.choice(INDArray source,
INDArray probs,
int numSamples,
Random rng)
This method returns new INDArray instance, sampled from Source array with probabilities given in Probs.
|
INDArray |
BaseNDArrayFactory.concat(int dimension,
INDArray... toConcat)
concatenate ndarrays along a dimension
|
static INDArray |
Nd4j.concat(int dimension,
INDArray... toConcat)
Concatneate ndarrays along a dimension
|
INDArray |
NDArrayFactory.concat(int dimension,
INDArray... toConcat)
Concatneate ndarrays along a dimension
|
INDArray |
NDArrayFactory.convertDataEx(DataTypeEx typeSrc,
INDArray source,
DataTypeEx typeDst) |
INDArray |
BaseBlasWrapper.copy(INDArray x,
INDArray y) |
INDArray |
BlasWrapper.copy(INDArray x,
INDArray y)
Compute y <- x (copy a matrix)
|
static INDArray |
Broadcast.copy(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast copy op.
|
static INDArray |
Nd4j.create(boolean[] data)
Create a vector based on a java boolean array.
|
static INDArray |
Nd4j.create(boolean[][] data)
Create 2D boolean array based on java 2d boolean array.
|
static INDArray |
Nd4j.create(boolean[][] data,
long... shape)
Create a boolean array with given shape based on java 2d boolean array.
|
static INDArray |
Nd4j.create(boolean[] data,
long[] shape,
DataType type)
|
static INDArray |
Nd4j.create(boolean[] data,
long[] shape,
long[] strides,
char order,
DataType type)
|
INDArray |
NDArrayFactory.create(boolean[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(boolean[] data,
long[] shape,
long[] stride,
DataType dataType,
MemoryWorkspace workspace) |
static INDArray |
Nd4j.create(byte[] data,
long[] shape,
DataType type)
|
static INDArray |
Nd4j.create(byte[] data,
long[] shape,
long[] strides,
char order,
DataType type)
|
INDArray |
NDArrayFactory.create(byte[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(byte[] data,
long[] shape,
long[] stride,
DataType dataType,
MemoryWorkspace workspace) |
static INDArray |
Nd4j.create(@NonNull Collection<String> strings,
long... shape) |
static INDArray |
Nd4j.create(@NonNull Collection<String> strings,
long[] shape,
char order) |
INDArray |
NDArrayFactory.create(Collection<String> strings,
long[] shape,
char order) |
static INDArray |
Nd4j.create(DataBuffer buffer)
Create an array based on the data buffer.
|
INDArray |
NDArrayFactory.create(DataBuffer data)
Creates a row vector with the data
|
static INDArray |
Nd4j.create(DataBuffer data,
int... shape)
Create an array based on the data buffer with given shape.
|
INDArray |
NDArrayFactory.create(DataBuffer data,
int[] shape)
Create an ndrray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(DataBuffer buffer,
int[] shape,
int[] stride,
char order,
long offset) |
INDArray |
NDArrayFactory.create(DataBuffer buffer,
int[] shape,
int[] stride,
char order,
long offset) |
static INDArray |
Nd4j.create(DataBuffer data,
int[] shape,
int[] strides,
long offset)
Create an array based on the data buffer with given shape, stride and offset.
|
INDArray |
NDArrayFactory.create(DataBuffer data,
int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(DataBuffer data,
int[] newShape,
int[] newStride,
long offset,
char ordering)
|
INDArray |
NDArrayFactory.create(DataBuffer data,
int[] newShape,
int[] newStride,
long offset,
char ordering) |
static INDArray |
Nd4j.create(DataBuffer data,
int[] shape,
long offset)
|
INDArray |
NDArrayFactory.create(DataBuffer buffer,
int[] shape,
long offset) |
static INDArray |
Nd4j.create(DataBuffer data,
long... shape)
|
INDArray |
NDArrayFactory.create(DataBuffer data,
long[] shape) |
static INDArray |
Nd4j.create(DataBuffer data,
long[] shape,
long[] strides,
long offset)
|
INDArray |
NDArrayFactory.create(DataBuffer data,
long[] shape,
long[] stride,
long offset) |
static INDArray |
Nd4j.create(DataBuffer data,
long[] newShape,
long[] newStride,
long offset,
char ordering)
|
INDArray |
NDArrayFactory.create(DataBuffer data,
long[] newShape,
long[] newStride,
long offset,
char ordering) |
static INDArray |
Nd4j.create(DataBuffer data,
long[] newShape,
long[] newStride,
long offset,
char ordering,
DataType dataType)
Create an array based on the data buffer with given shape, stride, offset and data type.
|
INDArray |
NDArrayFactory.create(DataBuffer data,
long[] newShape,
long[] newStride,
long offset,
char ordering,
DataType dataType) |
static INDArray |
Nd4j.create(DataBuffer data,
long[] newShape,
long[] newStride,
long offset,
long ews,
char ordering)
Create an array based on the data buffer with given shape, stride and offset.
|
INDArray |
NDArrayFactory.create(DataBuffer data,
long[] newShape,
long[] newStride,
long offset,
long ews,
char ordering) |
INDArray |
NDArrayFactory.create(DataBuffer data,
long rows,
long columns,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(DataType type,
long... shape)
Create an array with specified shape and datatype.
|
static INDArray |
Nd4j.create(@NonNull DataType dataType,
@NonNull long[] shape,
char ordering)
Create an array with given data type shape and ordering.
|
INDArray |
NDArrayFactory.create(DataType dataType,
long[] shape,
char ordering,
MemoryWorkspace workspace) |
static INDArray |
Nd4j.create(DataType dataType,
@NonNull long[] shape,
long[] strides,
char ordering)
Create an array with given shape, stride and ordering.
|
INDArray |
NDArrayFactory.create(DataType dataType,
long[] shape,
long[] strides,
char ordering,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(DataType dataType,
long[] shape,
long[] paddings,
long[] paddingOffsets,
char ordering,
MemoryWorkspace workspace)
Create an ndArray with padded Buffer
|
INDArray |
BaseNDArrayFactory.create(double[] data)
Creates a row vector with the data
|
static INDArray |
Nd4j.create(double[] data)
Create double array based on java double array.
|
INDArray |
NDArrayFactory.create(double[] data)
Creates a row vector with the data
|
static INDArray |
Nd4j.create(double[][] data)
Create 2D double array based on java 2d double array.
|
INDArray |
NDArrayFactory.create(double[][] data)
Create an ndarray with the given data layout
|
static INDArray |
Nd4j.create(double[][][] data)
Create a 3D double array based on the 3D java double array.
|
static INDArray |
Nd4j.create(double[][][][] data)
Create 4D double array based on 4D java double array.
|
static INDArray |
Nd4j.create(double[][] data,
char ordering)
Create a 2D double array based on a 2D java double array with given ordering.
|
INDArray |
NDArrayFactory.create(double[][] data,
char ordering)
Create a matrix from the given
data and ordering
|
INDArray |
BaseNDArrayFactory.create(double[] data,
char order) |
static INDArray |
Nd4j.create(double[] data,
char order)
Creates a row vector with the data
|
INDArray |
NDArrayFactory.create(double[] data,
char order) |
INDArray |
BaseNDArrayFactory.create(double[] data,
int[] shape)
Create an ndrray with the specified shape
|
static INDArray |
Nd4j.create(double[] data,
int[] shape)
Create an array of the specified shape initialized with values from a java 1d array.
|
INDArray |
NDArrayFactory.create(double[] data,
int[] shape)
Create an ndrray with the specified shape
|
static INDArray |
Nd4j.create(double[] data,
int[] shape,
char ordering)
Create an array withgiven shape and ordering based on a java double array.
|
INDArray |
NDArrayFactory.create(double[] data,
int[] shape,
char ordering) |
INDArray |
BaseNDArrayFactory.create(double[] data,
int[] shape,
int[] stride,
char order,
long offset) |
INDArray |
NDArrayFactory.create(double[] data,
int[] shape,
int[] stride,
char order,
long offset) |
abstract INDArray |
BaseNDArrayFactory.create(double[] data,
int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(double[] data,
int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(double[] data,
int[] shape,
int[] stride,
long offset,
char ordering) |
INDArray |
NDArrayFactory.create(double[] data,
int[] shape,
long offset) |
static INDArray |
Nd4j.create(double[] data,
int[] shape,
long offset,
char ordering)
Create an array.
|
static INDArray |
Nd4j.create(double[] data,
long... shape)
|
INDArray |
BaseNDArrayFactory.create(double[] data,
long[] shape) |
INDArray |
NDArrayFactory.create(double[] data,
long[] shape) |
static INDArray |
Nd4j.create(double[] data,
long[] shape,
char ordering)
|
INDArray |
NDArrayFactory.create(double[] data,
long[] shape,
char ordering) |
static INDArray |
Nd4j.create(double[] data,
long[] shape,
DataType type)
|
static INDArray |
Nd4j.create(double[] data,
long[] shape,
long[] strides,
char order,
DataType type)
|
INDArray |
NDArrayFactory.create(double[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(double[] data,
long[] shape,
long[] stride,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(double[] data,
long[] shape,
long[] stride,
long offset) |
INDArray |
NDArrayFactory.create(double[] data,
long[] shape,
long[] stride,
long offset,
char ordering) |
static INDArray |
Nd4j.create(double[] data,
long[] shape,
long offset,
char ordering)
|
INDArray |
NDArrayFactory.create(double[] data,
long[] shape,
long offset,
Character order) |
INDArray |
BaseNDArrayFactory.create(double[] data,
long rows,
long columns,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(double[] data,
long rows,
long columns,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(float[] data)
Creates a row vector with the data
|
static INDArray |
Nd4j.create(float[] data)
Creates a row vector with the data
|
INDArray |
NDArrayFactory.create(float[] data)
Creates a row vector with the data
|
static INDArray |
Nd4j.create(float[][] data)
Create 2D float array based on java 2d float array.
|
INDArray |
NDArrayFactory.create(float[][] floats) |
static INDArray |
Nd4j.create(float[][][] data)
Create a 3D float array based on the 3D java float array.
|
static INDArray |
Nd4j.create(float[][][][] data)
Create 4D float array based on 4D java float array.
|
static INDArray |
Nd4j.create(float[][] data,
char ordering)
Create 2D float array based on java 2d float array and ordering.
|
INDArray |
NDArrayFactory.create(float[][] data,
char ordering) |
INDArray |
BaseNDArrayFactory.create(float[] data,
char order) |
static INDArray |
Nd4j.create(float[] data,
char order)
Creates a row vector with the data
|
INDArray |
NDArrayFactory.create(float[] data,
char order) |
INDArray |
BaseNDArrayFactory.create(float[] data,
int[] shape)
Create an ndrray with the specified shape
|
static INDArray |
Nd4j.create(float[] data,
int[] shape)
Create an ndrray with the specified shape
|
INDArray |
NDArrayFactory.create(float[] data,
int[] shape)
Create an ndrray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(float[] data,
int[] shape,
char ordering) |
static INDArray |
Nd4j.create(float[] data,
int[] shape,
char ordering)
|
INDArray |
NDArrayFactory.create(float[] data,
int[] shape,
char ordering) |
INDArray |
BaseNDArrayFactory.create(float[] data,
int[] shape,
int[] stride,
char order,
long offset) |
INDArray |
NDArrayFactory.create(float[] data,
int[] shape,
int[] stride,
char order,
long offset) |
abstract INDArray |
BaseNDArrayFactory.create(float[] data,
int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(float[] data,
int[] shape,
int[] stride,
long offset)
Create an array of the specified type, shape and stride initialized with values from a java 1d array using offset.
|
INDArray |
NDArrayFactory.create(float[] data,
int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(float[] data,
int[] shape,
int[] stride,
long offset,
char ordering) |
INDArray |
BaseNDArrayFactory.create(float[] data,
int[] shape,
long offset) |
INDArray |
NDArrayFactory.create(float[] data,
int[] shape,
long offset) |
INDArray |
NDArrayFactory.create(float[] data,
int[] shape,
long offset,
Character order) |
static INDArray |
Nd4j.create(float[] data,
long... shape)
|
INDArray |
BaseNDArrayFactory.create(float[] data,
long[] shape) |
INDArray |
NDArrayFactory.create(float[] data,
long[] shape) |
static INDArray |
Nd4j.create(float[] data,
long[] shape,
char ordering)
|
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
char ordering) |
static INDArray |
Nd4j.create(float[] data,
long[] shape,
DataType type)
|
static INDArray |
Nd4j.create(float[] data,
long[] shape,
long[] strides,
char order,
DataType type)
|
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
long[] stride,
char order,
DataType dataType) |
abstract INDArray |
BaseNDArrayFactory.create(float[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
long[] stride,
char order,
long offset) |
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
long[] stride,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
long[] stride,
long offset) |
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
long[] stride,
long offset,
char ordering) |
INDArray |
NDArrayFactory.create(float[] data,
long[] shape,
long offset,
Character order) |
INDArray |
BaseNDArrayFactory.create(float[] data,
long rows,
long columns,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(float[] data,
long rows,
long columns,
int[] stride,
long offset) |
INDArray |
NDArrayFactory.create(float[] data,
long rows,
long columns,
int[] stride,
long offset,
char ordering) |
static INDArray |
Nd4j.create(int... shape)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(int columns)
Creates a row vector with the specified number of columns
|
INDArray |
BaseNDArrayFactory.create(int[] shape)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(int[] shape)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(int[][] data)
Create 2D double array based on java 2d double array.
|
static INDArray |
Nd4j.create(int[][][] data)
create 3D int array based on 3D java int array.
|
static INDArray |
Nd4j.create(int[][][][] data)
Create 4D int array based on 4D java int array.
|
static INDArray |
Nd4j.create(@NonNull int[] shape,
char ordering)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(int[] shape,
char ordering) |
INDArray |
NDArrayFactory.create(int[] shape,
DataBuffer buffer)
Create an ndarray with the given shape
and data
|
static INDArray |
Nd4j.create(int[] shape,
DataType dataType)
Create an array of given shape and data type.
|
INDArray |
BaseNDArrayFactory.create(int[] shape,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(int[] shape,
DataType dataType,
MemoryWorkspace workspace) |
static INDArray |
Nd4j.create(int[] sliceShape,
double[]... arrays)
See
#create(int[], float[]...) |
static INDArray |
Nd4j.create(int[] sliceShape,
float[]... arrays)
Create an ndarray based on the given data
|
INDArray |
BaseNDArrayFactory.create(int[] shape,
int[] stride)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(int[] shape,
int[] stride)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(int[] shape,
int[] stride)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(int[] shape,
int[] stride,
char ordering)
Creates an ndarray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(int[] data,
int[] shape,
int[] stride,
char order,
long offset) |
INDArray |
NDArrayFactory.create(int[] shape,
int[] ints1,
int[] stride,
char order,
long offset) |
INDArray |
BaseNDArrayFactory.create(int[] ints,
int[] ints1,
int[] stride,
long offset) |
INDArray |
NDArrayFactory.create(int[] ints,
int[] ints1,
int[] stride,
long offset) |
INDArray |
BaseNDArrayFactory.create(int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(int[] shape,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(int[] shape,
int[] stride,
long offset,
char ordering) |
INDArray |
NDArrayFactory.create(int[] shape,
int[] stride,
long offset,
char ordering) |
static INDArray |
Nd4j.create(int[] data,
long[] shape,
DataType type)
Create an array of the specified type and shape initialized with values from a java 1d array.
|
static INDArray |
Nd4j.create(int[] data,
long[] shape,
long[] strides,
char order,
DataType type)
Create an array of the specified type, shape and stride initialized with values from a java 1d array.
|
INDArray |
NDArrayFactory.create(int[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(int[] data,
long[] shape,
long[] stride,
DataType dataType,
MemoryWorkspace workspace) |
static INDArray |
Nd4j.create(int columns,
char order)
Creates a row vector with the specified number of columns
|
static INDArray |
Nd4j.create(int rows,
int columns,
int[] stride)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(int rows,
int columns,
int[] stride,
char ordering)
Create a 2D array with given rows, columns, stride and ordering.
|
static INDArray |
Nd4j.create(int rows,
int columns,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
static INDArray |
Nd4j.create(List<? extends Number> list)
Creates a row vector with the data
|
static INDArray |
Nd4j.create(List<INDArray> list,
int... shape)
Creates an array with the specified shape from a list of arrays.
|
abstract INDArray |
BaseNDArrayFactory.create(List<INDArray> list,
int[] shape)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(List<INDArray> list,
int[] shape)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(List<INDArray> list,
int[] shape,
char ordering) |
static INDArray |
Nd4j.create(List<INDArray> list,
long... shape)
|
INDArray |
NDArrayFactory.create(List<INDArray> list,
long[] shape) |
INDArray |
NDArrayFactory.create(List<INDArray> list,
long[] shape,
char ordering) |
static INDArray |
Nd4j.create(long... shape)
Creates an ndarray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(long columns)
Creates a row vector with the specified number of columns
|
INDArray |
NDArrayFactory.create(long columns)
Creates a row vector with the specified number of columns
|
INDArray |
BaseNDArrayFactory.create(long[] shape)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(long[] shape) |
static INDArray |
Nd4j.create(long[][] data)
Create 2D long array based on java 2d long array.
|
static INDArray |
Nd4j.create(@NonNull long[] shape,
char ordering)
Create an array with given shape and ordering.
|
INDArray |
NDArrayFactory.create(long[] shape,
char ordering) |
INDArray |
BaseNDArrayFactory.create(long[] shape,
long[] stride) |
static INDArray |
Nd4j.create(long[] shape,
long[] stride)
|
INDArray |
NDArrayFactory.create(long[] shape,
long[] stride) |
static INDArray |
Nd4j.create(long[] shape,
long[] stride,
char ordering)
|
static INDArray |
Nd4j.create(long[] data,
long[] shape,
DataType type)
|
INDArray |
BaseNDArrayFactory.create(long[] shape,
long[] stride,
long offset) |
INDArray |
NDArrayFactory.create(long[] shape,
long[] stride,
long offset) |
static INDArray |
Nd4j.create(long[] data,
long[] shape,
long[] strides,
char order,
DataType type)
|
INDArray |
NDArrayFactory.create(long[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(long[] data,
long[] shape,
long[] stride,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
BaseNDArrayFactory.create(long[] shape,
long[] stride,
long offset,
char ordering) |
static INDArray |
Nd4j.create(long[] shape,
long[] stride,
long offset,
char ordering)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(long[] shape,
long[] stride,
long offset,
char ordering) |
INDArray |
BaseNDArrayFactory.create(long rows,
long columns)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(long rows,
long columns)
Creates an ndarray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(long rows,
long columns,
char ordering) |
static INDArray |
Nd4j.create(long rows,
long columns,
char ordering)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(long rows,
long columns,
char ordering) |
INDArray |
BaseNDArrayFactory.create(long rows,
long columns,
int[] stride)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(long rows,
long columns,
int[] stride)
Creates an ndarray with the specified shape
|
INDArray |
BaseNDArrayFactory.create(long rows,
long columns,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(long rows,
long columns,
int[] stride,
long offset)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(long rows,
long columns,
long[] stride,
long offset) |
static INDArray |
Nd4j.create(LongShapeDescriptor descriptor)
See
Nd4j.create(LongShapeDescriptor, boolean) with initialize set to true. |
static INDArray |
Nd4j.create(LongShapeDescriptor descriptor,
boolean initialize)
Create an ndarray based on the given description,
|
static INDArray |
Nd4j.create(short[] data,
long[] shape,
DataType type)
|
static INDArray |
Nd4j.create(short[] data,
long[] shape,
long[] strides,
char order,
DataType type)
|
INDArray |
NDArrayFactory.create(short[] data,
long[] shape,
long[] stride,
char order,
DataType dataType,
MemoryWorkspace workspace) |
INDArray |
NDArrayFactory.create(short[] data,
long[] shape,
long[] stride,
DataType dataType,
MemoryWorkspace workspace) |
static INDArray |
Nd4j.create(String... strings) |
static INDArray |
Nd4j.createArrayFromShapeBuffer(DataBuffer data,
DataBuffer shapeInfo)
Create array based in data buffer and shape info,
|
static INDArray |
Nd4j.createArrayFromShapeBuffer(DataBuffer data,
Pair<DataBuffer,long[]> shapeInfo)
Create array based in data buffer and shape info,
|
static INDArray |
Nd4j.createFromArray(boolean... array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Boolean[] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(boolean[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Boolean[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(boolean[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Boolean[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(boolean[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Boolean[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(byte... array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Byte[] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(byte[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Byte[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(byte[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Byte[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(byte[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Byte[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(double... array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Double[] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(double[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Double[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(double[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Double[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(double[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Double[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(float... array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Float[] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(float[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Float[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(float[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Float[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(float[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Float[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(int... array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(int[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(int[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(int[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Integer[] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Integer[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Integer[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Integer[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(long... array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Long[] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(long[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Long[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(long[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Long[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(long[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Long[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(short... array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Short[] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(short[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Short[][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(short[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Short[][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(short[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromArray(Short[][][][] array)
This method creates INDArray from provided jvm array
|
static INDArray |
Nd4j.createFromFlatArray(FlatArray array)
|
static INDArray |
Nd4j.createFromNpyFile(@NonNull File file)
Create an INDArray from a given Numpy .npy file.
|
INDArray |
NDArrayFactory.createFromNpyFile(File file)
Create from a given numpy file.
|
INDArray |
NDArrayFactory.createFromNpyHeaderPointer(org.bytedeco.javacpp.Pointer pointer)
Create from an in memory numpy header.
|
static INDArray |
Nd4j.createFromNpyPointer(org.bytedeco.javacpp.Pointer pointer)
Create from an in memory numpy pointer
|
INDArray |
NDArrayFactory.createFromNpyPointer(org.bytedeco.javacpp.Pointer pointer)
Create from an in memory numpy pointer
|
static INDArray |
Nd4j.createNpyFromByteArray(@NonNull byte[] input)
Create an
INDArray from the given numpy input.The numpy input follows the format: https://docs.scipy.org/doc/numpy-1.14.0/neps/npy-format.html |
static INDArray |
Nd4j.createNpyFromInputStream(@NonNull InputStream is)
Create a numpy array based on the passed in input stream
|
static INDArray |
Nd4j.createUninitialized(DataType type,
long... shape) |
static INDArray |
Nd4j.createUninitialized(DataType type,
long[] shape,
char ordering)
Creates an *uninitialized* array with the specified data type, shape and ordering.
|
INDArray |
NDArrayFactory.createUninitialized(DataType dataType,
long[] shape,
char ordering,
MemoryWorkspace workspace) |
static INDArray |
Nd4j.createUninitialized(int... shape)
|
static INDArray |
Nd4j.createUninitialized(int[] shape,
char ordering)
Creates an *uninitialized* array with the specified shape and ordering.
NOTE: The underlying memory (DataBuffer) will not be initialized. |
INDArray |
NDArrayFactory.createUninitialized(int[] shape,
char ordering) |
static INDArray |
Nd4j.createUninitialized(long... shape)
Creates an *uninitialized* ndarray with the specified shape and default ordering.
NOTE: The underlying memory (DataBuffer) will not be initialized. |
static INDArray |
Nd4j.createUninitialized(long length)
This method creates an *uninitialized* ndarray of specified length and default ordering.
|
static INDArray |
Nd4j.createUninitialized(long[] shape,
char ordering)
Creates an *uninitialized* array with the specified shape and ordering.
|
INDArray |
NDArrayFactory.createUninitialized(long[] shape,
char ordering) |
static INDArray |
Nd4j.createUninitializedDetached(DataType dataType,
char ordering,
long... shape)
Create an uninitialized ndArray.
|
INDArray |
NDArrayFactory.createUninitializedDetached(DataType dataType,
char ordering,
long... shape)
Create an uninitialized ndArray.
|
static INDArray |
Nd4j.createUninitializedDetached(DataType dataType,
long... shape)
See
Nd4j.createUninitializedDetached(DataType, char, long...) with default ordering. |
static INDArray |
Nd4j.cumsum(INDArray compute)
See
cumsum(int) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.cumsum(INDArray compute,
int dimension)
See
cumsum(int) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.diag(INDArray x)
Creates a new matrix where the values of the given vector are the diagonal values of
the matrix if a vector is passed in, if a matrix is returns the kth diagonal
in the matrix
|
static INDArray |
Broadcast.div(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast divide op.
|
static INDArray |
Nd4j.empty()
This method creates "empty" INDArray with datatype determined by
Nd4j.dataType() |
static INDArray |
Nd4j.empty(DataType type)
This method creates "empty" INDArray of the specified datatype
|
INDArray |
NDArrayFactory.empty(DataType type) |
static INDArray |
Broadcast.eq(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast equal to op.
|
static INDArray[] |
Nd4j.exec(CustomOp op)
Execute the operation and return the result
|
static INDArray[] |
Nd4j.exec(CustomOp op,
OpContext context)
Execute the operation and return the result
|
static INDArray |
Nd4j.exec(Op op)
Execute the operation and return the result
|
static INDArray |
Nd4j.exec(Op op,
OpContext context) |
static INDArray |
Nd4j.expandDims(INDArray input,
int dimension)
Expand the array dimensions.
|
INDArray |
BaseNDArrayFactory.eye(long n)
Create the identity ndarray
|
static INDArray |
Nd4j.eye(long n)
Create the identity ndarray
|
INDArray |
NDArrayFactory.eye(long n)
Create the identity ndarray
|
static INDArray |
Nd4j.fromByteArray(@NonNull byte[] arr)
Read an ndarray from a byte array
|
static INDArray |
Nd4j.gemm(INDArray a,
INDArray b,
boolean transposeA,
boolean transposeB)
matrix multiply: implements op(a)*op(b)
where op(x) means transpose x (or not) depending on
setting of arguments transposea and transposeb.
so gemm(a,b,false,false) == a.mmul(b), gemm(a,b,true,false) == a.transpose().mmul(b) etc. |
static INDArray |
Nd4j.gemm(INDArray a,
INDArray b,
INDArray c,
boolean transposeA,
boolean transposeB,
double alpha,
double beta)
Matrix multiply: Implements c = alpha*op(a)*op(b) + beta*c where op(X) means transpose X (or not)
depending on setting of arguments transposeA and transposeB.
Note that matrix c MUST be fortran order, have zero offset and have c.data().length == c.length(). |
INDArray |
BaseBlasWrapper.gemv(double alpha,
INDArray a,
INDArray x,
double beta,
INDArray y) |
INDArray |
BlasWrapper.gemv(double alpha,
INDArray a,
INDArray x,
double beta,
INDArray y)
Deprecated.
|
INDArray |
BaseBlasWrapper.gemv(float alpha,
INDArray a,
INDArray x,
float beta,
INDArray y) |
INDArray |
BlasWrapper.gemv(float alpha,
INDArray a,
INDArray x,
float beta,
INDArray y)
Deprecated.
|
INDArray |
BaseBlasWrapper.gemv(Number alpha,
INDArray a,
INDArray x,
double beta,
INDArray y) |
INDArray |
BlasWrapper.gemv(Number alpha,
INDArray a,
INDArray x,
double beta,
INDArray y)
************************************************************************
BLAS Level 2
|
INDArray |
BaseBlasWrapper.ger(double alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BlasWrapper.ger(double alpha,
INDArray x,
INDArray y,
INDArray a)
Deprecated.
|
INDArray |
BaseBlasWrapper.ger(float alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BlasWrapper.ger(float alpha,
INDArray x,
INDArray y,
INDArray a)
Compute A <- alpha * x * y^T + A (general rank-1 update)
|
INDArray |
BaseBlasWrapper.ger(Number alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BlasWrapper.ger(Number alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BaseBlasWrapper.gesv(INDArray a,
int[] ipiv,
INDArray b) |
INDArray |
BlasWrapper.gesv(INDArray a,
int[] ipiv,
INDArray b)
************************************************************************
LAPACK
|
static INDArray |
Broadcast.gt(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast greater than op.
|
static INDArray |
Broadcast.gte(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast greater than or equal to op.
|
static INDArray |
Nd4j.hstack(Collection<INDArray> arrs)
Concatenates two matrices horizontally.
|
INDArray |
BaseNDArrayFactory.hstack(INDArray... arrs)
Concatenates two matrices horizontally.
|
static INDArray |
Nd4j.hstack(INDArray... arrs)
Concatenates two matrices horizontally.
|
INDArray |
NDArrayFactory.hstack(INDArray... arrs)
Concatenates two matrices horizontally.
|
static INDArray |
Nd4j.linspace(@NonNull DataType dataType,
double lower,
double step,
long num)
Generate a linearly spaced 1d vector of the specified datatype
|
static INDArray |
Nd4j.linspace(@NonNull DataType dtype,
long lower,
long num,
long step)
Generate a linearly spaced vector
|
static INDArray |
Nd4j.linspace(double lower,
double upper,
long num,
@NonNull DataType dataType)
Generate a linearly spaced 1d vector of the specified datatype
|
static INDArray |
Nd4j.linspace(long lower,
long upper,
long num)
Generate a linearly spaced vector with default data type
|
static INDArray |
Nd4j.linspace(long lower,
long upper,
long num,
@NonNull DataType dtype)
Generate a linearly spaced vector
|
static INDArray |
Broadcast.lt(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast less than op.
|
static INDArray |
Broadcast.lte(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast less than or equal to op.
|
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b)
Matrix multiplication/dot product
See
Nd4j.matmul(INDArray, INDArray, INDArray, boolean, boolean, boolean) |
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b,
boolean transposeA,
boolean transposeB,
boolean transposeResult)
Matrix multiplication/dot product.
See Nd4j.matmul(INDArray, INDArray, INDArray, boolean, boolean, boolean) |
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b,
INDArray result)
Matrix multiplication/dot product.
See Nd4j.matmul(INDArray, INDArray, INDArray, boolean, boolean, boolean) |
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b,
INDArray result,
boolean transposeA,
boolean transposeB,
boolean transposeResult)
Matrix multiplication/dot product
Depending on inputs dimensionality output result might be different.
|
static INDArray |
Nd4j.max(INDArray compute)
See
max(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Broadcast.max(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast max op.
|
static INDArray |
Nd4j.max(INDArray compute,
int dimension)
See
max(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.mean(INDArray compute)
See
mean(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.mean(INDArray compute,
int dimension)
See
mean(int...)with Integer.MAX_VALUE for full array reduction. |
static INDArray[] |
Nd4j.meshgrid(@NonNull INDArray x,
@NonNull INDArray y)
Meshgrid op.
|
static INDArray |
Nd4j.min(INDArray compute)
See
min(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Broadcast.min(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast min op.
|
static INDArray |
Nd4j.min(INDArray compute,
int dimension)
See
min(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Broadcast.mul(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast element-wise multiply op.
|
static INDArray |
Broadcast.neq(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast not equal to op.
|
static INDArray |
Nd4j.norm1(INDArray compute)
See
norm1(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.norm1(INDArray compute,
int dimension)
See
norm1(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.norm2(INDArray compute)
See
norm2(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.norm2(INDArray compute,
int dimension)
See
norm2(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.normmax(INDArray compute)
See
normmax(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.normmax(INDArray compute,
int dimension)
See
normmax(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.ones(DataType dataType,
long... shape)
Creates an array with the specified datatype and shape, with values all set to 1
|
static INDArray |
Nd4j.ones(int... shape)
Create an ndarray of ones
|
INDArray |
BaseNDArrayFactory.ones(int[] shape)
Create an ndarray of ones
|
INDArray |
NDArrayFactory.ones(int[] shape)
Create an ndarray of ones
|
static INDArray |
Nd4j.ones(long... shape)
See {@link #ones(int...
|
INDArray |
BaseNDArrayFactory.ones(long columns)
Creates a row vector with the specified number of columns
|
INDArray |
NDArrayFactory.ones(long columns)
Creates a row vector with the specified number of columns
|
INDArray |
BaseNDArrayFactory.ones(long[] shape) |
INDArray |
NDArrayFactory.ones(long[] shape) |
INDArray |
BaseNDArrayFactory.ones(long rows,
long columns)
Creates a row vector with the specified number of columns
|
INDArray |
NDArrayFactory.ones(long rows,
long columns)
Creates a row vector with the specified number of columns
|
static INDArray |
Nd4j.onesLike(INDArray arr)
Ones like
|
static INDArray |
Nd4j.pad(INDArray toPad,
INDArray padding)
See
#pad(INDArray, INDArray, Mode, double) with zero padding (zeros for padValue). |
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull INDArray padWidth,
Pad.Mode padMode,
double padValue)
Pad the given ndarray to the size along each dimension.
|
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
int... padWidth)
|
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull int[][] padWidth)
|
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull int[][] padWidth,
Pad.Mode padMode,
double padValue)
See
#pad(INDArray, INDArray, Mode, double). |
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull int[] padWidth,
Pad.Mode padMode,
double padValue)
See
#pad(INDArray, INDArray, Mode, double), uses padWidth for all dimensions. |
static INDArray |
Nd4j.pile(@NonNull Collection<INDArray> arrays)
This method stacks vertically examples with the same shape, increasing result dimensionality.
|
static INDArray |
Nd4j.pile(INDArray... arrays)
This method stacks vertically examples with the same shape, increasing result dimensionality.
|
static INDArray |
Nd4j.prepend(INDArray arr,
int padAmount,
double val,
int axis)
|
static INDArray |
Nd4j.prod(INDArray compute)
See
prod(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.prod(INDArray compute,
int dimension)
See
prod(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.pullRows(INDArray source,
INDArray destination,
int sourceDimension,
int... indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some
dimension and specified indexes.
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
INDArray destination,
int sourceDimension,
int[] indexes)
* This method produces concatenated array, that consist from tensors,
fetched from source array, against some dimension and specified indexes
in to the destination array
|
static INDArray |
Nd4j.pullRows(INDArray source,
int sourceDimension,
int... indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
INDArray |
BaseNDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
INDArray |
BaseNDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes,
char order)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
static INDArray |
Nd4j.pullRows(INDArray source,
int sourceDimension,
int[] indexes,
char order)
This method produces concatenated array,
that consist from tensors,
fetched from source array,
against some dimension and specified indexes
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes,
char order)
This method produces concatenated array, that consist from tensors,
fetched from source array, against some dimension and specified indexes
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
int sourceDimension,
long[] indexes) |
static INDArray |
Nd4j.rand(char order,
int... shape)
Create a random ndarray with the given shape and array order
Values are sampled from a uniform distribution over (0, 1)
|
INDArray |
BaseNDArrayFactory.rand(char order,
int[] shape)
Create a random ndarray with the given shape and order
|
INDArray |
NDArrayFactory.rand(char order,
int[] shape)
Create a random ndarray with the given shape, and specified output order
|
INDArray |
BaseNDArrayFactory.rand(char order,
long[] shape) |
INDArray |
NDArrayFactory.rand(char order,
long[] shape)
Create a random ndarray with the given shape
and specified output order
|
INDArray |
BaseNDArrayFactory.rand(char order,
long rows,
long columns)
Create a random (uniform 0-1) NDArray with the specified shape and order
|
INDArray |
NDArrayFactory.rand(char order,
long rows,
long columns)
Create a random (uniform 0-1) NDArray with the specified shape and order
|
static INDArray |
Nd4j.rand(@NonNull DataType dataType,
char order,
int... shape)
Deprecated.
|
static INDArray |
Nd4j.rand(@NonNull DataType dataType,
char order,
long... shape)
Create a random ndarray with the given shape, data type, and array order
Values are sampled from a uniform distribution over (0, 1)
|
static INDArray |
Nd4j.rand(@NonNull DataType dataType,
int... shape)
Create a random ndarray with the given shape and data type
Values are sampled from a uniform distribution over (0, 1)
|
static INDArray |
Nd4j.rand(@NonNull DataType dataType,
int[] shape,
char order)
Deprecated.
use {@link Nd4j#rand(DataType, char, long...))
|
static INDArray |
Nd4j.rand(@NonNull DataType dataType,
long... shape)
Create a random ndarray with values from a uniform distribution over (0, 1) with the given shape and data type
|
static INDArray |
Nd4j.rand(@NonNull Distribution dist,
long... shape)
Create a random ndarray with the given shape using the given rng
|
static INDArray |
Nd4j.rand(double min,
double max,
Random rng,
long... shape)
Generates a random matrix between min and max
|
static INDArray |
Nd4j.rand(INDArray target)
Fill the given ndarray with random numbers drawn from a uniform distribution
|
static INDArray |
Nd4j.rand(INDArray target,
@NonNull Distribution dist)
Fill the given ndarray with random numbers drawn from the given distribution
|
static INDArray |
Nd4j.rand(INDArray target,
double min,
double max,
Random rng)
Fill the given ndarray with random numbers drawn from a uniform distribution using the given RandomGenerator
|
static INDArray |
Nd4j.rand(INDArray target,
long seed)
Fill the given ndarray with random numbers drawn from a uniform distribution
|
static INDArray |
Nd4j.rand(INDArray target,
Random rng)
Fill the given ndarray with random numbers drawn from a uniform distribution using the given RandomGenerator
|
static INDArray |
Nd4j.rand(int... shape)
Create a random ndarray with values from a uniform distribution over (0, 1) with the given shape
|
INDArray |
BaseNDArrayFactory.rand(int[] shape)
Create a random ndarray with the given shape using
the current time as the seed
|
INDArray |
NDArrayFactory.rand(int[] shape)
Create a random ndarray with the given shape using
the current time as the seed
|
INDArray |
BaseNDArrayFactory.rand(int[] shape,
Distribution r)
Create a random ndarray with the given shape using the given rng
|
static INDArray |
Nd4j.rand(int[] shape,
@NonNull Distribution dist)
Deprecated.
|
INDArray |
NDArrayFactory.rand(int[] shape,
Distribution r)
Create a random ndarray with the given shape using the given rng
|
INDArray |
BaseNDArrayFactory.rand(int[] shape,
double min,
double max,
Random rng) |
static INDArray |
Nd4j.rand(int[] shape,
double min,
double max,
Random rng)
|
INDArray |
NDArrayFactory.rand(int[] shape,
double min,
double max,
Random rng) |
INDArray |
BaseNDArrayFactory.rand(int[] shape,
float min,
float max,
Random rng)
Generates a random matrix between min and max
|
INDArray |
NDArrayFactory.rand(int[] shape,
float min,
float max,
Random rng)
Generates a random matrix between min and max
|
INDArray |
BaseNDArrayFactory.rand(int[] shape,
long seed)
Create a random ndarray with the given shape using the given rng
|
static INDArray |
Nd4j.rand(int[] shape,
long seed)
Deprecated.
|
INDArray |
NDArrayFactory.rand(int[] shape,
long seed)
Create a random ndarray with the given shape using the given rng
|
INDArray |
BaseNDArrayFactory.rand(int[] shape,
Random r)
Create a random ndarray with the given shape using the given rng
|
static INDArray |
Nd4j.rand(int[] shape,
Random rng)
Deprecated.
|
INDArray |
NDArrayFactory.rand(int[] shape,
Random r)
Create a random ndarray with the given shape using the given rng
|
static INDArray |
Nd4j.rand(int rows,
int columns,
Random rng)
Create a random ndarray with the given shape using the given rng
|
static INDArray |
Nd4j.rand(long... shape)
See
Nd4j.rand(int[]) |
INDArray |
BaseNDArrayFactory.rand(long[] shape) |
INDArray |
NDArrayFactory.rand(long[] shape)
Create a random ndarray with the given shape using
the current time as the seed
|
static INDArray |
Nd4j.rand(long[] shape,
@NonNull Distribution dist)
Deprecated.
|
INDArray |
BaseNDArrayFactory.rand(long[] shape,
double min,
double max,
Random rng) |
static INDArray |
Nd4j.rand(long[] shape,
double min,
double max,
Random rng)
|
INDArray |
NDArrayFactory.rand(long[] shape,
double min,
double max,
Random rng) |
INDArray |
BaseNDArrayFactory.rand(long[] shape,
float min,
float max,
Random rng) |
INDArray |
NDArrayFactory.rand(long[] shape,
float min,
float max,
Random rng) |
INDArray |
BaseNDArrayFactory.rand(long[] shape,
long seed) |
INDArray |
NDArrayFactory.rand(long[] shape,
long seed) |
INDArray |
BaseNDArrayFactory.rand(long[] shape,
Random r) |
INDArray |
NDArrayFactory.rand(long[] shape,
Random r) |
static INDArray |
Nd4j.rand(long seed,
long... shape)
Create a random ndarray with values from a uniform distribution over (0, 1) with the given shape
using given seed
|
INDArray |
BaseNDArrayFactory.rand(long rows,
long columns)
Create a random ndarray with the given shape using
the current time as the seed
|
INDArray |
NDArrayFactory.rand(long rows,
long columns)
Create a random ndarray with the given shape using
the current time as the seed
|
INDArray |
BaseNDArrayFactory.rand(long rows,
long columns,
double min,
double max,
Random rng) |
INDArray |
NDArrayFactory.rand(long rows,
long columns,
double min,
double max,
Random rng) |
INDArray |
BaseNDArrayFactory.rand(long rows,
long columns,
float min,
float max,
Random rng)
Generates a random matrix between min and max
|
INDArray |
NDArrayFactory.rand(long rows,
long columns,
float min,
float max,
Random rng)
Generates a random matrix between min and max
|
INDArray |
BaseNDArrayFactory.rand(long rows,
long columns,
long seed)
Create a random ndarray with the given shape using the given rng
|
INDArray |
NDArrayFactory.rand(long rows,
long columns,
long seed)
Create a random ndarray with the given shape using the given rng
|
INDArray |
BaseNDArrayFactory.rand(long rows,
long columns,
Random r)
Create a random ndarray with the given shape using the given rng
|
INDArray |
NDArrayFactory.rand(long rows,
long columns,
Random r)
Create a random ndarray with the given shape using the given rng
|
static INDArray |
Nd4j.rand(Random rng,
long... shape)
Create a random ndarray with the given shape using the given RandomGenerator
|
static INDArray |
Nd4j.randn(char order,
int... shape)
Random normal N(0,1) with the specified shape and array order
|
INDArray |
BaseNDArrayFactory.randn(char order,
int[] shape)
Random normal using the current time stamp
as the seed
|
INDArray |
NDArrayFactory.randn(char order,
int[] shape)
Random normal N(0,1) with the specified shape and order
|
static INDArray |
Nd4j.randn(char order,
long... shape)
Random normal N(0,1) with the specified shape and array order
|
INDArray |
BaseNDArrayFactory.randn(char order,
long[] shape) |
INDArray |
NDArrayFactory.randn(char order,
long[] shape)
Random normal N(0,1) with the specified shape and order
|
INDArray |
BaseNDArrayFactory.randn(char order,
long rows,
long columns)
Generate a random normal N(0,1) with the specified order and shape
|
INDArray |
NDArrayFactory.randn(char order,
long rows,
long columns)
Random normal N(0,1), with specified output order
|
static INDArray |
Nd4j.randn(@NonNull DataType dataType,
char order,
long... shape)
Random normal N(0,1) with the specified shape and array order
|
static INDArray |
Nd4j.randn(@NonNull DataType dataType,
@NonNull int[] shape)
Create a ndarray of the given shape and data type with values from N(0,1)
|
static INDArray |
Nd4j.randn(@NonNull DataType dataType,
long... shape)
Create a ndarray of the given shape and data type with values from N(0,1)
|
static INDArray |
Nd4j.randn(double mean,
double stddev,
INDArray target,
Random rng) |
static INDArray |
Nd4j.randn(double mean,
double stddev,
long[] shape,
Random rng) |
static INDArray |
Nd4j.randn(INDArray target)
Fill the given ndarray with random numbers drawn from a normal distribution
|
static INDArray |
Nd4j.randn(INDArray target,
long seed)
Fill the given ndarray with random numbers drawn from a normal distribution
|
static INDArray |
Nd4j.randn(INDArray target,
Random rng)
Fill the given ndarray with random numbers drawn from a normal distribution utilizing the given random generator
|
INDArray |
BaseNDArrayFactory.randn(int[] shape)
Random normal N(0,1) with the specified shape and
|
static INDArray |
Nd4j.randn(@NonNull int[] shape)
Create a ndarray of the given shape with values from N(0,1)
|
INDArray |
NDArrayFactory.randn(int[] shape)
Random normal N(0,1) using the current time stamp
as the seed
|
INDArray |
BaseNDArrayFactory.randn(int[] shape,
long seed)
Random normal using the specified seed
|
INDArray |
NDArrayFactory.randn(int[] shape,
long seed)
Random normal using the specified seed
|
INDArray |
BaseNDArrayFactory.randn(int[] shape,
Random r)
Random normal using the given rng
|
static INDArray |
Nd4j.randn(int[] shape,
Random r)
Deprecated.
|
INDArray |
NDArrayFactory.randn(int[] shape,
Random r)
Random normal using the given rng
|
static INDArray |
Nd4j.randn(long... shape)
Create a ndarray of the given shape with values from N(0,1).
|
INDArray |
BaseNDArrayFactory.randn(long[] shape) |
INDArray |
NDArrayFactory.randn(long[] shape)
Random normal N(0,1) using the current time stamp
as the seed
|
INDArray |
BaseNDArrayFactory.randn(long[] shape,
long seed) |
INDArray |
NDArrayFactory.randn(long[] shape,
long seed)
Random normal using the specified seed
|
INDArray |
BaseNDArrayFactory.randn(long[] shape,
Random r) |
static INDArray |
Nd4j.randn(long[] shape,
Random r)
Deprecated.
|
INDArray |
NDArrayFactory.randn(long[] shape,
Random r) |
static INDArray |
Nd4j.randn(long seed,
int[] shape)
Deprecated.
|
INDArray |
BaseNDArrayFactory.randn(long rows,
long columns)
Random normal using the current time stamp
as the seed
|
INDArray |
NDArrayFactory.randn(long rows,
long columns)
Random normal (N(0,1)) using the current time stamp
as the seed
|
static INDArray |
Nd4j.randn(long seed,
@NonNull long[] shape)
Random normal N(0, 1) using the specified seed
|
INDArray |
BaseNDArrayFactory.randn(long rows,
long columns,
long seed)
Random normal using the specified seed
|
INDArray |
NDArrayFactory.randn(long rows,
long columns,
long seed)
Random normal using the specified seed
|
INDArray |
BaseNDArrayFactory.randn(long rows,
long columns,
Random r)
Random normal using the given rng
|
INDArray |
NDArrayFactory.randn(long rows,
long columns,
Random r)
Random normal using the given rng
|
static INDArray |
Nd4j.randn(Random r,
long... shape)
Random normal using the given rng
|
static INDArray |
Nd4j.randomBernoulli(double p,
@NonNull INDArray target)
Fill the specified array with values generated according to a binomial distribution with probability p: i.e.,
values 0 with probability (1-p) or value 1 with probability p
|
static INDArray |
Nd4j.randomBernoulli(double p,
long... shape)
Generate a random array according to a binomial distribution with probability p: i.e., values 0 with probability
(1-p) or value 1 with probability p
|
static INDArray |
Nd4j.randomBinomial(int nTrials,
double p,
INDArray target)
Fill the target array with random values generated according to a binomial distribution with the specified
number of trials and probability
|
static INDArray |
Nd4j.randomBinomial(int nTrials,
double p,
long... shape)
Generate an array with random values generated according to a binomial distribution with the specified
number of trials and probability
|
static INDArray |
Nd4j.randomExponential(double lambda,
INDArray target)
Exponential distribution: P(x) = lambda * exp(-lambda * x)
|
static INDArray |
Nd4j.randomExponential(double lambda,
long... shape)
Exponential distribution: P(x) = lambda * exp(-lambda * x)
|
static INDArray |
Broadcast.rdiv(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast reverse division op.
|
static INDArray |
Nd4j.read(DataInputStream dis)
Read in an ndarray from a data input stream
|
static INDArray |
Nd4j.read(InputStream reader)
Raad an ndarray from an input stream
See
Nd4j.read(DataInputStream) |
static INDArray |
Nd4j.readBinary(File read)
Read a binary ndarray from the given file
|
static INDArray |
Nd4j.readNpy(@NonNull File file)
Create an INDArray from a given Numpy .npy file.
|
static INDArray |
Nd4j.readNpy(@NonNull String path)
Create an INDArray from a given Numpy .npy file.
|
static INDArray |
Nd4j.readNumpy(@NonNull DataType dataType,
@NonNull InputStream filePath,
@NonNull String split,
@NonNull Charset charset)
Read array from input stream.
|
static INDArray |
Nd4j.readNumpy(DataType dataType,
String filePath)
Read array.
See Nd4j.readNumpy(DataType, InputStream, String , Charset) with default split and UTF-8 encoding. |
static INDArray |
Nd4j.readNumpy(DataType dataType,
String filePath,
String split)
Read array via input stream.
|
static INDArray |
Nd4j.readNumpy(@NonNull InputStream filePath,
@NonNull String split)
Read line via input streams
|
static INDArray |
Nd4j.readNumpy(String filePath)
Read line via input streams
|
static INDArray |
Nd4j.readNumpy(String filePath,
String split)
Read line via input streams
|
static INDArray |
Nd4j.readTxt(String filePath)
Read line via input streams
|
static INDArray |
Nd4j.readTxtString(InputStream ndarray)
Read line via input streams
|
static INDArray |
Nd4j.repeat(INDArray n,
int num)
Create an n x (shape)
ndarray where the ndarray is repeated num times
|
INDArray |
BaseNDArrayFactory.reverse(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
static INDArray |
Nd4j.reverse(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
INDArray |
NDArrayFactory.reverse(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
static INDArray |
Nd4j.rollAxis(INDArray a,
int axis)
Roll the specified axis backwards,
until it lies in a given position.
|
static INDArray |
Nd4j.rollAxis(INDArray a,
int axis,
int start)
Roll the specified axis backwards,
until it lies in a given position.
|
INDArray |
BaseNDArrayFactory.rot(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
INDArray |
NDArrayFactory.rot(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
static INDArray |
Broadcast.rsub(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast reverse subtraction op.
|
INDArray |
BaseBlasWrapper.scal(double alpha,
INDArray x) |
INDArray |
BlasWrapper.scal(double alpha,
INDArray x)
Deprecated.
|
INDArray |
BaseBlasWrapper.scal(float alpha,
INDArray x) |
INDArray |
BlasWrapper.scal(float alpha,
INDArray x)
Deprecated.
|
static INDArray |
Nd4j.scalar(boolean value)
Create a scalar NDArray with the specified value and BOOLEAN datatype
|
INDArray |
BaseNDArrayFactory.scalar(DataType dataType) |
INDArray |
NDArrayFactory.scalar(DataType dataType)
Create a scalar nd array with the data type
and a default value depending on the data type.
|
static INDArray |
Nd4j.scalar(DataType dataType,
Number value)
Create a scalar ndarray with the specified value and datatype
|
INDArray |
BaseNDArrayFactory.scalar(double value)
Create a scalar nd array with the specified value and offset
|
static INDArray |
Nd4j.scalar(double value)
Create a scalar nd array with the specified value
|
INDArray |
NDArrayFactory.scalar(double value)
Create a scalar nd array with the specified value and offset
|
INDArray |
BaseNDArrayFactory.scalar(double value,
long offset)
Create a scalar nd array with the specified value and offset
|
INDArray |
NDArrayFactory.scalar(double value,
long offset)
Create a scalar nd array with the specified value and offset
|
INDArray |
BaseNDArrayFactory.scalar(float value) |
static INDArray |
Nd4j.scalar(float value)
Create a scalar NDArray with the specified value and FLOAT datatype
|
INDArray |
NDArrayFactory.scalar(float value)
Create a scalar nd array with the specified value and offset
|
INDArray |
BaseNDArrayFactory.scalar(float value,
long offset)
Create a scalar nd array with the specified value and offset
|
INDArray |
NDArrayFactory.scalar(float value,
long offset)
Create a scalar nd array with the specified value and offset
|
static INDArray |
Nd4j.scalar(int value)
Create a scalar NDArray with the specified value and INT datatype
|
INDArray |
BaseNDArrayFactory.scalar(int value,
long offset)
Create a scalar nd array with the specified value and offset
|
INDArray |
NDArrayFactory.scalar(int value,
long offset) |
static INDArray |
Nd4j.scalar(long value)
Create a scalar NDArray with the specified value and LONG datatype
|
INDArray |
BaseNDArrayFactory.scalar(Number value)
Create a scalar ndarray with the specified offset
|
static INDArray |
Nd4j.scalar(Number value)
Create a scalar ndarray with the specified value
|
INDArray |
NDArrayFactory.scalar(Number value)
Create a scalar ndarray with the specified offset
|
static INDArray |
Nd4j.scalar(@NonNull String string) |
static INDArray |
Nd4j.sort(INDArray ndarray,
boolean ascending)
Sort all elements of an array.
|
INDArray |
NDArrayFactory.sort(INDArray x,
boolean descending) |
INDArray |
NDArrayFactory.sort(INDArray x,
boolean descending,
int... dimensions) |
static INDArray |
Nd4j.sort(INDArray ndarray,
int dimension,
boolean ascending)
Sort an ndarray along a particular dimension
Note that the input array is modified in-place. |
static INDArray |
Nd4j.sortColumns(INDArray in,
int rowIdx,
boolean ascending)
Sort (shuffle) the columns of a 2d array according to the value at a specified row.
|
INDArray |
NDArrayFactory.sortCooIndices(INDArray x) |
static INDArray |
Nd4j.sortRows(INDArray in,
int colIdx,
boolean ascending)
Sort (shuffle) the rows of a 2d array according to the value at a specified column.
|
static INDArray[] |
Nd4j.sortWithIndices(INDArray ndarray,
int dimension,
boolean ascending)
Sort an ndarray along a particular dimension.
Note that the input array is modified in-place. |
static INDArray |
Nd4j.specialConcat(int dimension,
INDArray... toConcat)
Concatneate ndarrays along a dimension
PLEASE NOTE: This method is special for GPU backend, it works on HOST side only.
|
INDArray |
NDArrayFactory.specialConcat(int dimension,
INDArray... toConcat)
Concatenate ndarrays along a dimension
PLEASE NOTE: This method is special for GPU backend, it works on HOST side only.
|
static INDArray |
Nd4j.squeeze(INDArray input,
int dimension)
Squeeze : removes a dimension of size 1
|
static INDArray |
Nd4j.stack(int axis,
INDArray... values)
Stack a set of N SDVariables of rank X into one rank X+1 variable.
|
static INDArray |
Nd4j.std(INDArray compute)
See
std(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.std(INDArray compute,
int dimension)
See
std(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.stripOnes(INDArray toStrip)
Reshapes an ndarray to remove leading 1s
|
static INDArray |
Broadcast.sub(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast subtraction op.
|
static INDArray |
Nd4j.sum(INDArray compute)
See
sum(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.sum(INDArray compute,
int dimension)
See
sum(int...)with Integer.MAX_VALUE for full array reduction. |
INDArray |
BaseBlasWrapper.swap(INDArray x,
INDArray y) |
INDArray |
BlasWrapper.swap(INDArray x,
INDArray y)
Compute x <-> y (swap two matrices)
|
INDArray |
BaseBlasWrapper.sysv(char uplo,
INDArray a,
int[] ipiv,
INDArray b) |
INDArray |
BlasWrapper.sysv(char uplo,
INDArray a,
int[] ipiv,
INDArray b) |
static INDArray[] |
Nd4j.tear(INDArray tensor,
int... dimensions)
This method does the opposite to pile/vstack/hstack - it returns independent TAD copies along given dimensions
|
INDArray[] |
NDArrayFactory.tear(INDArray tensor,
int... dimensions) |
static INDArray |
Nd4j.tensorMmul(INDArray a,
INDArray b,
INDArray result,
int[][] axes)
Tensor matrix multiplication.
|
static INDArray |
Nd4j.tensorMmul(INDArray a,
INDArray b,
int[][] axes)
Tensor matrix multiplication.
|
static INDArray |
Nd4j.tile(INDArray tile,
int... repeat)
An alias for repmat
|
static INDArray |
Nd4j.toFlattened(char order,
Collection<INDArray> matrices)
Create a long row vector of all of the given ndarrays
|
INDArray |
NDArrayFactory.toFlattened(char order,
Collection<INDArray> matrices)
Returns a flattened ndarray with all elements in each ndarray
regardless of dimension.
|
INDArray |
BaseNDArrayFactory.toFlattened(char order,
INDArray... matrices) |
static INDArray |
Nd4j.toFlattened(char order,
INDArray... matrices)
Create a long row vector of all of the given ndarrays/
|
INDArray |
NDArrayFactory.toFlattened(char order,
INDArray... matrices)
Flatten all of the ndarrays in to one long vector
|
INDArray |
BaseNDArrayFactory.toFlattened(Collection<INDArray> matrices)
Returns a vector with all of the elements in every nd array
equal to the sum of the lengths of the ndarrays
|
static INDArray |
Nd4j.toFlattened(Collection<INDArray> matrices)
Create a long row vector of all of the given ndarrays
|
INDArray |
NDArrayFactory.toFlattened(Collection<INDArray> matrices)
/**
Returns a flattened ndarray with all of the elements in each ndarray
regardless of dimension
|
INDArray |
BaseNDArrayFactory.toFlattened(INDArray... matrices) |
static INDArray |
Nd4j.toFlattened(INDArray... matrices)
Create a long row vector of all of the given ndarrays
|
INDArray |
NDArrayFactory.toFlattened(INDArray... matrices)
Flatten all of the ndarrays in to one long vector
|
INDArray |
BaseNDArrayFactory.toFlattened(int length,
Iterator<? extends INDArray>... matrices) |
INDArray |
NDArrayFactory.toFlattened(int length,
Iterator<? extends INDArray>... matrices)
Returns a flattened ndarray with all of the elements in each ndarray
regardless of dimension
|
static INDArray |
Nd4j.tri(int n)
See
Nd4j.tri(int,int,int) with m = n, k=0. |
static INDArray |
Nd4j.tri(int n,
int k)
See
Nd4j.tri(int,int,int) with m = n. |
static INDArray |
Nd4j.tri(int n,
int m,
int k)
Like the scipy function tri.
|
static INDArray |
Nd4j.triu(INDArray m,
int k)
Upper triangle of an array.
|
INDArray |
BaseNDArrayFactory.valueArrayOf(int[] shape,
double value)
Creates an ndarray with the specified value
as the only value in the ndarray
|
static INDArray |
Nd4j.valueArrayOf(int[] shape,
double value)
Creates an ndarray with the specified value
as the only value in the ndarray.
|
INDArray |
NDArrayFactory.valueArrayOf(int[] shape,
double value)
Creates an ndarray with the specified value
as the only value in the ndarray
|
INDArray |
BaseNDArrayFactory.valueArrayOf(long[] shape,
double value) |
static INDArray |
Nd4j.valueArrayOf(long[] shape,
double value)
|
INDArray |
NDArrayFactory.valueArrayOf(long[] shape,
double value) |
static INDArray |
Nd4j.valueArrayOf(long[] shape,
double value,
DataType type)
Creates an ndarray with the specified value
as the only value in the ndarray.
|
static INDArray |
Nd4j.valueArrayOf(long[] shape,
float value)
Creates an ndarray with the specified value as the only value in the FLOAT32 datatype NDArray.
|
static INDArray |
Nd4j.valueArrayOf(long[] shape,
int value)
Creates an ndarray with the specified value as the only value in the INTEGER datatype NDArray.
|
static INDArray |
Nd4j.valueArrayOf(long[] shape,
long value,
DataType type)
|
static INDArray |
Nd4j.valueArrayOf(long num,
double value)
Creates a row vector ndarray with the specified value
as the only value in the ndarray
Some people may know this as np.full
|
INDArray |
BaseNDArrayFactory.valueArrayOf(long rows,
long columns,
double value)
Creates a row vector with the specified number of columns
|
static INDArray |
Nd4j.valueArrayOf(long rows,
long columns,
double value)
Creates a row vector with the specified number of columns
Some people may know this as np.full
|
INDArray |
NDArrayFactory.valueArrayOf(long rows,
long columns,
double value)
Creates a row vector with the specified number of columns
|
static INDArray |
Nd4j.var(INDArray compute)
See
var(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.var(INDArray compute,
int dimension)
See
var(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.vstack(Collection<INDArray> arrs)
Concatenates two matrices vertically.
|
INDArray |
BaseNDArrayFactory.vstack(INDArray... arrs)
Concatenates two matrices vertically.
|
static INDArray |
Nd4j.vstack(INDArray... arrs)
Concatenates two matrices vertically.
|
INDArray |
NDArrayFactory.vstack(INDArray... arrs)
Concatenates two matrices vertically.
|
static INDArray[] |
Nd4j.where(INDArray condition,
INDArray x,
INDArray y)
Similar to numpy.where operation.
|
static INDArray |
Nd4j.zeros(DataType dataType,
long... shape)
Creates an array with the specified data tyoe and shape initialized with zero.
|
static INDArray |
Nd4j.zeros(int... shape)
Create an ndarray of zeros
|
INDArray |
BaseNDArrayFactory.zeros(int[] shape)
Create an ndarray of zeros
|
INDArray |
NDArrayFactory.zeros(int[] shape)
Create an ndarray of zeros
|
static INDArray |
Nd4j.zeros(int[] shape,
char order)
Create an ndarray of zeros
|
static INDArray |
Nd4j.zeros(int[] shape,
DataType dataType) |
static INDArray |
Nd4j.zeros(int[] shape,
int[] stride)
|
static INDArray |
Nd4j.zeros(int[] shape,
int[] stride,
long offset)
|
static INDArray |
Nd4j.zeros(int columns,
char order)
Create a 1D float array in soecified order initialized with zero.
|
static INDArray |
Nd4j.zeros(int rows,
int columns,
char ordering)
Create a 2D array initialized with zeros.
|
static INDArray |
Nd4j.zeros(int rows,
int columns,
int[] stride)
See
#create(int, int, int[], char) |
static INDArray |
Nd4j.zeros(int rows,
int columns,
int[] stride,
long offset)
|
static INDArray |
Nd4j.zeros(long... shape)
Create an ndarray of zeros
|
INDArray |
BaseNDArrayFactory.zeros(long columns)
Creates a row vector with the specified number of columns
|
INDArray |
NDArrayFactory.zeros(long columns)
Creates a row vector with the specified number of columns
|
INDArray |
BaseNDArrayFactory.zeros(long[] shape) |
INDArray |
NDArrayFactory.zeros(long[] shape) |
static INDArray |
Nd4j.zeros(long[] shape,
char order)
|
INDArray |
BaseNDArrayFactory.zeros(long rows,
long columns)
Creates a row vector with the specified number of columns
|
INDArray |
NDArrayFactory.zeros(long rows,
long columns)
Creates a row vector with the specified number of columns
|
static INDArray |
Nd4j.zerosLike(INDArray arr)
Empty like
|
| Modifier and Type | Method and Description |
|---|---|
static Map<String,INDArray> |
Nd4j.createFromNpzFile(File file) |
Map<String,INDArray> |
NDArrayFactory.createFromNpzFile(File file)
Create a Map
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Nd4j.accumulate(INDArray... arrays)
This method sums given arrays and stores them to a new array
|
static INDArray |
Nd4j.accumulate(INDArray target,
Collection<INDArray> arrays)
This method sums given arrays and stores them to a given target array
|
INDArray |
NDArrayFactory.accumulate(INDArray target,
INDArray... arrays)
This method sums given arrays to target
|
INDArray |
NDArrayFactory.accumulate(INDArray target,
INDArray... arrays)
This method sums given arrays to target
|
static INDArray |
Nd4j.accumulate(INDArray target,
INDArray[] arrays)
This method sums given arrays and stores them to a given target array
|
static INDArray |
Nd4j.accumulate(INDArray target,
INDArray[] arrays)
This method sums given arrays and stores them to a given target array
|
static INDArray |
Broadcast.add(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast add op.
|
static INDArray |
Broadcast.amax(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast absolute max op.
|
static INDArray |
Broadcast.amin(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast absolute min op.
|
static INDArray |
Nd4j.append(INDArray arr,
int padAmount,
double val,
int axis)
Append the given array with the specified value size along a particular axis.
|
INDArray |
BaseNDArrayFactory.appendBias(INDArray... vectors)
Merge the vectors and append a bias.
|
static INDArray |
Nd4j.appendBias(INDArray... vectors) |
INDArray |
NDArrayFactory.appendBias(INDArray... vectors) |
static INDArray |
Nd4j.argMax(INDArray arr,
int... dimension)
Get the maximum values for a dimension.
|
static INDArray |
Nd4j.argMin(INDArray arr,
int... dimension)
See
Nd4j.argMax(INDArray, int...) but return minimum values. |
static void |
DataTypeValidation.assertDouble(INDArray... d) |
static void |
DataTypeValidation.assertDouble(INDArray d) |
static void |
DataTypeValidation.assertFloat(INDArray... d2) |
static void |
DataTypeValidation.assertFloat(INDArray d2) |
static void |
DataTypeValidation.assertSameDataType(INDArray... indArrays) |
double |
BaseBlasWrapper.asum(INDArray x) |
double |
BlasWrapper.asum(INDArray x)
Compute || x ||_1 (1-norm, sum of absolute values)
|
INDArray |
NDArrayFactory.average(INDArray[] arrays)
This method averages input arrays, and returns averaged array
|
INDArray |
NDArrayFactory.average(INDArray target,
Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array
|
INDArray |
NDArrayFactory.average(INDArray target,
INDArray[] arrays)
This method averages input arrays, and returns averaged array
|
INDArray |
NDArrayFactory.average(INDArray target,
INDArray[] arrays)
This method averages input arrays, and returns averaged array
|
static INDArray |
Nd4j.averageAndPropagate(INDArray[] arrays)
This method averages input arrays, and returns averaged array.
|
static INDArray |
Nd4j.averageAndPropagate(INDArray target,
Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array.
|
INDArray |
BaseBlasWrapper.axpy(double da,
INDArray dx,
INDArray dy) |
INDArray |
BlasWrapper.axpy(double da,
INDArray dx,
INDArray dy)
Deprecated.
|
INDArray |
BaseBlasWrapper.axpy(float da,
INDArray dx,
INDArray dy) |
INDArray |
BlasWrapper.axpy(float da,
INDArray dx,
INDArray dy)
Deprecated.
|
INDArray |
BaseBlasWrapper.axpy(Number da,
INDArray dx,
INDArray dy) |
INDArray |
BlasWrapper.axpy(Number da,
INDArray dx,
INDArray dy)
Compute y <- y + x * alpha
|
INDArray |
BaseNDArrayFactory.bilinearProducts(INDArray curr,
INDArray in)
Returns a column vector where each entry is the nth bilinear
product of the nth slices of the two tensors.
|
INDArray |
NDArrayFactory.bilinearProducts(INDArray curr,
INDArray in)
Returns a column vector where each entry is the nth bilinear
product of the nth slices of the two tensors.
|
static INDArray |
Nd4j.choice(INDArray source,
INDArray probs,
INDArray target)
This method samples value from Source array to Target,the default random number generator.
|
static INDArray |
Nd4j.choice(@NonNull INDArray source,
@NonNull INDArray probs,
@NonNull INDArray target,
Random rng)
This method samples value from Source array to Target, with probabilites provided in Probs argument
|
static INDArray |
Nd4j.choice(@NonNull INDArray source,
@NonNull INDArray probs,
@NonNull INDArray target,
Random rng)
This method samples value from Source array to Target, with probabilites provided in Probs argument
|
static INDArray |
Nd4j.choice(@NonNull INDArray source,
@NonNull INDArray probs,
@NonNull INDArray target,
Random rng)
This method samples value from Source array to Target, with probabilites provided in Probs argument
|
static INDArray |
Nd4j.choice(INDArray source,
INDArray probs,
int numSamples)
This method returns new INDArray instance, sampled from Source array with probabilities given in Probs
using the default random number generator.
|
static INDArray |
Nd4j.choice(INDArray source,
INDArray probs,
int numSamples,
Random rng)
This method returns new INDArray instance, sampled from Source array with probabilities given in Probs.
|
static void |
Nd4j.clearNans(INDArray arr)
Clear nans from an ndarray
|
INDArray |
BaseNDArrayFactory.concat(int dimension,
INDArray... toConcat)
concatenate ndarrays along a dimension
|
static INDArray |
Nd4j.concat(int dimension,
INDArray... toConcat)
Concatneate ndarrays along a dimension
|
INDArray |
NDArrayFactory.concat(int dimension,
INDArray... toConcat)
Concatneate ndarrays along a dimension
|
INDArray |
NDArrayFactory.convertDataEx(DataTypeEx typeSrc,
INDArray source,
DataTypeEx typeDst) |
static org.bytedeco.javacpp.Pointer |
Nd4j.convertToNumpy(INDArray arr)
Converts an
INDArray to a numpy struct. |
org.bytedeco.javacpp.Pointer |
NDArrayFactory.convertToNumpy(INDArray array)
Convert an
INDArray
to a numpy array. |
void |
BaseNDArrayFactory.copy(INDArray a,
INDArray b)
Copy a to b
|
static void |
Nd4j.copy(INDArray a,
INDArray b)
Copy a to b
|
INDArray |
BaseBlasWrapper.copy(INDArray x,
INDArray y) |
INDArray |
BlasWrapper.copy(INDArray x,
INDArray y)
Compute y <- x (copy a matrix)
|
void |
NDArrayFactory.copy(INDArray a,
INDArray b)
Copy a to b
|
static INDArray |
Broadcast.copy(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast copy op.
|
static INDArray |
Nd4j.cumsum(INDArray compute)
See
cumsum(int) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.cumsum(INDArray compute,
int dimension)
See
cumsum(int) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.diag(INDArray x)
Creates a new matrix where the values of the given vector are the diagonal values of
the matrix if a vector is passed in, if a matrix is returns the kth diagonal
in the matrix
|
static INDArray |
Broadcast.div(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast divide op.
|
double |
BaseBlasWrapper.dot(INDArray x,
INDArray y) |
double |
BlasWrapper.dot(INDArray x,
INDArray y)
Compute x^T * y (dot product)
|
static INDArray |
Broadcast.eq(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast equal to op.
|
static INDArray |
Nd4j.expandDims(INDArray input,
int dimension)
Expand the array dimensions.
|
int |
BaseBlasWrapper.geev(char jobvl,
char jobvr,
INDArray A,
INDArray WR,
INDArray WI,
INDArray VL,
INDArray VR) |
int |
BlasWrapper.geev(char jobvl,
char jobvr,
INDArray A,
INDArray WR,
INDArray WI,
INDArray VL,
INDArray VR) |
void |
BaseBlasWrapper.gelsd(INDArray A,
INDArray B) |
void |
BlasWrapper.gelsd(INDArray A,
INDArray B)
Generalized Least Squares via *GELSD.
|
static INDArray |
Nd4j.gemm(INDArray a,
INDArray b,
boolean transposeA,
boolean transposeB)
matrix multiply: implements op(a)*op(b)
where op(x) means transpose x (or not) depending on
setting of arguments transposea and transposeb.
so gemm(a,b,false,false) == a.mmul(b), gemm(a,b,true,false) == a.transpose().mmul(b) etc. |
static INDArray |
Nd4j.gemm(INDArray a,
INDArray b,
INDArray c,
boolean transposeA,
boolean transposeB,
double alpha,
double beta)
Matrix multiply: Implements c = alpha*op(a)*op(b) + beta*c where op(X) means transpose X (or not)
depending on setting of arguments transposeA and transposeB.
Note that matrix c MUST be fortran order, have zero offset and have c.data().length == c.length(). |
INDArray |
BaseBlasWrapper.gemv(double alpha,
INDArray a,
INDArray x,
double beta,
INDArray y) |
INDArray |
BlasWrapper.gemv(double alpha,
INDArray a,
INDArray x,
double beta,
INDArray y)
Deprecated.
|
INDArray |
BaseBlasWrapper.gemv(float alpha,
INDArray a,
INDArray x,
float beta,
INDArray y) |
INDArray |
BlasWrapper.gemv(float alpha,
INDArray a,
INDArray x,
float beta,
INDArray y)
Deprecated.
|
INDArray |
BaseBlasWrapper.gemv(Number alpha,
INDArray a,
INDArray x,
double beta,
INDArray y) |
INDArray |
BlasWrapper.gemv(Number alpha,
INDArray a,
INDArray x,
double beta,
INDArray y)
************************************************************************
BLAS Level 2
|
void |
BaseBlasWrapper.geqrf(INDArray A,
INDArray tau) |
void |
BlasWrapper.geqrf(INDArray A,
INDArray tau) |
INDArray |
BaseBlasWrapper.ger(double alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BlasWrapper.ger(double alpha,
INDArray x,
INDArray y,
INDArray a)
Deprecated.
|
INDArray |
BaseBlasWrapper.ger(float alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BlasWrapper.ger(float alpha,
INDArray x,
INDArray y,
INDArray a)
Compute A <- alpha * x * y^T + A (general rank-1 update)
|
INDArray |
BaseBlasWrapper.ger(Number alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BlasWrapper.ger(Number alpha,
INDArray x,
INDArray y,
INDArray a) |
INDArray |
BaseBlasWrapper.gesv(INDArray a,
int[] ipiv,
INDArray b) |
INDArray |
BlasWrapper.gesv(INDArray a,
int[] ipiv,
INDArray b)
************************************************************************
LAPACK
|
static INDArray |
Broadcast.gt(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast greater than op.
|
static INDArray |
Broadcast.gte(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast greater than or equal to op.
|
INDArray |
BaseNDArrayFactory.hstack(INDArray... arrs)
Concatenates two matrices horizontally.
|
static INDArray |
Nd4j.hstack(INDArray... arrs)
Concatenates two matrices horizontally.
|
INDArray |
NDArrayFactory.hstack(INDArray... arrs)
Concatenates two matrices horizontally.
|
int |
BaseBlasWrapper.iamax(INDArray x) |
int |
BlasWrapper.iamax(INDArray x)
Compute index of element with largest absolute value (index of absolute
value maximum)
|
static boolean |
NDValidation.isSameType(INDArray[] x) |
static boolean |
NDValidation.isSameType(INDArray x,
INDArray y) |
static INDArray |
Broadcast.lt(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast less than op.
|
static INDArray |
Broadcast.lte(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast less than or equal to op.
|
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b)
Matrix multiplication/dot product
See
Nd4j.matmul(INDArray, INDArray, INDArray, boolean, boolean, boolean) |
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b,
boolean transposeA,
boolean transposeB,
boolean transposeResult)
Matrix multiplication/dot product.
See Nd4j.matmul(INDArray, INDArray, INDArray, boolean, boolean, boolean) |
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b,
INDArray result)
Matrix multiplication/dot product.
See Nd4j.matmul(INDArray, INDArray, INDArray, boolean, boolean, boolean) |
static INDArray |
Nd4j.matmul(INDArray a,
INDArray b,
INDArray result,
boolean transposeA,
boolean transposeB,
boolean transposeResult)
Matrix multiplication/dot product
Depending on inputs dimensionality output result might be different.
|
static INDArray |
Nd4j.max(INDArray compute)
See
max(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Broadcast.max(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast max op.
|
static INDArray |
Nd4j.max(INDArray compute,
int dimension)
See
max(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.mean(INDArray compute)
See
mean(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.mean(INDArray compute,
int dimension)
See
mean(int...)with Integer.MAX_VALUE for full array reduction. |
static INDArray[] |
Nd4j.meshgrid(@NonNull INDArray x,
@NonNull INDArray y)
Meshgrid op.
|
static INDArray[] |
Nd4j.meshgrid(@NonNull INDArray x,
@NonNull INDArray y)
Meshgrid op.
|
static INDArray |
Nd4j.min(INDArray compute)
See
min(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Broadcast.min(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast min op.
|
static INDArray |
Nd4j.min(INDArray compute,
int dimension)
See
min(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Broadcast.mul(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast element-wise multiply op.
|
static INDArray |
Broadcast.neq(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast not equal to op.
|
static INDArray |
Nd4j.norm1(INDArray compute)
See
norm1(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.norm1(INDArray compute,
int dimension)
See
norm1(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.norm2(INDArray compute)
See
norm2(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.norm2(INDArray compute,
int dimension)
See
norm2(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.normmax(INDArray compute)
See
normmax(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.normmax(INDArray compute,
int dimension)
See
normmax(int...) with Integer.MAX_VALUE for full array reduction. |
double |
BaseBlasWrapper.nrm2(INDArray x) |
double |
BlasWrapper.nrm2(INDArray x)
Compute || x ||_2 (2-norm)
|
static INDArray |
Nd4j.onesLike(INDArray arr)
Ones like
|
void |
BaseBlasWrapper.ormqr(char side,
char trans,
INDArray A,
INDArray tau,
INDArray C) |
void |
BlasWrapper.ormqr(char side,
char trans,
INDArray A,
INDArray tau,
INDArray C) |
static INDArray |
Nd4j.pad(INDArray toPad,
INDArray padding)
See
#pad(INDArray, INDArray, Mode, double) with zero padding (zeros for padValue). |
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull INDArray padWidth,
Pad.Mode padMode,
double padValue)
Pad the given ndarray to the size along each dimension.
|
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull INDArray padWidth,
Pad.Mode padMode,
double padValue)
Pad the given ndarray to the size along each dimension.
|
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
int... padWidth)
|
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull int[][] padWidth)
|
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull int[][] padWidth,
Pad.Mode padMode,
double padValue)
See
#pad(INDArray, INDArray, Mode, double). |
static INDArray |
Nd4j.pad(@NonNull INDArray toPad,
@NonNull int[] padWidth,
Pad.Mode padMode,
double padValue)
See
#pad(INDArray, INDArray, Mode, double), uses padWidth for all dimensions. |
static INDArray |
Nd4j.pile(INDArray... arrays)
This method stacks vertically examples with the same shape, increasing result dimensionality.
|
void |
BaseBlasWrapper.posv(char uplo,
INDArray A,
INDArray B) |
void |
BlasWrapper.posv(char uplo,
INDArray A,
INDArray B) |
static INDArray |
Nd4j.prepend(INDArray arr,
int padAmount,
double val,
int axis)
|
static INDArray |
Nd4j.prod(INDArray compute)
See
prod(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.prod(INDArray compute,
int dimension)
See
prod(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.pullRows(INDArray source,
INDArray destination,
int sourceDimension,
int... indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some
dimension and specified indexes.
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
INDArray destination,
int sourceDimension,
int[] indexes)
* This method produces concatenated array, that consist from tensors,
fetched from source array, against some dimension and specified indexes
in to the destination array
|
static INDArray |
Nd4j.pullRows(INDArray source,
int sourceDimension,
int... indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
INDArray |
BaseNDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
INDArray |
BaseNDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes,
char order)
This method produces concatenated array, that consist from tensors, fetched from source array, against some dimension and specified indexes
|
static INDArray |
Nd4j.pullRows(INDArray source,
int sourceDimension,
int[] indexes,
char order)
This method produces concatenated array,
that consist from tensors,
fetched from source array,
against some dimension and specified indexes
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
int sourceDimension,
int[] indexes,
char order)
This method produces concatenated array, that consist from tensors,
fetched from source array, against some dimension and specified indexes
|
INDArray |
NDArrayFactory.pullRows(INDArray source,
int sourceDimension,
long[] indexes) |
static INDArray |
Nd4j.rand(INDArray target)
Fill the given ndarray with random numbers drawn from a uniform distribution
|
static INDArray |
Nd4j.rand(INDArray target,
@NonNull Distribution dist)
Fill the given ndarray with random numbers drawn from the given distribution
|
static INDArray |
Nd4j.rand(INDArray target,
double min,
double max,
Random rng)
Fill the given ndarray with random numbers drawn from a uniform distribution using the given RandomGenerator
|
static INDArray |
Nd4j.rand(INDArray target,
long seed)
Fill the given ndarray with random numbers drawn from a uniform distribution
|
static INDArray |
Nd4j.rand(INDArray target,
Random rng)
Fill the given ndarray with random numbers drawn from a uniform distribution using the given RandomGenerator
|
static INDArray |
Nd4j.randn(double mean,
double stddev,
INDArray target,
Random rng) |
static INDArray |
Nd4j.randn(INDArray target)
Fill the given ndarray with random numbers drawn from a normal distribution
|
static INDArray |
Nd4j.randn(INDArray target,
long seed)
Fill the given ndarray with random numbers drawn from a normal distribution
|
static INDArray |
Nd4j.randn(INDArray target,
Random rng)
Fill the given ndarray with random numbers drawn from a normal distribution utilizing the given random generator
|
static INDArray |
Nd4j.randomBernoulli(double p,
@NonNull INDArray target)
Fill the specified array with values generated according to a binomial distribution with probability p: i.e.,
values 0 with probability (1-p) or value 1 with probability p
|
static INDArray |
Nd4j.randomBinomial(int nTrials,
double p,
INDArray target)
Fill the target array with random values generated according to a binomial distribution with the specified
number of trials and probability
|
static INDArray |
Nd4j.randomExponential(double lambda,
INDArray target)
Exponential distribution: P(x) = lambda * exp(-lambda * x)
|
static INDArray |
Broadcast.rdiv(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast reverse division op.
|
static INDArray |
Nd4j.repeat(INDArray n,
int num)
Create an n x (shape)
ndarray where the ndarray is repeated num times
|
INDArray |
BaseNDArrayFactory.reverse(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
static INDArray |
Nd4j.reverse(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
INDArray |
NDArrayFactory.reverse(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
static INDArray |
Nd4j.rollAxis(INDArray a,
int axis)
Roll the specified axis backwards,
until it lies in a given position.
|
static INDArray |
Nd4j.rollAxis(INDArray a,
int axis,
int start)
Roll the specified axis backwards,
until it lies in a given position.
|
INDArray |
BaseNDArrayFactory.rot(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
INDArray |
NDArrayFactory.rot(INDArray reverse)
Reverses the passed in matrix such that m[0] becomes m[m.length - 1] etc
|
void |
BaseNDArrayFactory.rot90(INDArray toRotate)
Rotate a matrix 90 degrees
|
static void |
Nd4j.rot90(INDArray toRotate)
Rotate a matrix 90 degrees
|
void |
NDArrayFactory.rot90(INDArray toRotate)
Rotate a matrix 90 degrees
|
static INDArray |
Broadcast.rsub(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast reverse subtraction op.
|
static void |
Nd4j.saveBinary(INDArray arr,
File saveTo)
Save an ndarray to the given file
|
void |
BaseBlasWrapper.saxpy(double alpha,
INDArray x,
INDArray y) |
void |
BlasWrapper.saxpy(double alpha,
INDArray x,
INDArray y)
Deprecated.
|
void |
BaseBlasWrapper.saxpy(float alpha,
INDArray x,
INDArray y) |
void |
BlasWrapper.saxpy(float alpha,
INDArray x,
INDArray y)
Deprecated.
|
INDArray |
BaseBlasWrapper.scal(double alpha,
INDArray x) |
INDArray |
BlasWrapper.scal(double alpha,
INDArray x)
Deprecated.
|
INDArray |
BaseBlasWrapper.scal(float alpha,
INDArray x) |
INDArray |
BlasWrapper.scal(float alpha,
INDArray x)
Deprecated.
|
static void |
Nd4j.scatterUpdate(ScatterUpdate.UpdateOp op,
@NonNull INDArray array,
@NonNull INDArray indices,
@NonNull INDArray updates,
int... axis)
Deprecated.
|
static void |
Nd4j.scatterUpdate(ScatterUpdate.UpdateOp op,
@NonNull INDArray array,
@NonNull INDArray indices,
@NonNull INDArray updates,
int... axis)
Deprecated.
|
static void |
Nd4j.scatterUpdate(ScatterUpdate.UpdateOp op,
@NonNull INDArray array,
@NonNull INDArray indices,
@NonNull INDArray updates,
int... axis)
Deprecated.
|
static long[] |
Nd4j.shape(INDArray arr)
Returns the shape of the ndarray
|
static void |
Nd4j.shuffle(INDArray toShuffle,
int... dimension)
In place shuffle of an ndarray
along a specified set of dimensions
|
static void |
Nd4j.shuffle(INDArray toShuffle,
Random random,
int... dimension)
In place shuffle of an ndarray
along a specified set of dimensions
|
void |
NDArrayFactory.shuffle(INDArray array,
Random rnd,
int... dimension)
In place shuffle of an ndarray
along a specified set of dimensions
|
static INDArray |
Nd4j.sort(INDArray ndarray,
boolean ascending)
Sort all elements of an array.
|
INDArray |
NDArrayFactory.sort(INDArray x,
boolean descending) |
INDArray |
NDArrayFactory.sort(INDArray x,
boolean descending,
int... dimensions) |
static INDArray |
Nd4j.sort(INDArray ndarray,
int dimension,
boolean ascending)
Sort an ndarray along a particular dimension
Note that the input array is modified in-place. |
static INDArray |
Nd4j.sortColumns(INDArray in,
int rowIdx,
boolean ascending)
Sort (shuffle) the columns of a 2d array according to the value at a specified row.
|
INDArray |
NDArrayFactory.sortCooIndices(INDArray x) |
static INDArray |
Nd4j.sortRows(INDArray in,
int colIdx,
boolean ascending)
Sort (shuffle) the rows of a 2d array according to the value at a specified column.
|
static INDArray[] |
Nd4j.sortWithIndices(INDArray ndarray,
int dimension,
boolean ascending)
Sort an ndarray along a particular dimension.
Note that the input array is modified in-place. |
static INDArray |
Nd4j.specialConcat(int dimension,
INDArray... toConcat)
Concatneate ndarrays along a dimension
PLEASE NOTE: This method is special for GPU backend, it works on HOST side only.
|
INDArray |
NDArrayFactory.specialConcat(int dimension,
INDArray... toConcat)
Concatenate ndarrays along a dimension
PLEASE NOTE: This method is special for GPU backend, it works on HOST side only.
|
static INDArray |
Nd4j.squeeze(INDArray input,
int dimension)
Squeeze : removes a dimension of size 1
|
static INDArray |
Nd4j.stack(int axis,
INDArray... values)
Stack a set of N SDVariables of rank X into one rank X+1 variable.
|
static INDArray |
Nd4j.std(INDArray compute)
See
std(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.std(INDArray compute,
int dimension)
See
std(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.stripOnes(INDArray toStrip)
Reshapes an ndarray to remove leading 1s
|
static INDArray |
Broadcast.sub(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Broadcast subtraction op.
|
static INDArray |
Nd4j.sum(INDArray compute)
See
sum(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.sum(INDArray compute,
int dimension)
See
sum(int...)with Integer.MAX_VALUE for full array reduction. |
INDArray |
BaseBlasWrapper.swap(INDArray x,
INDArray y) |
INDArray |
BlasWrapper.swap(INDArray x,
INDArray y)
Compute x <-> y (swap two matrices)
|
int |
BaseBlasWrapper.syev(char jobz,
char uplo,
INDArray a,
INDArray w) |
int |
BlasWrapper.syev(char jobz,
char uplo,
INDArray a,
INDArray w) |
int |
BaseBlasWrapper.syevd(char jobz,
char uplo,
INDArray A,
INDArray w) |
int |
BlasWrapper.syevd(char jobz,
char uplo,
INDArray A,
INDArray w) |
int |
BaseBlasWrapper.syevr(char jobz,
char range,
char uplo,
INDArray a,
double vl,
double vu,
int il,
int iu,
double abstol,
INDArray w,
INDArray z,
int[] isuppz) |
int |
BlasWrapper.syevr(char jobz,
char range,
char uplo,
INDArray a,
double vl,
double vu,
int il,
int iu,
double abstol,
INDArray w,
INDArray z,
int[] isuppz)
Deprecated.
|
int |
BaseBlasWrapper.syevr(char jobz,
char range,
char uplo,
INDArray a,
float vl,
float vu,
int il,
int iu,
float abstol,
INDArray w,
INDArray z,
int[] isuppz) |
int |
BlasWrapper.syevr(char jobz,
char range,
char uplo,
INDArray a,
float vl,
float vu,
int il,
int iu,
float abstol,
INDArray w,
INDArray z,
int[] isuppz)
Deprecated.
|
int |
BaseBlasWrapper.syevr(char jobz,
char range,
char uplo,
INDArray a,
float vl,
float vu,
int il,
int iu,
Number abstol,
INDArray w,
INDArray z,
int[] isuppz) |
int |
BlasWrapper.syevr(char jobz,
char range,
char uplo,
INDArray a,
float vl,
float vu,
int il,
int iu,
Number abstol,
INDArray w,
INDArray z,
int[] isuppz) |
int |
BaseBlasWrapper.syevx(char jobz,
char range,
char uplo,
INDArray a,
double vl,
double vu,
int il,
int iu,
double abstol,
INDArray w,
INDArray z) |
int |
BlasWrapper.syevx(char jobz,
char range,
char uplo,
INDArray a,
double vl,
double vu,
int il,
int iu,
double abstol,
INDArray w,
INDArray z) |
int |
BaseBlasWrapper.syevx(char jobz,
char range,
char uplo,
INDArray a,
float vl,
float vu,
int il,
int iu,
float abstol,
INDArray w,
INDArray z) |
int |
BlasWrapper.syevx(char jobz,
char range,
char uplo,
INDArray a,
float vl,
float vu,
int il,
int iu,
float abstol,
INDArray w,
INDArray z) |
int |
BaseBlasWrapper.sygvd(int itype,
char jobz,
char uplo,
INDArray A,
INDArray B,
INDArray W) |
int |
BlasWrapper.sygvd(int itype,
char jobz,
char uplo,
INDArray A,
INDArray B,
INDArray W) |
INDArray |
BaseBlasWrapper.sysv(char uplo,
INDArray a,
int[] ipiv,
INDArray b) |
INDArray |
BlasWrapper.sysv(char uplo,
INDArray a,
int[] ipiv,
INDArray b) |
static INDArray[] |
Nd4j.tear(INDArray tensor,
int... dimensions)
This method does the opposite to pile/vstack/hstack - it returns independent TAD copies along given dimensions
|
INDArray[] |
NDArrayFactory.tear(INDArray tensor,
int... dimensions) |
static INDArray |
Nd4j.tensorMmul(INDArray a,
INDArray b,
INDArray result,
int[][] axes)
Tensor matrix multiplication.
|
static INDArray |
Nd4j.tensorMmul(INDArray a,
INDArray b,
int[][] axes)
Tensor matrix multiplication.
|
static INDArray |
Nd4j.tile(INDArray tile,
int... repeat)
An alias for repmat
|
static byte[] |
Nd4j.toByteArray(@NonNull INDArray arr)
Convert an ndarray to a byte array
|
INDArray |
BaseNDArrayFactory.toFlattened(char order,
INDArray... matrices) |
static INDArray |
Nd4j.toFlattened(char order,
INDArray... matrices)
Create a long row vector of all of the given ndarrays/
|
INDArray |
NDArrayFactory.toFlattened(char order,
INDArray... matrices)
Flatten all of the ndarrays in to one long vector
|
INDArray |
BaseNDArrayFactory.toFlattened(INDArray... matrices) |
static INDArray |
Nd4j.toFlattened(INDArray... matrices)
Create a long row vector of all of the given ndarrays
|
INDArray |
NDArrayFactory.toFlattened(INDArray... matrices)
Flatten all of the ndarrays in to one long vector
|
static byte[] |
Nd4j.toNpyByteArray(INDArray input)
Converts an
INDArray to a byte array |
static INDArray |
Nd4j.triu(INDArray m,
int k)
Upper triangle of an array.
|
static void |
NDValidation.validateBool(String opName,
INDArray v)
Validate that the operation is being applied on a boolean type INDArray
|
static void |
NDValidation.validateBool(String opName,
INDArray v1,
INDArray v2)
Validate that the operation is being applied on boolean INDArrays
|
static void |
NDValidation.validateBool(String opName,
String inputName,
INDArray v)
Validate that the operation is being applied on a boolean type INDArray
|
static void |
Broadcast.validateBroadcastDims(INDArray x,
INDArray y,
INDArray z,
int... dimensions)
Validate the broadcast dimensions for manual broadcast ops such as
BroadcastMulOp. |
static void |
NDValidation.validateFloatingPoint(String opName,
INDArray v)
Validate that the operation is being applied on an floating point type INDArray
|
static void |
NDValidation.validateFloatingPoint(String opName,
String inputName,
INDArray v)
Validate that the operation is being applied on a floating point type INDArray
|
static void |
NDValidation.validateInteger(String opName,
INDArray v)
Validate that the operation is being applied on an integer type INDArray
|
static void |
NDValidation.validateInteger(String opName,
String inputName,
INDArray v)
Validate that the operation is being applied on an integer type INDArray
|
static void |
NDValidation.validateInteger(String opName,
String inputName,
INDArray[] v)
Validate that the operation is being applied on an integer type INDArray []
|
static void |
NDValidation.validateNumerical(String opName,
INDArray v)
Validate that the operation is being applied on a numerical INDArray (not boolean or utf8).
|
static void |
NDValidation.validateNumerical(String opName,
INDArray[] v)
Validate that the operation is being applied on numerical INDArrays (not boolean or utf8).
|
static void |
NDValidation.validateNumerical(String opName,
INDArray v1,
INDArray v2)
Validate that the operation is being applied on numerical INDArrays (not boolean or utf8).
|
static void |
NDValidation.validateNumerical(String opName,
String inputName,
INDArray v)
Validate that the operation is being applied on a numerical INDArray (not boolean or utf8).
|
static void |
NDValidation.validateNumerical(String opName,
String inputName,
INDArray[] v)
Validate that the operation is being applied on numerical INDArrays (not boolean or utf8).
|
static void |
NDValidation.validateSameType(String opName,
boolean numericalOnly,
INDArray... vars)
Validate that the operation is being applied on array with the exact same datatypes (which may optionally be
restricted to numerical INDArrays only (not boolean or utf8))
|
static void |
Broadcast.validateShapesNoDimCase(INDArray x,
INDArray y,
INDArray z) |
static INDArray |
Nd4j.var(INDArray compute)
See
var(int...) with Integer.MAX_VALUE for full array reduction. |
static INDArray |
Nd4j.var(INDArray compute,
int dimension)
See
var(int...) with Integer.MAX_VALUE for full array reduction. |
INDArray |
BaseNDArrayFactory.vstack(INDArray... arrs)
Concatenates two matrices vertically.
|
static INDArray |
Nd4j.vstack(INDArray... arrs)
Concatenates two matrices vertically.
|
INDArray |
NDArrayFactory.vstack(INDArray... arrs)
Concatenates two matrices vertically.
|
static INDArray[] |
Nd4j.where(INDArray condition,
INDArray x,
INDArray y)
Similar to numpy.where operation.
|
static void |
Nd4j.write(INDArray arr,
DataOutputStream dataOutputStream)
Write an ndarray to the specified outputstream
|
static void |
Nd4j.write(OutputStream writer,
INDArray write)
Y
Write an ndarray to a writer
|
static void |
Nd4j.writeAsNumpy(INDArray arr,
File file)
|
static void |
Nd4j.writeAsNumpy(INDArray arr,
OutputStream writeTo)
Writes an array to an output stream
|
static void |
Nd4j.writeTxt(INDArray write,
String filePath)
Write NDArray to a text file
|
static void |
Nd4j.writeTxt(INDArray write,
String filePath,
int precision)
Deprecated.
Defaults to scientific notation with 18 digits after the decimal
Use
Nd4j.writeTxt(INDArray, String) |
static void |
Nd4j.writeTxt(INDArray write,
String filePath,
String split)
Deprecated.
custom col and higher dimension separators are no longer supported; uses ","
Use
Nd4j.writeTxt(INDArray, String) |
static void |
Nd4j.writeTxt(INDArray write,
String filePath,
String split,
int precision)
Deprecated.
Precision is no longer used. Split is no longer used.
Defaults to scientific notation with 18 digits after the decimal
Use
Nd4j.writeTxt(INDArray, String) |
static INDArray |
Nd4j.zerosLike(INDArray arr)
Empty like
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Nd4j.accumulate(INDArray target,
Collection<INDArray> arrays)
This method sums given arrays and stores them to a given target array
|
INDArray |
NDArrayFactory.average(Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array
|
INDArray |
NDArrayFactory.average(INDArray target,
Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array
|
static INDArray |
Nd4j.averageAndPropagate(Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array.
|
static INDArray |
Nd4j.averageAndPropagate(INDArray target,
Collection<INDArray> arrays)
This method averages input arrays, and returns averaged array.
|
static INDArray |
Nd4j.create(List<INDArray> list,
int... shape)
Creates an array with the specified shape from a list of arrays.
|
abstract INDArray |
BaseNDArrayFactory.create(List<INDArray> list,
int[] shape)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(List<INDArray> list,
int[] shape)
Creates an ndarray with the specified shape
|
INDArray |
NDArrayFactory.create(List<INDArray> list,
int[] shape,
char ordering) |
static INDArray |
Nd4j.create(List<INDArray> list,
long... shape)
|
INDArray |
NDArrayFactory.create(List<INDArray> list,
long[] shape) |
INDArray |
NDArrayFactory.create(List<INDArray> list,
long[] shape,
char ordering) |
static INDArray |
Nd4j.hstack(Collection<INDArray> arrs)
Concatenates two matrices horizontally.
|
static void |
Nd4j.shuffle(Collection<INDArray> toShuffle,
int... dimension)
Symmetric in place shuffle of an ndarray
along a specified set of dimensions
|
static void |
Nd4j.shuffle(Collection<INDArray> toShuffle,
Random rnd,
int... dimension)
Symmetric in place shuffle of an ndarray
along a specified set of dimensions
|
void |
NDArrayFactory.shuffle(Collection<INDArray> array,
Random rnd,
int... dimension)
Symmetric in place shuffle of an ndarray
along a specified set of dimensions.
|
static void |
Nd4j.shuffle(List<INDArray> toShuffle,
Random rnd,
List<int[]> dimensions)
Symmetric in place shuffle of an ndarray
along a variable dimensions
|
void |
NDArrayFactory.shuffle(List<INDArray> array,
Random rnd,
List<int[]> dimensions)
Symmetric in place shuffle of an ndarray
along a specified set of dimensions.
|
static INDArray |
Nd4j.toFlattened(char order,
Collection<INDArray> matrices)
Create a long row vector of all of the given ndarrays
|
INDArray |
NDArrayFactory.toFlattened(char order,
Collection<INDArray> matrices)
Returns a flattened ndarray with all elements in each ndarray
regardless of dimension.
|
INDArray |
BaseNDArrayFactory.toFlattened(Collection<INDArray> matrices)
Returns a vector with all of the elements in every nd array
equal to the sum of the lengths of the ndarrays
|
static INDArray |
Nd4j.toFlattened(Collection<INDArray> matrices)
Create a long row vector of all of the given ndarrays
|
INDArray |
NDArrayFactory.toFlattened(Collection<INDArray> matrices)
/**
Returns a flattened ndarray with all of the elements in each ndarray
regardless of dimension
|
static INDArray |
Nd4j.vstack(Collection<INDArray> arrs)
Concatenates two matrices vertically.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
NDMath.abs(INDArray x)
Elementwise absolute value operation: out = abs(x)
|
INDArray |
NDLoss.absoluteDifference(INDArray label,
INDArray predictions,
INDArray weights)
Absolute difference loss:
sum_i abs( label[i] - predictions[i] ) |
INDArray |
NDLoss.absoluteDifference(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Absolute difference loss:
sum_i abs( label[i] - predictions[i] ) |
INDArray |
NDMath.acos(INDArray x)
Elementwise acos (arccosine, inverse cosine) operation: out = arccos(x)
|
INDArray |
NDMath.acosh(INDArray x)
Elementwise acosh (inverse hyperbolic cosine) function: out = acosh(x)
|
INDArray |
NDMath.add(INDArray x,
double value)
Scalar add operation, out = in + scalar
|
INDArray |
NDMath.add(INDArray x,
INDArray y)
Pairwise addition operation, out = x + y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDImage.adjustContrast(INDArray in,
double factor)
Adjusts contrast of RGB or grayscale images.
|
INDArray |
NDImage.adjustHue(INDArray in,
double delta)
Adjust hue of RGB image
|
INDArray |
NDImage.adjustSaturation(INDArray in,
double factor)
Adjust saturation of RGB images
|
INDArray |
NDBase.all(INDArray x,
int... dimensions)
Boolean and array reduction operation, optionally along specified dimensions
|
INDArray |
NDBitwise.and(INDArray x,
INDArray y)
Bitwise AND operation.
|
INDArray |
NDMath.and(INDArray x,
INDArray y)
Boolean AND operation: elementwise (x != 0) && (y != 0)
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. Returns an array with values 1 where condition is satisfied, or value 0 otherwise. |
INDArray |
NDBase.any(INDArray x,
int... dimensions)
Boolean or array reduction operation, optionally along specified dimensions
|
INDArray |
NDBase.argmax(INDArray in,
boolean keepDims,
int... dimensions)
Argmax array reduction operation, optionally along specified dimensions.
Output values are the index of the maximum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.argmax(INDArray in,
int... dimensions)
Argmax array reduction operation, optionally along specified dimensions.
Output values are the index of the maximum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.argmin(INDArray in,
boolean keepDims,
int... dimensions)
Argmin array reduction operation, optionally along specified dimensions.
Output values are the index of the minimum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.argmin(INDArray in,
int... dimensions)
Argmin array reduction operation, optionally along specified dimensions.
Output values are the index of the minimum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.asin(INDArray x)
Elementwise asin (arcsin, inverse sine) operation: out = arcsin(x)
|
INDArray |
NDMath.asinh(INDArray x)
Elementwise asinh (inverse hyperbolic sine) function: out = asinh(x)
|
INDArray |
NDBase.assign(INDArray x,
INDArray y)
Assign the contents of y to x.
Y must be broadcastable to x or the same shape. |
INDArray |
NDMath.asum(INDArray in,
boolean keepDims,
int... dimensions)
Absolute sum array reduction operation, optionally along specified dimensions: out = sum(abs(x))
|
INDArray |
NDMath.asum(INDArray in,
int... dimensions)
Absolute sum array reduction operation, optionally along specified dimensions: out = sum(abs(x))
|
INDArray |
NDMath.atan(INDArray x)
Elementwise atan (arctangent, inverse tangent) operation: out = arctangent(x)
|
INDArray |
NDMath.atan2(INDArray y,
INDArray x)
Elementwise atan (arctangent, inverse tangent) operation: out = atan2(x,y).
Similar to atan(y/x) but sigts of x and y are used to determine the location of the result |
INDArray |
NDMath.atanh(INDArray x)
Elementwise atanh (inverse hyperbolic tangent) function: out = atanh(x)
|
INDArray |
NDCNN.avgPooling2d(INDArray input,
Pooling2DConfig Pooling2DConfig)
2D Convolution layer operation - average pooling 2d
|
INDArray |
NDCNN.avgPooling3d(INDArray input,
Pooling3DConfig Pooling3DConfig)
3D convolution layer operation - average pooling 3d
|
INDArray[] |
NDBase.batchMmul(INDArray[] inputsA,
INDArray... inputsB)
Matrix multiply a batch of matrices.
|
INDArray[] |
NDBase.batchMmul(INDArray[] inputsA,
INDArray[] inputsB,
boolean transposeA,
boolean transposeB)
Matrix multiply a batch of matrices.
|
INDArray |
NDNN.batchNorm(INDArray input,
INDArray mean,
INDArray variance,
INDArray gamma,
INDArray beta,
double epsilon,
int... axis)
|
INDArray |
NDCNN.batchToSpace(INDArray x,
int[] blocks,
int[] croppingTop,
int... croppingBottom)
Convolution 2d layer batch to space operation on 4d input.
Reduces input batch dimension by rearranging data into a larger spatial dimensions |
INDArray |
NDRandom.bernoulli(double p,
DataType datatype,
long... shape)
Generate a new random INDArray, where values are randomly sampled according to a Bernoulli distribution,
with the specified probability. |
INDArray |
NDNN.biasAdd(INDArray input,
INDArray bias,
boolean nchw)
Bias addition operation: a special case of addition, typically used with CNN 4D activations and a 1D bias vector
|
INDArray |
NDRandom.binomial(int nTrials,
double p,
DataType datatype,
long... shape)
Generate a new random INDArray, where values are randomly sampled according to a Binomial distribution,
with the specified number of trials and probability. |
INDArray |
NDBitwise.bitRotl(INDArray x,
INDArray shift)
Roll integer bits to the left, i.e.
|
INDArray |
NDBitwise.bitRotr(INDArray x,
INDArray shift)
Roll integer bits to the right, i.e.
|
INDArray |
NDBitwise.bitsHammingDistance(INDArray x,
INDArray y)
Bitwise Hamming distance reduction over all elements of both input arrays.
For example, if x=01100000 and y=1010000 then the bitwise Hamming distance is 2 (due to differences at positions 0 and 1) Inputs must satisfy the following constraints: Must be same types: isSameType(x, y) |
INDArray |
NDBitwise.bitShift(INDArray x,
INDArray shift)
Shift integer bits to the left, i.e.
|
INDArray |
NDMath.bitShift(INDArray x,
INDArray shift)
Bit shift operation
|
INDArray |
NDBitwise.bitShiftRight(INDArray x,
INDArray shift)
Shift integer bits to the right, i.e.
|
INDArray |
NDMath.bitShiftRight(INDArray x,
INDArray shift)
Right bit shift operation
|
INDArray |
NDMath.bitShiftRotl(INDArray x,
INDArray shift)
Cyclic bit shift operation
|
INDArray |
NDMath.bitShiftRotr(INDArray x,
INDArray shift)
Cyclic right shift operation
|
INDArray |
NDBase.castTo(INDArray arg,
DataType datatype)
Cast the array to a new datatype - for example, Integer -> Float
|
INDArray |
NDMath.ceil(INDArray x)
Element-wise ceiling function: out = ceil(x).
Rounds each value up to the nearest integer value (if not already an integer) |
INDArray |
NDLinalg.cholesky(INDArray input)
Computes the Cholesky decomposition of one or more square matrices.
|
INDArray |
NDMath.clipByAvgNorm(INDArray x,
double clipValue,
int... dimensions)
Clips tensor values to a maximum average L2-norm.
|
INDArray |
NDBase.clipByNorm(INDArray x,
double clipValue)
Returns a clipped ndarray such that the input is normalized so that its L2 norm
is <= the specified value. |
INDArray |
NDMath.clipByNorm(INDArray x,
double clipValue,
int... dimensions)
Clipping by L2 norm, optionally along dimension(s)
if l2Norm(x,dimension) < clipValue, then input is returned unmodifed Otherwise, out[i] = in[i] * clipValue / l2Norm(in, dimensions) where each value is clipped according to the corresponding l2Norm along the specified dimensions |
INDArray |
NDBase.clipByNorm(INDArray x,
INDArray clipValue,
INDArray dimensions)
Returns a clipped ndarray such that the input is normalized so that its L2 norm
is <= the specified value. |
INDArray |
NDBase.clipByValue(INDArray x,
double clipValueMin,
double clipValueMax)
Return the clipped ndarray containing values no smaller or larger than the given min and max.
|
INDArray |
NDMath.clipByValue(INDArray x,
double clipValueMin,
double clipValueMax)
Element-wise clipping function:
out[i] = in[i] if in[i] >= clipValueMin and in[i] <= clipValueMax out[i] = clipValueMin if in[i] < clipValueMin out[i] = clipValueMax if in[i] > clipValueMax |
INDArray |
NDBase.clipByValue(INDArray x,
INDArray clipValueMin,
INDArray clipValueMax)
Return the clipped ndarray containing values no smaller or larger than the given min and max.
|
INDArray |
NDCNN.col2Im(INDArray in,
Conv2DConfig Conv2DConfig)
col2im operation for use in 2D convolution operations.
|
INDArray |
NDBase.concat(int dimension,
INDArray... inputs)
Concatenate a set of inputs along the specified dimension.
Note that inputs must have identical rank and identical dimensions, other than the dimension to stack on. For example, if 2 inputs have shape [a, x, c] and [a, y, c] and dimension = 1, then the output has shape [a, x+y, c] Inputs must satisfy the following constraints: Input arrays must all be the same datatype: isSameType(inputs) |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
DataType dataType)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
INDArray weights)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
INDArray weights,
int numClasses)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. For example, if labels = [0, 1, 1], predicted = [0, 2, 1], numClasses = 4, and weights = [1, 2, 3] [1, 0, 0, 0] [0, 3, 2, 0] [0, 0, 0, 0] [0, 0, 0, 0] |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
int numClasses)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. For example, if labels = [0, 1, 1], predicted = [0, 2, 1], and numClasses=4 then output is: [1, 0, 0, 0] [0, 1, 1, 0] [0, 0, 0, 0] [0, 0, 0, 0] |
INDArray |
NDCNN.conv1d(INDArray input,
INDArray weights,
Conv1DConfig Conv1DConfig)
Conv1d operation.
|
INDArray |
NDCNN.conv1d(INDArray input,
INDArray weights,
INDArray bias,
Conv1DConfig Conv1DConfig)
Conv1d operation.
|
INDArray |
NDCNN.conv2d(INDArray layerInput,
INDArray weights,
Conv2DConfig Conv2DConfig)
2D Convolution operation with optional bias
|
INDArray |
NDCNN.conv2d(INDArray layerInput,
INDArray weights,
INDArray bias,
Conv2DConfig Conv2DConfig)
2D Convolution operation with optional bias
|
INDArray |
NDCNN.conv3d(INDArray input,
INDArray weights,
Conv3DConfig Conv3DConfig)
Convolution 3D operation with optional bias
|
INDArray |
NDCNN.conv3d(INDArray input,
INDArray weights,
INDArray bias,
Conv3DConfig Conv3DConfig)
Convolution 3D operation with optional bias
|
INDArray |
NDMath.cos(INDArray x)
Elementwise cosine operation: out = cos(x)
|
INDArray |
NDMath.cosh(INDArray x)
Elementwise cosh (hyperbolic cosine) operation: out = cosh(x)
|
INDArray |
NDMath.cosineDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Cosine distance reduction operation.
|
INDArray |
NDLoss.cosineDistance(INDArray label,
INDArray predictions,
INDArray weights,
int dimension)
Cosine distance loss:
1 - cosineSimilarity(x,y) or 1 - sum_i label[i] * prediction[i], which isequivalent to cosine distance when both the predictions and labels are normalized. Note: This loss function assumes that both the predictions and labels are normalized to have unit l2 norm. If this is not the case, you should normalize them first by dividing by norm2(String, SDVariable, boolean, int...) along the cosine distance dimension (with keepDims=true). |
INDArray |
NDLoss.cosineDistance(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
int dimension)
Cosine distance loss:
1 - cosineSimilarity(x,y) or 1 - sum_i label[i] * prediction[i], which isequivalent to cosine distance when both the predictions and labels are normalized. Note: This loss function assumes that both the predictions and labels are normalized to have unit l2 norm. If this is not the case, you should normalize them first by dividing by norm2(String, SDVariable, boolean, int...) along the cosine distance dimension (with keepDims=true). |
INDArray |
NDMath.cosineDistance(INDArray x,
INDArray y,
int... dimensions)
Cosine distance reduction operation.
|
INDArray |
NDMath.cosineSimilarity(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Cosine similarity pairwise reduction operation.
|
INDArray |
NDMath.cosineSimilarity(INDArray x,
INDArray y,
int... dimensions)
Cosine similarity pairwise reduction operation.
|
INDArray |
NDMath.countNonZero(INDArray in,
boolean keepDims,
int... dimensions)
Count non zero array reduction operation, optionally along specified dimensions: out = count(x != 0)
|
INDArray |
NDMath.countNonZero(INDArray in,
int... dimensions)
Count non zero array reduction operation, optionally along specified dimensions: out = count(x != 0)
|
INDArray |
NDMath.countZero(INDArray in,
boolean keepDims,
int... dimensions)
Count zero array reduction operation, optionally along specified dimensions: out = count(x == 0)
|
INDArray |
NDMath.countZero(INDArray in,
int... dimensions)
Count zero array reduction operation, optionally along specified dimensions: out = count(x == 0)
|
INDArray |
NDBase.create(INDArray shape,
DataType dataType)
Return a newly created variable, with the specified shape and data type.
|
INDArray |
NDBase.create(INDArray shape,
DataType dataType,
String order,
boolean initialize)
Return a newly created variable, with the specified shape and data type.
|
INDArray |
NDNN.cReLU(INDArray x)
Concatenates a ReLU which selects only the positive part of the activation with a ReLU which selects only the negative part of the activation.
|
INDArray |
NDImage.cropAndResize(INDArray image,
INDArray cropBoxes,
INDArray boxIndices,
INDArray cropOutSize)
Given an input image and some crop boxes, extract out the image subsets and resize them to the specified size.
|
INDArray |
NDImage.cropAndResize(INDArray image,
INDArray cropBoxes,
INDArray boxIndices,
INDArray cropOutSize,
double extrapolationValue)
Given an input image and some crop boxes, extract out the image subsets and resize them to the specified size.
|
INDArray |
NDLinalg.cross(INDArray a,
INDArray b)
Computes pairwise cross product.
|
INDArray |
NDMath.cross(INDArray a,
INDArray b)
Returns the pair-wise cross product of equal size arrays a and b: a x b = ||a||x||b|| sin(theta).
Can take rank 1 or above inputs (of equal shapes), but note that the last dimension must have dimension 3 |
INDArray |
NDLoss.ctcLoss(INDArray targetLabels,
INDArray logitInput,
INDArray targetLabelLengths,
INDArray logitInputLengths)
CTC Loss: Connectionist Temporal Classification Loss.
|
INDArray |
NDMath.cube(INDArray x)
Element-wise cube function: out = x^3
|
INDArray |
NDBase.cumprod(INDArray in,
boolean exclusive,
boolean reverse,
int... axis)
Cumulative product operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a*b, a*b*c] exclusive=true, reverse=false, [0, a, a*b] exclusive=false, reverse=true: [a*b*c, b*c, c] exclusive=true, reverse=true: [b*c, c, 0] |
INDArray |
NDBase.cumprod(INDArray in,
int... axis)
Cumulative product operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a*b, a*b*c] exclusive=true, reverse=false, [0, a, a*b] exclusive=false, reverse=true: [a*b*c, b*c, c] exclusive=true, reverse=true: [b*c, c, 0] |
INDArray |
NDBase.cumsum(INDArray in,
boolean exclusive,
boolean reverse,
int... axis)
Cumulative sum operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a+b, a+b+c] exclusive=true, reverse=false, [0, a, a+b] exclusive=false, reverse=true: [a+b+c, b+c, c] exclusive=true, reverse=true: [b+c, c, 0] |
INDArray |
NDBase.cumsum(INDArray in,
int... axis)
Cumulative sum operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a+b, a+b+c] exclusive=true, reverse=false, [0, a, a+b] exclusive=false, reverse=true: [a+b+c, b+c, c] exclusive=true, reverse=true: [b+c, c, 0] |
INDArray |
NDCNN.deconv2d(INDArray layerInput,
INDArray weights,
DeConv2DConfig DeConv2DConfig)
2D deconvolution operation with optional bias
|
INDArray |
NDCNN.deconv2d(INDArray layerInput,
INDArray weights,
INDArray bias,
DeConv2DConfig DeConv2DConfig)
2D deconvolution operation with optional bias
|
INDArray |
NDCNN.deconv3d(INDArray input,
INDArray weights,
DeConv3DConfig DeConv3DConfig)
3D CNN deconvolution operation with or without optional bias
|
INDArray |
NDCNN.deconv3d(INDArray input,
INDArray weights,
INDArray bias,
DeConv3DConfig DeConv3DConfig)
3D CNN deconvolution operation with or without optional bias
|
INDArray |
NDCNN.depthToSpace(INDArray x,
int blockSize,
DataFormat dataFormat)
Convolution 2d layer batch to space operation on 4d input.
Reduces input channels dimension by rearranging data into a larger spatial dimensions Example: if input has shape [mb, 8, 2, 2] and block size is 2, then output size is [mb, 8/(2*2), 2*2, 2*2] = [mb, 2, 4, 4] |
INDArray |
NDCNN.depthWiseConv2d(INDArray layerInput,
INDArray depthWeights,
Conv2DConfig Conv2DConfig)
Depth-wise 2D convolution operation with optional bias
|
INDArray |
NDCNN.depthWiseConv2d(INDArray layerInput,
INDArray depthWeights,
INDArray bias,
Conv2DConfig Conv2DConfig)
Depth-wise 2D convolution operation with optional bias
|
INDArray |
NDLinalg.diag_part(INDArray input)
Calculates diagonal tensor.
|
INDArray |
NDLinalg.diag(INDArray input)
Calculates diagonal tensor.
|
INDArray |
NDMath.diag(INDArray x)
Returns an output variable with diagonal values equal to the specified values; off-diagonal values will be set to 0
For example, if input = [1,2,3], then output is given by: [ 1, 0, 0] [ 0, 2, 0] [ 0, 0, 3] Higher input ranks are also supported: if input has shape [a,...,R-1] then output[i,...,k,i,...,k] = input[i,...,k]. i.e., for input rank R, output has rank 2R |
INDArray |
NDMath.diagPart(INDArray x)
Extract the diagonal part from the input array.
If input is [ 1, 0, 0] [ 0, 2, 0] [ 0, 0, 3] then output is [1, 2, 3]. Supports higher dimensions: in general, out[i,...,k] = in[i,...,k,i,...,k] |
INDArray |
NDCNN.dilation2D(INDArray df,
INDArray weights,
int[] strides,
int[] rates,
boolean isSameMode)
TODO doc string
|
INDArray |
NDMath.div(INDArray x,
double value)
Scalar division operation, out = in / scalar
|
INDArray |
NDMath.div(INDArray x,
INDArray y)
Pairwise division operation, out = x / y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.dot(INDArray x,
INDArray y,
int... dimensions)
Pairwise dot product reduction along dimension
output = sum(i=0 ... |
INDArray |
NDNN.dotProductAttention(INDArray queries,
INDArray keys,
INDArray values,
INDArray mask,
boolean scaled)
This operation performs dot product attention on the given timeseries input with the given queries
out = sum(similarity(k_i, q) * v_i) similarity(k, q) = softmax(k * q) where x * q is the dot product of x and q Optionally with normalization step: similarity(k, q) = softmax(k * q / sqrt(size(q)) See also "Attention is all you need" (https://arxiv.org/abs/1706.03762, p. |
INDArray |
NDNN.dropout(INDArray input,
double inputRetainProbability)
Dropout operation
|
INDArray |
NDNN.dropoutInverted(INDArray input,
double p)
Dropout inverted operation.
|
INDArray[] |
NDBase.dynamicPartition(INDArray x,
INDArray partitions,
int numPartitions)
Dynamically partition the input variable values into the specified number of paritions, using the indices.
Example: |
INDArray |
NDBase.dynamicStitch(INDArray[] indices,
INDArray... x)
Dynamically merge the specified input arrays into a single array, using the specified indices
|
INDArray |
NDNN.elu(INDArray x)
Element-wise exponential linear unit (ELU) function:
out = x if x > 0 out = a * (exp(x) - 1) if x <= 0 with constant a = 1.0 |
INDArray |
NDMath.embeddingLookup(INDArray x,
INDArray indices,
PartitionMode PartitionMode)
Looks up ids in a list of embedding tensors.
|
INDArray |
NDMath.entropy(INDArray in,
boolean keepDims,
int... dimensions)
Entropy reduction: -sum(x * log(x))
|
INDArray |
NDMath.entropy(INDArray in,
int... dimensions)
Entropy reduction: -sum(x * log(x))
|
INDArray |
NDBase.eq(INDArray x,
double y)
Equals operation: elementwise x == y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.eq(INDArray x,
INDArray y)
Equal to operation: elementwise x == y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDMath.erf(INDArray x)
Element-wise Gaussian error function - out = erf(in)
|
INDArray |
NDMath.erfc(INDArray x)
Element-wise complementary Gaussian error function - out = erfc(in) = 1 - erf(in)
|
INDArray |
NDMath.euclideanDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Euclidean distance (l2 norm, l2 distance) reduction operation.
|
INDArray |
NDMath.euclideanDistance(INDArray x,
INDArray y,
int... dimensions)
Euclidean distance (l2 norm, l2 distance) reduction operation.
|
INDArray |
NDMath.exp(INDArray x)
Elementwise exponent function: out = exp(x) = 2.71828...^x
|
INDArray |
NDBase.expandDims(INDArray x,
int axis)
Reshape the input by adding a 1 at the specified location.
For example, if input has shape [a, b], then output shape is: axis = 0: [1, a, b] axis = 1: [a, 1, b] axis = 2: [a, b, 1] |
INDArray |
NDMath.expm1(INDArray x)
Elementwise 1.0 - exponent function: out = 1.0 - exp(x) = 1.0 - 2.71828...^x
|
INDArray |
NDRandom.exponential(double lambda,
DataType datatype,
long... shape)
Generate a new random INDArray, where values are randomly sampled according to a exponential distribution:
P(x) = lambda * exp(-lambda * x) Inputs must satisfy the following constraints: Must be positive: lambda > 0 |
INDArray |
NDImage.extractImagePatches(INDArray image,
int[] kSizes,
int[] strides,
int[] rates,
boolean sameMode)
Given an input image, extract out image patches (of size kSizes - h x w) and place them in the depth dimension.
|
INDArray |
NDCNN.extractImagePatches(INDArray input,
int kH,
int kW,
int sH,
int sW,
int rH,
int rW,
boolean sameMode)
Extract image patches
|
INDArray |
NDMath.eye(INDArray rows)
As per eye(String, int) but with the number of rows specified as a scalar INDArray
|
INDArray |
NDMath.eye(INDArray rows,
INDArray cols)
As per eye(int, int) bit with the number of rows/columns specified as scalar INDArrays
|
INDArray |
NDMath.eye(int rows)
Generate an identity matrix with the specified number of rows and columns.
|
INDArray |
NDMath.eye(int rows,
int cols)
As per eye(String, int, int, DataType) but with the default datatype, Eye.DEFAULT_DTYPE
|
INDArray |
NDMath.eye(int rows,
int cols,
DataType dataType,
int... dimensions)
Generate an identity matrix with the specified number of rows and columns
Example: |
INDArray |
NDBase.fill(INDArray shape,
DataType dataType,
double value)
Generate an output variable with the specified (dynamic) shape with all elements set to the specified value
|
INDArray |
NDMath.firstIndex(INDArray in,
Condition condition,
boolean keepDims,
int... dimensions)
First index reduction operation.
Returns a variable that contains the index of the first element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.firstIndex(INDArray in,
Condition condition,
int... dimensions)
First index reduction operation.
Returns a variable that contains the index of the first element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.floor(INDArray x)
Element-wise floor function: out = floor(x).
Rounds each value down to the nearest integer value (if not already an integer) |
INDArray |
NDMath.floorDiv(INDArray x,
INDArray y)
Pairwise floor division operation, out = floor(x / y)
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.floorMod(INDArray x,
double value)
Scalar floor modulus operation
|
INDArray |
NDMath.floorMod(INDArray x,
INDArray y)
Pairwise Modulus division operation
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.gather(INDArray df,
INDArray indices,
int axis)
Gather slices from the input variable where the indices are specified as dynamic array values.
Output shape is same as input shape, except for axis dimension, which has size equal to indices.length. |
INDArray |
NDBase.gather(INDArray df,
int[] indices,
int axis)
Gather slices from the input variable where the indices are specified as fixed int[] values.
Output shape is same as input shape, except for axis dimension, which has size equal to indices.length. |
INDArray |
NDBase.gatherNd(INDArray df,
INDArray indices)
Gather slices from df with shape specified by indices.
|
INDArray |
NDNN.gelu(INDArray x)
GELU activation function - Gaussian Error Linear Units
For more details, see Gaussian Error Linear Units (GELUs) - https://arxiv.org/abs/1606.08415 This method uses the sigmoid approximation |
INDArray |
NDRNN.gru(INDArray x,
INDArray hLast,
INDArray Wx,
INDArray Wh,
INDArray biases)
The GRU operation.
|
INDArray[] |
NDRNN.gruCell(INDArray x,
INDArray hLast,
GRUWeights GRUWeights)
The GRU cell.
|
INDArray |
NDBase.gt(INDArray x,
double y)
Greater than operation: elementwise x > y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.gt(INDArray x,
INDArray y)
Greater than operation: elementwise x > y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.gte(INDArray x,
double y)
Greater than or equals operation: elementwise x >= y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.gte(INDArray x,
INDArray y)
Greater than or equal to operation: elementwise x >= y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDMath.hammingDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Hamming distance reduction operation.
|
INDArray |
NDMath.hammingDistance(INDArray x,
INDArray y,
int... dimensions)
Hamming distance reduction operation.
|
INDArray |
NDNN.hardSigmoid(INDArray x)
Element-wise hard sigmoid function:
out[i] = 0 if in[i] <= -2.5 out[1] = 0.2*in[i]+0.5 if -2.5 < in[i] < 2.5 out[i] = 1 if in[i] >= 2.5 |
INDArray |
NDNN.hardTanh(INDArray x)
Element-wise hard tanh function:
out[i] = -1 if in[i] <= -1 out[1] = in[i] if -1 < in[i] < 1 out[i] = 1 if in[i] >= 1 |
INDArray |
NDNN.hardTanhDerivative(INDArray x)
Derivative (dOut/dIn) of the element-wise hard Tanh function - hardTanh(INDArray)
|
INDArray |
NDLoss.hingeLoss(INDArray label,
INDArray predictions,
INDArray weights)
Hinge loss: a loss function used for training classifiers.
Implements L = max(0, 1 - t * predictions) where t is the label values after internally converting to {-1,1}from the user specified {0,1}. |
INDArray |
NDLoss.hingeLoss(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Hinge loss: a loss function used for training classifiers.
Implements L = max(0, 1 - t * predictions) where t is the label values after internally converting to {-1,1}from the user specified {0,1}. |
INDArray |
NDImage.hsvToRgb(INDArray input)
Converting image from HSV to RGB format
|
INDArray |
NDLoss.huberLoss(INDArray label,
INDArray predictions,
INDArray weights,
double delta)
Huber loss function, used for robust regression.
|
INDArray |
NDLoss.huberLoss(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double delta)
Huber loss function, used for robust regression.
|
INDArray |
NDMath.iamax(INDArray in,
boolean keepDims,
int... dimensions)
Index of the max absolute value: argmax(abs(in))
see argmax(String, INDArray, boolean, int...) |
INDArray |
NDMath.iamax(INDArray in,
int... dimensions)
Index of the max absolute value: argmax(abs(in))
see argmax(String, INDArray, boolean, int...) |
INDArray |
NDMath.iamin(INDArray in,
boolean keepDims,
int... dimensions)
Index of the min absolute value: argmin(abs(in))
see argmin(String, INDArray, boolean, int...) |
INDArray |
NDMath.iamin(INDArray in,
int... dimensions)
Index of the min absolute value: argmin(abs(in))
see argmin(String, INDArray, boolean, int...) |
INDArray |
NDBase.identity(INDArray input)
Elementwise identity operation: out = x
|
INDArray |
NDCNN.im2Col(INDArray in,
Conv2DConfig Conv2DConfig)
im2col operation for use in 2D convolution operations.
|
INDArray |
NDImage.imageResize(INDArray input,
INDArray size,
boolean preserveAspectRatio,
boolean antialias,
ImageResizeMethod ImageResizeMethod)
Resize images to size using the specified method.
|
INDArray |
NDImage.imageResize(INDArray input,
INDArray size,
ImageResizeMethod ImageResizeMethod)
Resize images to size using the specified method.
|
INDArray |
NDBase.invertPermutation(INDArray input)
Compute the inverse permutation indices for a permutation operation
Example: if input is [2, 0, 1] then output is [1, 2, 0] The idea is that x.permute(input).permute(invertPermutation(input)) == x |
INDArray |
NDMath.isFinite(INDArray x)
Is finite operation: elementwise isFinite(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isInfinite(INDArray x)
Is infinite operation: elementwise isInfinite(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isMax(INDArray x)
Is maximum operation: elementwise x == max(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isNaN(INDArray x)
Is Not a Number operation: elementwise isNaN(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isNonDecreasing(INDArray x)
Is the array non decreasing?
An array is non-decreasing if for every valid i, x[i] <= x[i+1]. |
INDArray |
NDBase.isNumericTensor(INDArray x)
Is the director a numeric tensor? In the current version of ND4J/SameDiff, this always returns true/1
|
INDArray |
NDMath.isStrictlyIncreasing(INDArray x)
Is the array strictly increasing?
An array is strictly increasing if for every valid i, x[i] < x[i+1]. |
INDArray |
NDMath.jaccardDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Jaccard similarity reduction operation.
|
INDArray |
NDMath.jaccardDistance(INDArray x,
INDArray y,
int... dimensions)
Jaccard similarity reduction operation.
|
INDArray |
NDLoss.l2Loss(INDArray var)
L2 loss: 1/2 * sum(x^2)
|
INDArray |
NDMath.lastIndex(INDArray in,
Condition condition,
boolean keepDims,
int... dimensions)
Last index reduction operation.
Returns a variable that contains the index of the last element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.lastIndex(INDArray in,
Condition condition,
int... dimensions)
Last index reduction operation.
Returns a variable that contains the index of the last element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDNN.layerNorm(INDArray input,
INDArray gain,
boolean channelsFirst,
int... dimensions)
Apply Layer Normalization
y = gain * standardize(x) + bias |
INDArray |
NDNN.layerNorm(INDArray input,
INDArray gain,
INDArray bias,
boolean channelsFirst,
int... dimensions)
Apply Layer Normalization
y = gain * standardize(x) + bias |
INDArray |
NDNN.leakyRelu(INDArray x,
double alpha)
Element-wise leaky ReLU function:
out = x if x >= 0.0 out = alpha * x if x < cutoff Alpha value is most commonly set to 0.01 |
INDArray |
NDNN.leakyReluDerivative(INDArray x,
double alpha)
Leaky ReLU derivative: dOut/dIn given input.
|
INDArray |
NDBitwise.leftShift(INDArray x,
INDArray y)
Bitwise left shift operation.
|
INDArray |
NDBitwise.leftShiftCyclic(INDArray x,
INDArray y)
Bitwise left cyclical shift operation.
|
INDArray |
NDNN.linear(INDArray input,
INDArray weights,
INDArray bias)
Linear layer operation: out = mmul(in,w) + bias
Note that bias array is optional |
INDArray |
NDBase.linspace(DataType dataType,
double start,
double stop,
long number)
Create a new 1d array with values evenly spaced between values 'start' and 'stop'
For example, linspace(start=3.0, stop=4.0, number=3) will generate [3.0, 3.5, 4.0] |
INDArray |
NDBase.linspace(INDArray start,
INDArray stop,
INDArray number,
DataType dataType)
Create a new 1d array with values evenly spaced between values 'start' and 'stop'
For example, linspace(start=3.0, stop=4.0, number=3) will generate [3.0, 3.5, 4.0] |
INDArray[] |
NDMath.listDiff(INDArray x,
INDArray y)
Calculates difference between inputs X and Y.
|
INDArray |
NDCNN.localResponseNormalization(INDArray input,
LocalResponseNormalizationConfig LocalResponseNormalizationConfig)
2D convolution layer operation - local response normalization
|
INDArray |
NDMath.log(INDArray x)
Element-wise logarithm function (base e - natural logarithm): out = log(x)
|
INDArray |
NDMath.log(INDArray x,
double base)
Element-wise logarithm function (with specified base): out = log_{base}(x)
|
INDArray |
NDMath.log1p(INDArray x)
Elementwise natural logarithm function: out = log_e (1 + x)
|
INDArray |
NDLinalg.logdet(INDArray input)
Calculates log of determinant.
|
INDArray |
NDMath.logEntropy(INDArray in,
boolean keepDims,
int... dimensions)
Log entropy reduction: log(-sum(x * log(x)))
|
INDArray |
NDMath.logEntropy(INDArray in,
int... dimensions)
Log entropy reduction: log(-sum(x * log(x)))
|
INDArray |
NDLoss.logLoss(INDArray label,
INDArray predictions)
Log loss, i.e., binary cross entropy loss, usually used for binary multi-label classification.
|
INDArray |
NDLoss.logLoss(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double epsilon)
Log loss, i.e., binary cross entropy loss, usually used for binary multi-label classification.
|
INDArray |
NDRandom.logNormal(double mean,
double stddev,
DataType datatype,
long... shape)
Generate a new random INDArray, where values are randomly sampled according to a Log Normal distribution,
i.e., log(x) ~ N(mean, stdev) |
INDArray |
NDLoss.logPoisson(INDArray label,
INDArray predictions,
INDArray weights,
boolean full)
Log poisson loss: a loss function used for training classifiers.
Implements L = exp(c) - z * c where c is log(predictions) and z is labels. |
INDArray |
NDLoss.logPoisson(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
boolean full)
Log poisson loss: a loss function used for training classifiers.
Implements L = exp(c) - z * c where c is log(predictions) and z is labels. |
INDArray |
NDNN.logSigmoid(INDArray x)
Element-wise sigmoid function: out[i] = log(sigmoid(in[i]))
|
INDArray |
NDNN.logSoftmax(INDArray x)
Log softmax activation
|
INDArray |
NDNN.logSoftmax(INDArray x,
int dimension)
Log softmax activation
|
INDArray |
NDMath.logSumExp(INDArray input,
int... dimensions)
Log-sum-exp reduction (optionally along dimension).
Computes log(sum(exp(x)) |
INDArray |
NDRNN.lstmblock(INDArray maxTSLength,
INDArray x,
INDArray cLast,
INDArray yLast,
LSTMWeights LSTMWeights,
LSTMConfiguration LSTMConfiguration)
The LSTM block
|
INDArray |
NDRNN.lstmblock(INDArray x,
LSTMWeights LSTMWeights,
LSTMConfiguration LSTMConfiguration)
The LSTM block
|
INDArray[] |
NDRNN.lstmCell(INDArray x,
INDArray cLast,
INDArray yLast,
LSTMWeights LSTMWeights,
LSTMConfiguration LSTMConfiguration)
The LSTM cell.
|
INDArray[] |
NDRNN.lstmLayer(INDArray x,
INDArray cLast,
INDArray yLast,
INDArray maxTSLength,
LSTMLayerWeights LSTMLayerWeights,
LSTMLayerConfig LSTMLayerConfig)
Long Short-Term Memory layer - Hochreiter 1997.
SUPPORTS following data formats: for unidirectional: TNS: shapes [timeLength, numExamples, inOutSize] NST: shapes [numExamples, inOutSize, timeLength] NTS: shapes [numExamples, timeLength, inOutSize] for bidirectional: T2NS: shapes [timeLength, 2, numExamples, inOutSize] (for ONNX) SUPPORTS following direction modes: FWD: forward BWD: backward BIDIR_SUM: bidirectional sum BIDIR_CONCAT: bidirectional concat BIDIR_EXTRA_DIM: bidirectional extra output dim (in conjunction with format dataFormat - T2NS) You may use different gate configurations: specify gate/cell/out aplha/beta and numbers of activations for gate/cell/out described in activations enum ("RELU","SIGMOID","AFFINE","LEAKY_RELU","THRESHHOLD_RELU","SCALED_TAHN","HARD_SIGMOID","ELU","SOFTSIGN","SOFTPLUS") Also this layer supports MKLDNN (DNNL) and cuDNN acceleration |
INDArray[] |
NDRNN.lstmLayer(INDArray x,
LSTMLayerWeights LSTMLayerWeights,
LSTMLayerConfig LSTMLayerConfig)
Long Short-Term Memory layer - Hochreiter 1997.
SUPPORTS following data formats: for unidirectional: TNS: shapes [timeLength, numExamples, inOutSize] NST: shapes [numExamples, inOutSize, timeLength] NTS: shapes [numExamples, timeLength, inOutSize] for bidirectional: T2NS: shapes [timeLength, 2, numExamples, inOutSize] (for ONNX) SUPPORTS following direction modes: FWD: forward BWD: backward BIDIR_SUM: bidirectional sum BIDIR_CONCAT: bidirectional concat BIDIR_EXTRA_DIM: bidirectional extra output dim (in conjunction with format dataFormat - T2NS) You may use different gate configurations: specify gate/cell/out aplha/beta and numbers of activations for gate/cell/out described in activations enum ("RELU","SIGMOID","AFFINE","LEAKY_RELU","THRESHHOLD_RELU","SCALED_TAHN","HARD_SIGMOID","ELU","SOFTSIGN","SOFTPLUS") Also this layer supports MKLDNN (DNNL) and cuDNN acceleration |
INDArray |
NDLinalg.lstsq(INDArray matrix,
INDArray rhs,
double l2_reguralizer)
Solver for linear squares problems.
|
INDArray |
NDLinalg.lstsq(INDArray matrix,
INDArray rhs,
double l2_reguralizer,
boolean fast)
Solver for linear squares problems.
|
INDArray |
NDBase.lt(INDArray x,
double y)
Less than operation: elementwise x < y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.lt(INDArray x,
INDArray y)
Less than operation: elementwise x < y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.lte(INDArray x,
double y)
Less than or equals operation: elementwise x <= y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.lte(INDArray x,
INDArray y)
Less than or equal to operation: elementwise x <= y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDLinalg.lu(INDArray input)
Computes LU decomposition.
|
INDArray |
NDMath.manhattanDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Manhattan distance (l1 norm, l1 distance) reduction operation.
|
INDArray |
NDMath.manhattanDistance(INDArray x,
INDArray y,
int... dimensions)
Manhattan distance (l1 norm, l1 distance) reduction operation.
|
INDArray |
NDBase.matchCondition(INDArray in,
Condition condition)
Returns a boolean mask of equal shape to the input, where the condition is satisfied - value 1 where satisfied, 0 otherwise
|
INDArray |
NDBase.matchConditionCount(INDArray in,
Condition condition)
Returns a count of the number of elements that satisfy the condition
|
INDArray |
NDBase.matchConditionCount(INDArray in,
Condition condition,
boolean keepDim,
int... dimensions)
Returns a count of the number of elements that satisfy the condition (for each slice along the specified dimensions)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.matchConditionCount(INDArray in,
Condition condition,
int... dimensions)
Returns a count of the number of elements that satisfy the condition (for each slice along the specified dimensions)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDLinalg.matmul(INDArray a,
INDArray b)
Performs matrix mutiplication on input tensors.
|
INDArray[] |
NDLinalg.matrixBandPart(INDArray input,
int minLower,
int maxUpper)
Copy a tensor setting outside a central band in each innermost matrix.
|
INDArray |
NDMath.matrixDeterminant(INDArray in)
Matrix determinant op.
|
INDArray |
NDMath.matrixInverse(INDArray in)
Matrix inverse op.
|
INDArray |
NDBase.max(INDArray x,
boolean keepDims,
int... dimensions)
Max array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.max(INDArray first,
INDArray second)
Element-wise maximum operation: out[i] = max(first[i], second[i])
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.max(INDArray x,
INDArray y)
Pairwise max operation, out = max(x, y)
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.max(INDArray x,
int... dimensions)
Max array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDCNN.maxPooling2d(INDArray input,
Pooling2DConfig Pooling2DConfig)
2D Convolution layer operation - max pooling 2d
|
INDArray |
NDCNN.maxPooling3d(INDArray input,
Pooling3DConfig Pooling3DConfig)
3D convolution layer operation - max pooling 3d operation.
|
INDArray[] |
NDCNN.maxPoolWithArgmax(INDArray input,
Pooling2DConfig Pooling2DConfig)
2D Convolution layer operation - Max pooling on the input and outputs both max values and indices
|
INDArray |
NDBase.mean(INDArray x,
boolean keepDims,
int... dimensions)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
boolean keepDims,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.mean(INDArray x,
INDArray dimensions)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
INDArray dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.mean(INDArray x,
INDArray dimensions,
boolean keepDims)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
INDArray dimensions,
boolean keepDims)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.mean(INDArray x,
int... dimensions)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDLoss.meanPairwiseSquaredError(INDArray label,
INDArray predictions,
INDArray weights)
Mean pairwise squared error.
MPWSE loss calculates the difference between pairs of consecutive elements in the predictions and labels arrays. For example, if predictions = [p0, p1, p2] and labels are [l0, l1, l2] then MPWSE is: [((p0-p1) - (l0-l1))^2 + ((p0-p2) - (l0-l2))^2 + ((p1-p2) - (l1-l2))^2] / 3 |
INDArray |
NDLoss.meanPairwiseSquaredError(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Mean pairwise squared error.
MPWSE loss calculates the difference between pairs of consecutive elements in the predictions and labels arrays. For example, if predictions = [p0, p1, p2] and labels are [l0, l1, l2] then MPWSE is: [((p0-p1) - (l0-l1))^2 + ((p0-p2) - (l0-l2))^2 + ((p1-p2) - (l1-l2))^2] / 3 |
INDArray |
NDLoss.meanSquaredError(INDArray label,
INDArray predictions,
INDArray weights)
Mean squared error loss function.
|
INDArray |
NDLoss.meanSquaredError(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Mean squared error loss function.
|
INDArray |
NDBase.merge(INDArray x,
INDArray y)
The merge operation is a control operation that forwards the either of the inputs to the output, when
the first of them becomes available. |
INDArray |
NDMath.mergeAdd(INDArray... inputs)
Merge add function: merges an arbitrary number of equal shaped arrays using element-wise addition:
out = sum_i in[i] |
INDArray |
NDMath.mergeAvg(INDArray... inputs)
Merge average function: merges an arbitrary number of equal shaped arrays using element-wise mean operation:
out = mean_i in[i] |
INDArray |
NDMath.mergeMax(INDArray... inputs)
Merge max function: merges an arbitrary number of equal shaped arrays using element-wise maximum operation:
out = max_i in[i] |
INDArray |
NDMath.mergeMaxIndex(INDArray... x)
Return array of max elements indices with along tensor dimensions
|
INDArray |
NDMath.mergeMaxIndex(INDArray[] x,
DataType dataType)
Return array of max elements indices with along tensor dimensions
|
INDArray[] |
NDMath.meshgrid(INDArray[] inputs,
boolean cartesian)
Broadcasts parameters for evaluation on an N-D grid.
|
INDArray |
NDBase.min(INDArray x,
boolean keepDims,
int... dimensions)
Minimum array reduction operation, optionally along specified dimensions.
|
INDArray |
NDBase.min(INDArray first,
INDArray second)
Element-wise minimum operation: out[i] = min(first[i], second[i])
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.min(INDArray x,
INDArray y)
Pairwise max operation, out = min(x, y)
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.min(INDArray x,
int... dimensions)
Minimum array reduction operation, optionally along specified dimensions.
|
INDArray |
NDBase.minMax(int datatype,
int minOrMax)
Return a scalar array reflecting the min or max value for a given data type.
|
INDArray |
NDLinalg.mmul(INDArray x,
INDArray y)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDBase.mmul(INDArray x,
INDArray y)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDLinalg.mmul(INDArray x,
INDArray y,
boolean transposeX,
boolean transposeY,
boolean transposeZ)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDBase.mmul(INDArray x,
INDArray y,
boolean transposeX,
boolean transposeY,
boolean transposeZ)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDMath.mod(INDArray x,
INDArray y)
Pairwise modulus (remainder) operation, out = x % y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray[] |
NDMath.moments(INDArray input,
INDArray axes,
boolean keepDims)
Calculate the mean and (population) variance for the input variable, for the specified axis
|
INDArray[] |
NDMath.moments(INDArray input,
int[] axes,
boolean keepDims)
Calculate the mean and (population) variance for the input variable, for the specified axis
|
INDArray |
NDMath.mul(INDArray x,
double value)
Scalar multiplication operation, out = in * scalar
|
INDArray |
NDMath.mul(INDArray x,
INDArray y)
Pairwise multiplication operation, out = x * y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDNN.multiHeadDotProductAttention(INDArray queries,
INDArray keys,
INDArray values,
INDArray Wq,
INDArray Wk,
INDArray Wv,
INDArray Wo,
INDArray mask,
boolean scaled)
This performs multi-headed dot product attention on the given timeseries input
out = concat(head_1, head_2, ..., head_n) * Wo head_i = dot_product_attention(Wq_i*q, Wk_i*k, Wv_i*v) Optionally with normalization when calculating the attention for each head. See also "Attention is all you need" (https://arxiv.org/abs/1706.03762, pp. |
INDArray |
NDMath.neg(INDArray x)
Elementwise negative operation: out = -x
|
INDArray |
NDBase.neq(INDArray x,
double y)
Not equals operation: elementwise x != y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.neq(INDArray x,
INDArray y)
Not equal to operation: elementwise x != y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDImage.nonMaxSuppression(INDArray boxes,
INDArray scores,
int maxOutSize,
double iouThreshold,
double scoreThreshold)
Greedily selects a subset of bounding boxes in descending order of score
|
INDArray |
NDBase.norm1(INDArray x,
boolean keepDims,
int... dimensions)
Norm1 (L1 norm) reduction operation: The output contains the L1 norm for each tensor/subset along the specified dimensions:
out = sum_i abs(x[i]) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm1(INDArray in,
boolean keepDims,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.norm1(INDArray in,
INDArray dimensions)
Sum of absolute differences.
|
INDArray |
NDMath.norm1(INDArray in,
INDArray dimensions,
boolean keepDims)
Sum of absolute differences.
|
INDArray |
NDBase.norm1(INDArray x,
int... dimensions)
Norm1 (L1 norm) reduction operation: The output contains the L1 norm for each tensor/subset along the specified dimensions:
out = sum_i abs(x[i]) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm1(INDArray in,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.norm2(INDArray x,
boolean keepDims,
int... dimensions)
Norm2 (L2 norm) reduction operation: The output contains the L2 norm for each tensor/subset along the specified dimensions:
out = sqrt(sum_i x[i]^2) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm2(INDArray in,
boolean keepDims,
int... dimensions)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray |
NDMath.norm2(INDArray in,
INDArray dimensions)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray |
NDMath.norm2(INDArray in,
INDArray dimensions,
boolean keepDims)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray |
NDBase.norm2(INDArray x,
int... dimensions)
Norm2 (L2 norm) reduction operation: The output contains the L2 norm for each tensor/subset along the specified dimensions:
out = sqrt(sum_i x[i]^2) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm2(INDArray in,
int... dimensions)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray |
NDRandom.normal(double mean,
double stddev,
DataType datatype,
long... shape)
Generate a new random INDArray, where values are randomly sampled according to a Gaussian (normal) distribution,
N(mean, stdev) |
INDArray[] |
NDMath.normalizeMoments(INDArray counts,
INDArray means,
INDArray variances,
double shift)
Calculate the mean and variance from the sufficient statistics
|
INDArray |
NDRandom.normalTruncated(double mean,
double stddev,
DataType datatype,
long... shape)
Generate a new random INDArray, where values are randomly sampled according to a Gaussian (normal) distribution,
N(mean, stdev). |
INDArray |
NDBase.normmax(INDArray x,
boolean keepDims,
int... dimensions)
Max norm (infinity norm) reduction operation: The output contains the max norm for each tensor/subset along the
specified dimensions: out = max(abs(x[i])) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.normMax(INDArray in,
boolean keepDims,
int... dimensions)
Differences between max absolute value
|
INDArray |
NDMath.normMax(INDArray in,
INDArray dimensions)
Differences between max absolute value
|
INDArray |
NDMath.normMax(INDArray in,
INDArray dimensions,
boolean keepDims)
Differences between max absolute value
|
INDArray |
NDBase.normmax(INDArray x,
int... dimensions)
Max norm (infinity norm) reduction operation: The output contains the max norm for each tensor/subset along the
specified dimensions: out = max(abs(x[i])) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.normMax(INDArray in,
int... dimensions)
Differences between max absolute value
|
INDArray |
NDBase.oneHot(INDArray indices,
int depth)
Convert the array to a one-hot array with values 0 and 1 for each entry
If input has shape [ a, ..., n] then output has shape [ a, ..., n, depth], with out[i, ..., j, in[i,...,j]] = 1 with other values being set to 0 see oneHot(SDVariable, int, int, double, double) |
INDArray |
NDBase.oneHot(INDArray indices,
int depth,
int axis,
double on,
double off)
Convert the array to a one-hot array with values and for each entry
If input has shape [ a, ..., n] then output has shape [ a, ..., n, depth], with {out[i, ..., j, in[i,...,j]] with other values being set to |
INDArray |
NDBase.oneHot(INDArray indices,
int depth,
int axis,
double on,
double off,
DataType dataType)
Convert the array to a one-hot array with values and for each entry
If input has shape [ a, ..., n] then output has shape [ a, ..., n, depth], with {out[i, ..., j, in[i,...,j]] with other values being set to |
INDArray |
NDBase.onesLike(INDArray input)
Return a variable of all 1s, with the same shape as the input variable.
|
INDArray |
NDBase.onesLike(INDArray input,
DataType dataType)
As per onesLike(String, SDVariable) but the output datatype may be specified
|
INDArray |
NDBitwise.or(INDArray x,
INDArray y)
Bitwise OR operation.
|
INDArray |
NDMath.or(INDArray x,
INDArray y)
Boolean OR operation: elementwise (x != 0) || (y != 0)
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. Returns an array with values 1 where condition is satisfied, or value 0 otherwise. |
INDArray |
NDNN.pad(INDArray input,
INDArray padding,
double constant)
Padding operation
|
INDArray |
NDImage.pad(INDArray input,
INDArray padding,
Mode Mode,
double padValue)
Pads an image according to the given padding type
|
INDArray |
NDNN.pad(INDArray input,
INDArray padding,
PadMode PadMode,
double constant)
Padding operation
|
INDArray |
NDBase.permute(INDArray x,
INDArray dimensions)
Array permutation operation: permute the dimensions according to the specified permutation indices.
Example: if input has shape [a,b,c] and dimensions = [2,0,1] the output has shape [c,a,b] |
INDArray |
NDBase.permute(INDArray x,
int... dimensions)
Array permutation operation: permute the dimensions according to the specified permutation indices.
Example: if input has shape [a,b,c] and dimensions = [2,0,1] the output has shape [c,a,b] |
INDArray |
NDMath.pow(INDArray x,
double value)
Element-wise power function: out = x^value
|
INDArray |
NDMath.pow(INDArray x,
INDArray y)
Element-wise (broadcastable) power function: out = x[i]^y[i]
|
INDArray |
NDNN.preciseGelu(INDArray x)
GELU activation function - Gaussian Error Linear Units
For more details, see Gaussian Error Linear Units (GELUs) - https://arxiv.org/abs/1606.08415 This method uses the precise method |
INDArray |
NDNN.prelu(INDArray input,
INDArray alpha,
int... sharedAxes)
PReLU (Parameterized Rectified Linear Unit) operation.
|
INDArray |
NDBase.prod(INDArray x,
boolean keepDims,
int... dimensions)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
boolean keepDims,
int... dimensions)
The max of an array along each dimension
|
INDArray |
NDBase.prod(INDArray x,
INDArray dimensions)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
INDArray dimensions)
The product of an array long each dimension
|
INDArray |
NDBase.prod(INDArray x,
INDArray dimensions,
boolean keepDims)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
INDArray dimensions,
boolean keepDims)
The product of an array long each dimension
|
INDArray |
NDBase.prod(INDArray x,
int... dimensions)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
int... dimensions)
The max of an array along each dimension
|
INDArray[] |
NDLinalg.qr(INDArray input)
Computes the QR decompositions of input matrix.
|
INDArray[] |
NDLinalg.qr(INDArray input,
boolean full)
Computes the QR decompositions of input matrix.
|
INDArray |
NDImage.randomCrop(INDArray input,
INDArray shape)
Randomly crops image
|
INDArray |
NDBase.range(double from,
double to,
double step,
DataType dataType)
Create a new variable with a 1d array, where the values start at from and increment by step
up to (but not including) limit. For example, range(1.0, 3.0, 0.5) will return [1.0, 1.5, 2.0, 2.5] |
INDArray |
NDBase.range(INDArray from,
INDArray to,
INDArray step,
DataType dataType)
Create a new variable with a 1d array, where the values start at from and increment by step
up to (but not including) limit. For example, range(1.0, 3.0, 0.5) will return [1.0, 1.5, 2.0, 2.5] |
INDArray |
NDBase.rank(INDArray in)
Returns the rank (number of dimensions, i.e., length(shape)) of the specified INDArray as a 0D scalar variable
|
INDArray |
NDMath.rationalTanh(INDArray x)
Rational Tanh Approximation elementwise function, as described in the paper:
Compact Convolutional Neural Network Cascade for Face Detection This is a faster Tanh approximation |
INDArray |
NDMath.rdiv(INDArray x,
double value)
Scalar reverse division operation, out = scalar / in
|
INDArray |
NDMath.rdiv(INDArray x,
INDArray y)
Pairwise reverse division operation, out = y / x
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.reciprocal(INDArray x)
Element-wise reciprocal (inverse) function: out[i] = 1 / in[i]
|
INDArray |
NDMath.rectifiedTanh(INDArray x)
Rectified tanh operation: max(0, tanh(in))
|
INDArray |
NDMath.reduceAMax(INDArray in,
boolean keepDims,
int... dimensions)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAMax(INDArray in,
INDArray dimensions)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAMax(INDArray in,
INDArray dimensions,
boolean keepDims)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAMax(INDArray in,
int... dimensions)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
boolean keepDims,
int... dimensions)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
INDArray dimensions)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
INDArray dimensions,
boolean keepDims)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
int... dimensions)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
boolean keepDims,
int... dimensions)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
INDArray dimensions)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
INDArray dimensions,
boolean keepDims)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
int... dimensions)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceMax(INDArray in,
boolean keepDims,
int... dimensions)
The max of an array along each dimension
|
INDArray |
NDMath.reduceMax(INDArray in,
INDArray dimensions)
The max of an array long each dimension
|
INDArray |
NDMath.reduceMax(INDArray in,
INDArray dimensions,
boolean keepDims)
The max of an array long each dimension
|
INDArray |
NDMath.reduceMax(INDArray in,
int... dimensions)
The max of an array along each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
boolean keepDims,
int... dimensions)
The minimum of an array along each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
INDArray dimensions)
The minimum of an array long each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
INDArray dimensions,
boolean keepDims)
The minimum of an array long each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
int... dimensions)
The minimum of an array along each dimension
|
INDArray |
NDNN.relu(INDArray x,
double cutoff)
Element-wise rectified linear function with specified cutoff:
out[i] = in[i] if in[i] >= cutoff out[i] = 0 otherwise |
INDArray |
NDNN.relu6(INDArray x,
double cutoff)
Element-wise "rectified linear 6" function with specified cutoff:
out[i] = min(max(in, cutoff), 6) |
INDArray |
NDNN.reluLayer(INDArray input,
INDArray weights,
INDArray bias)
ReLU (Rectified Linear Unit) layer operation: out = relu(mmul(in,w) + bias)
Note that bias array is optional |
INDArray |
NDBase.repeat(INDArray input,
INDArray repeats,
int axis)
A tensor with the shape of input minus the specified axis with elements repeated along the specified axis.
|
INDArray |
NDBase.replaceWhere(INDArray update,
double value,
Condition condition)
Element-wise replace where condition:
out[i] = value if condition(update[i]) is satisfied, or out[i] = update[i] if condition(update[i]) is NOT satisfied |
INDArray |
NDBase.replaceWhere(INDArray update,
INDArray from,
Condition condition)
Element-wise replace where condition:
out[i] = from[i] if condition(update[i]) is satisfied, or out[i] = update[i] if condition(update[i]) is NOT satisfied |
INDArray |
NDBase.reshape(INDArray x,
INDArray shape)
Reshape the input variable to the specified (fixed) shape.
|
INDArray |
NDBase.reshape(INDArray x,
long... shape)
Reshape the input variable to the specified (fixed) shape.
|
INDArray |
NDImage.resizeBiCubic(INDArray input,
INDArray size,
boolean alignCorners,
boolean alignPixelCenters)
Resize images to size using the specified method.
|
INDArray |
NDImage.resizeBiLinear(INDArray input,
int height,
int width,
boolean alignCorners,
boolean halfPixelCenters)
Resize images to size using the specified method.
|
INDArray |
NDBase.reverse(INDArray x,
int... dimensions)
Reverse the values of an array for the specified dimensions
If input is: [ 1, 2, 3] [ 4, 5, 6] then reverse(in, 0): [3, 2, 1] [6, 5, 4] reverse(in, 1): [4, 5, 6] [1, 2 3] |
INDArray |
NDBase.reverseSequence(INDArray x,
INDArray seq_lengths)
Reverse sequence op: for each slice along dimension seqDimension, the first seqLength values are reversed
|
INDArray |
NDBase.reverseSequence(INDArray x,
INDArray seq_lengths,
int seqDim,
int batchDim)
Reverse sequence op: for each slice along dimension seqDimension, the first seqLength values are reversed
|
INDArray |
NDImage.rgbToHsv(INDArray input)
Converting array from HSV to RGB format
|
INDArray |
NDImage.rgbToYiq(INDArray input)
Converting array from RGB to YIQ format
|
INDArray |
NDImage.rgbToYuv(INDArray input)
Converting array from RGB to YUV format
|
INDArray |
NDBitwise.rightShift(INDArray x,
INDArray y)
Bitwise right shift operation.
|
INDArray |
NDBitwise.rightShiftCyclic(INDArray x,
INDArray y)
Bitwise right cyclical shift operation.
|
INDArray |
NDMath.round(INDArray x)
Element-wise round function: out = round(x).
Rounds (up or down depending on value) to the nearest integer value. |
INDArray |
NDMath.rsqrt(INDArray x)
Element-wise reciprocal (inverse) of square root: out = 1.0 / sqrt(x)
|
INDArray |
NDMath.rsub(INDArray x,
double value)
Scalar reverse subtraction operation, out = scalar - in
|
INDArray |
NDMath.rsub(INDArray x,
INDArray y)
Pairwise reverse subtraction operation, out = y - x
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.scalarFloorMod(INDArray in,
double value)
Element-wise scalar floor modulus operation: out = floorMod(in, value).
i.e., returns the remainder after division by 'value' |
INDArray |
NDBase.scalarMax(INDArray in,
double value)
Element-wise scalar maximum operation: out = max(in, value)
|
INDArray |
NDBase.scalarMin(INDArray in,
double value)
Element-wise scalar minimum operation: out = min(in, value)
|
INDArray |
NDBase.scalarSet(INDArray in,
double set)
Return a variable with equal shape to the input, but all elements set to value 'set'
|
INDArray |
NDBase.scatterAdd(INDArray ref,
INDArray indices,
INDArray updates)
Scatter addition operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterDiv(INDArray ref,
INDArray indices,
INDArray updates)
Scatter division operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterMax(INDArray ref,
INDArray indices,
INDArray updates)
Scatter max operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterMin(INDArray ref,
INDArray indices,
INDArray updates)
Scatter min operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterMul(INDArray ref,
INDArray indices,
INDArray updates)
Scatter multiplication operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterSub(INDArray ref,
INDArray indices,
INDArray updates)
Scatter subtraction operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterUpdate(INDArray ref,
INDArray indices,
INDArray updates)
Scatter update operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.segmentMax(INDArray data,
INDArray segmentIds)
Segment max operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentMean(INDArray data,
INDArray segmentIds)
Segment mean operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentMin(INDArray data,
INDArray segmentIds)
Segment min operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentProd(INDArray data,
INDArray segmentIds)
Segment product operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentSum(INDArray data,
INDArray segmentIds)
Segment sum operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDNN.selu(INDArray x)
Element-wise SeLU function - Scaled exponential Lineal Unit: see Self-Normalizing Neural Networks
out[i] = scale * alpha * (exp(in[i])-1) if in[i]>0, or 0 if in[i] <= 0 Uses default scale and alpha values. |
INDArray |
NDCNN.separableConv2d(INDArray layerInput,
INDArray depthWeights,
INDArray pointWeights,
Conv2DConfig Conv2DConfig)
Separable 2D convolution operation with optional bias
|
INDArray |
NDCNN.separableConv2d(INDArray layerInput,
INDArray depthWeights,
INDArray pointWeights,
INDArray bias,
Conv2DConfig Conv2DConfig)
Separable 2D convolution operation with optional bias
|
INDArray |
NDBase.sequenceMask(INDArray lengths,
DataType dataType)
see sequenceMask(String, SDVariable, SDVariable, DataType)
|
INDArray |
NDBase.sequenceMask(INDArray lengths,
INDArray maxLen,
DataType dataType)
Generate a sequence mask (with values 0 or 1) based on the specified lengths
Specifically, out[i, ..., k, j] = (j < lengths[i, ..., k] ? 1.0 : 0.0) |
INDArray |
NDBase.sequenceMask(INDArray lengths,
int maxLen,
DataType dataType)
Generate a sequence mask (with values 0 or 1) based on the specified lengths
Specifically, out[i, ..., k, j] = (j < lengths[i, ..., k] ? 1.0 : 0.0) |
INDArray |
NDMath.setDiag(INDArray in,
INDArray diag)
Set the diagonal value to the specified values
If input is [ a, b, c] [ d, e, f] [ g, h, i] and diag = [ 1, 2, 3] then output is [ 1, b, c] [ d, 2, f] [ g, h, 3] |
INDArray[] |
NDBase.setShape(INDArray input,
INDArray shape)
Sets an inplace shape on the passed in input.
|
INDArray |
NDMath.shannonEntropy(INDArray in,
boolean keepDims,
int... dimensions)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDMath.shannonEntropy(INDArray in,
INDArray dimensions)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDMath.shannonEntropy(INDArray in,
INDArray dimensions,
boolean keepDims)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDMath.shannonEntropy(INDArray in,
int... dimensions)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDBase.shape(INDArray input)
Returns the shape of the specified INDArray as a 1D INDArray
|
INDArray |
NDNN.sigmoid(INDArray x)
Element-wise sigmoid function: out[i] = 1.0/(1+exp(-in[i]))
|
INDArray |
NDLoss.sigmoidCrossEntropy(INDArray label,
INDArray predictionLogits,
INDArray weights)
Sigmoid cross entropy: applies the sigmoid activation function on the input logits (input "pre-sigmoid preductions")
and implements the binary cross entropy loss function. |
INDArray |
NDLoss.sigmoidCrossEntropy(INDArray label,
INDArray predictionLogits,
INDArray weights,
LossReduce lossReduce,
double labelSmoothing)
Sigmoid cross entropy: applies the sigmoid activation function on the input logits (input "pre-sigmoid preductions")
and implements the binary cross entropy loss function. |
INDArray |
NDNN.sigmoidDerivative(INDArray x,
INDArray wrt)
Element-wise sigmoid function derivative: dL/dIn given input and dL/dOut
|
INDArray |
NDMath.sign(INDArray x)
Element-wise sign (signum) function:
out = -1 if in < 0 out = 0 if in = 0 out = 1 if in > 0 |
INDArray |
NDMath.sin(INDArray x)
Elementwise sine operation: out = sin(x)
|
INDArray |
NDMath.sinh(INDArray x)
Elementwise sinh (hyperbolic sine) operation: out = sinh(x)
|
INDArray |
NDBase.size(INDArray in)
Returns the size (number of elements, i.e., prod(shape)) of the specified INDArray as a 0D scalar variable
|
INDArray |
NDBase.sizeAt(INDArray in,
int dimension)
Returns a rank 0 (scalar) variable for the size of the specified dimension.
For example, if X has shape [10,20,30] then sizeAt(X,1)=20. |
INDArray |
NDBase.slice(INDArray input,
INDArray begin,
INDArray size)
Get a subset of the specified input, by specifying the first element and the size of the array.
For example, if input is: [a, b, c] [d, e, f] then slice(input, begin=[0,1], size=[2,1] will return: [b] [e] Note that for each dimension i, begin[i] + size[i] <= input.size(i) |
INDArray |
NDBase.slice(INDArray input,
int[] begin,
int... size)
Get a subset of the specified input, by specifying the first element and the size of the array.
For example, if input is: [a, b, c] [d, e, f] then slice(input, begin=[0,1], size=[2,1] will return: [b] [e] Note that for each dimension i, begin[i] + size[i] <= input.size(i) |
INDArray |
NDNN.softmax(INDArray x)
Softmax activation, along the specified dimension
|
INDArray |
NDNN.softmax(INDArray x,
int dimension)
Softmax activation, along the specified dimension
|
INDArray |
NDLoss.softmaxCrossEntropy(INDArray oneHotLabels,
INDArray logitPredictions,
INDArray weights)
Applies the softmax activation function to the input, then implement multi-class cross entropy:
-sum_classes label[i] * log(p[c]) where p = softmax(logits)If LossReduce#NONE is used, returned shape is [numExamples] out for [numExamples, numClasses] predicitons/labels; otherwise, the output is a scalar. |
INDArray |
NDLoss.softmaxCrossEntropy(INDArray oneHotLabels,
INDArray logitPredictions,
INDArray weights,
LossReduce lossReduce,
double labelSmoothing)
Applies the softmax activation function to the input, then implement multi-class cross entropy:
-sum_classes label[i] * log(p[c]) where p = softmax(logits)If LossReduce#NONE is used, returned shape is [numExamples] out for [numExamples, numClasses] predicitons/labels; otherwise, the output is a scalar. |
INDArray |
NDNN.softmaxDerivative(INDArray x,
INDArray wrt,
int dimension)
Softmax derivative function
|
INDArray |
NDNN.softplus(INDArray x)
Element-wise softplus function: out = log(exp(x) + 1)
|
INDArray |
NDNN.softsign(INDArray x)
Element-wise softsign function: out = x / (abs(x) + 1)
|
INDArray |
NDNN.softsignDerivative(INDArray x)
Element-wise derivative (dOut/dIn) of the softsign function softsign(INDArray)
|
INDArray |
NDLinalg.solve(INDArray matrix,
INDArray rhs)
Solver for systems of linear equations.
|
INDArray |
NDLinalg.solve(INDArray matrix,
INDArray rhs,
boolean adjoint)
Solver for systems of linear equations.
|
INDArray |
NDCNN.spaceToBatch(INDArray x,
int[] blocks,
int[] paddingTop,
int... paddingBottom)
Convolution 2d layer space to batch operation on 4d input.
Increases input batch dimension by rearranging data from spatial dimensions into batch dimension |
INDArray |
NDCNN.spaceToDepth(INDArray x,
int blockSize,
DataFormat dataFormat)
Convolution 2d layer space to depth operation on 4d input.
Increases input channels (reduced spatial dimensions) by rearranging data into a larger channels dimension Example: if input has shape [mb, 2, 4, 4] and block size is 2, then output size is [mb, 8/(2*2), 2*2, 2*2] = [mb, 2, 4, 4] |
INDArray |
NDLoss.sparseSoftmaxCrossEntropy(INDArray logits,
INDArray labels)
As per softmaxCrossEntropy(String, SDVariable, SDVariable, LossReduce) but the labels variable
is represented as an integer array instead of the equivalent one-hot array. i.e., if logits are rank N, then labels have rank N-1 |
INDArray |
NDBase.sparseToDense(INDArray indices,
INDArray shape,
INDArray values)
Create a dense matrix equivalent of a sparse matrix based on the given input.
|
INDArray |
NDBase.sparseToDense(INDArray indices,
INDArray shape,
INDArray values,
INDArray defaultValue)
Create a dense matrix equivalent of a sparse matrix based on the given input.
|
INDArray[] |
NDBase.split(INDArray input,
INDArray numSplit,
int splitDim)
Split a value in to a list of ndarrays.
|
INDArray[] |
NDBase.split(INDArray input,
int numSplit,
int splitDim)
Split a value in to a list of ndarrays.
|
INDArray[] |
NDBase.splitV(INDArray input,
INDArray sizes,
int numSplit,
int splitDim)
Split a value in to a list of ndarrays with varying sizes
according to the sizes parameter. |
INDArray |
NDMath.sqrt(INDArray x)
Element-wise square root function: out = sqrt(x)
|
INDArray |
NDMath.square(INDArray x)
Element-wise square function: out = x^2
|
INDArray |
NDMath.squaredDifference(INDArray x,
INDArray y)
Pairwise squared difference operation.
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.squaredNorm(INDArray x,
boolean keepDims,
int... dimensions)
Squared L2 norm: see norm2(String, SDVariable, boolean, int...)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.squaredNorm(INDArray in,
boolean keepDims,
int... dimensions)
Sum of squared differences.
|
INDArray |
NDMath.squaredNorm(INDArray in,
INDArray dimensions)
Sum of squared differences.
|
INDArray |
NDMath.squaredNorm(INDArray in,
INDArray dimensions,
boolean keepDims)
Sum of squared differences.
|
INDArray |
NDBase.squaredNorm(INDArray x,
int... dimensions)
Squared L2 norm: see norm2(String, SDVariable, boolean, int...)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.squaredNorm(INDArray in,
int... dimensions)
Sum of squared differences.
|
INDArray |
NDBase.squeeze(INDArray x,
int axis)
Remove a single dimension of size 1.
For example, if input has shape [a,b,1,c] then squeeze(input, 2) returns an array of shape [a,b,c] |
INDArray |
NDRNN.sru(INDArray x,
INDArray initialC,
INDArray mask,
SRUWeights SRUWeights)
The SRU layer.
|
INDArray |
NDRNN.sru(INDArray x,
INDArray initialC,
SRUWeights SRUWeights)
The SRU layer.
|
INDArray |
NDRNN.sruCell(INDArray x,
INDArray cLast,
SRUWeights SRUWeights)
The SRU layer.
|
INDArray |
NDBase.stack(int axis,
INDArray... values)
Stack a set of N INDArray of rank X into one rank X+1 variable.
If inputs have shape [a,b,c] then output has shape: axis = 0: [N,a,b,c] axis = 1: [a,N,b,c] axis = 2: [a,b,N,c] axis = 3: [a,b,c,N] see unstack(String[], SDVariable, int, int) |
INDArray |
NDBase.standardDeviation(INDArray x,
boolean biasCorrected,
boolean keepDims,
int... dimensions)
Standard deviation array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.standardDeviation(INDArray x,
boolean biasCorrected,
int... dimensions)
Standard deviation array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.standardize(INDArray x,
int... dimensions)
Standardize input variable along given axis
|
INDArray |
NDMath.step(INDArray x,
double value)
Elementwise step function:
out(x) = 1 if x >= cutoff out(x) = 0 otherwise |
INDArray |
NDBase.stridedSlice(INDArray in,
INDArray begin,
INDArray end,
INDArray strides)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDBase.stridedSlice(INDArray in,
INDArray begin,
INDArray end,
INDArray strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDBase.stridedSlice(INDArray in,
long[] begin,
long[] end,
long... strides)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDBase.stridedSlice(INDArray in,
long[] begin,
long[] end,
long[] strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDMath.sub(INDArray x,
double value)
Scalar subtraction operation, out = in - scalar
|
INDArray |
NDMath.sub(INDArray x,
INDArray y)
Pairwise subtraction operation, out = x - y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.sum(INDArray x,
boolean keepDims,
int... dimensions)
Sum array reduction operation, optionally along specified dimensions.
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.sum(INDArray in,
boolean keepDims,
int... dimensions)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDMath.sum(INDArray in,
INDArray dimensions)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDMath.sum(INDArray in,
INDArray dimensions,
boolean keepDims)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDBase.sum(INDArray x,
int... dimensions)
Sum array reduction operation, optionally along specified dimensions.
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.sum(INDArray in,
int... dimensions)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDLinalg.svd(INDArray input,
boolean fullUV,
boolean computeUV)
Calculates singular value decomposition.
|
INDArray |
NDLinalg.svd(INDArray input,
boolean fullUV,
boolean computeUV,
int switchNum)
Calculates singular value decomposition.
|
INDArray |
NDNN.swish(INDArray x)
Element-wise "swish" function: out = x * sigmoid(b*x) with b=1.0
See: https://arxiv.org/abs/1710.05941 |
INDArray[] |
NDBase.switchOp(INDArray x,
INDArray predicate)
Switch operation
Predicate - if false, values are output to left (first) branch/output; if true, to right (second) branch/output |
INDArray |
NDMath.tan(INDArray x)
Elementwise tangent operation: out = tan(x)
|
INDArray |
NDNN.tanh(INDArray x)
Elementwise tanh (hyperbolic tangent) operation: out = tanh(x)
|
INDArray |
NDMath.tanh(INDArray x)
Elementwise tanh (hyperbolic tangent) operation: out = tanh(x)
|
INDArray |
NDBase.tensorMmul(INDArray x,
INDArray y,
int[] dimensionsX,
int... dimensionsY)
//TODO: Ops must be documented.
|
INDArray |
NDBase.tensorMmul(INDArray x,
INDArray y,
int[] dimensionsX,
int[] dimensionsY,
boolean transposeX,
boolean transposeY,
boolean transposeZ)
//TODO: Ops must be documented.
|
INDArray |
NDBase.tile(INDArray x,
INDArray repeat)
Repeat (tile) the input tensor the specified number of times.
For example, if input is [1, 2] [3, 4] and repeat is [2, 3] then output is [1, 2, 1, 2, 1, 2] [3, 4, 3, 4, 3, 4] [1, 2, 1, 2, 1, 2] [3, 4, 3, 4, 3, 4] |
INDArray |
NDBase.tile(INDArray x,
int... repeat)
see tile(String, SDVariable, int...)
|
INDArray |
NDMath.trace(INDArray in)
Matrix trace operation
For rank 2 matrices, the output is a scalar vith the trace - i.e., sum of the main diagonal. For higher rank inputs, output[a,b,c] = trace(in[a,b,c,:,:]) |
INDArray |
NDBase.transpose(INDArray x)
Matrix transpose operation: If input has shape [a,b] output has shape [b,a]
|
INDArray |
NDLinalg.tri(DataType dataType,
int row,
int column,
int diagonal)
An array with ones at and below the given diagonal and zeros elsewhere.
|
INDArray |
NDLinalg.tri(int row,
int column)
An array with ones at and below the given diagonal and zeros elsewhere.
|
INDArray |
NDLinalg.triangularSolve(INDArray matrix,
INDArray rhs,
boolean lower,
boolean adjoint)
Solver for systems of linear questions.
|
INDArray |
NDLinalg.triu(INDArray input)
Upper triangle of an array.
|
INDArray |
NDLinalg.triu(INDArray input,
int diag)
Upper triangle of an array.
|
INDArray |
NDRandom.uniform(double min,
double max,
DataType datatype,
long... shape)
Generate a new random INDArray, where values are randomly sampled according to a uniform distribution,
U(min,max) |
INDArray |
NDBase.unsortedSegmentMax(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment max operation.
|
INDArray |
NDBase.unsortedSegmentMean(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment mean operation.
|
INDArray |
NDBase.unsortedSegmentMin(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment min operation.
|
INDArray |
NDBase.unsortedSegmentProd(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment product operation.
|
INDArray |
NDBase.unsortedSegmentSqrtN(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment sqrtN operation.
|
INDArray |
NDBase.unsortedSegmentSum(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment sum operation.
|
INDArray[] |
NDBase.unstack(INDArray value,
int axis,
int num)
Unstack a variable of rank X into N rank X-1 variables by taking slices along the specified axis.
If input has shape [a,b,c] then output has shape: axis = 0: [b,c] axis = 1: [a,c] axis = 2: [a,b] |
INDArray |
NDCNN.upsampling2d(INDArray input,
int scale)
Upsampling layer for 2D inputs.
scale is used for both height and width dimensions. |
INDArray |
NDCNN.upsampling2d(INDArray input,
int scaleH,
int scaleW,
boolean nchw)
2D Convolution layer operation - Upsampling 2d
|
INDArray |
NDCNN.upsampling3d(INDArray input,
boolean ncdhw,
int scaleD,
int scaleH,
int scaleW)
3D Convolution layer operation - Upsampling 3d
|
INDArray |
NDBase.variance(INDArray x,
boolean biasCorrected,
boolean keepDims,
int... dimensions)
Variance array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.variance(INDArray x,
boolean biasCorrected,
int... dimensions)
Variance array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.where(INDArray condition)
Returns elements that are true from the given condition array
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.where(INDArray x,
INDArray condition)
Similar to numpy where, takes elements from x or y depending on whether the condition at a given element is true or false
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.where(INDArray x,
INDArray y,
INDArray condition)
Similar to numpy where, takes elements from x or y depending on whether the condition at a given element is true or false
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.whereNumpy(INDArray x,
INDArray y,
INDArray condition)
As implemented in numpy, Return elements chosen from x or y depending on condition.
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBitwise.xor(INDArray x,
INDArray y)
Bitwise XOR operation (exclusive OR).
|
INDArray |
NDMath.xor(INDArray x,
INDArray y)
Boolean XOR (exclusive OR) operation: elementwise (x != 0) XOR (y != 0)
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. Returns an array with values 1 where condition is satisfied, or value 0 otherwise. |
INDArray |
NDImage.yiqToRgb(INDArray input)
Converting image from YIQ to RGB format
|
INDArray |
NDImage.yuvToRgb(INDArray input)
Converting image from YUV to RGB format
|
INDArray |
NDMath.zeroFraction(INDArray input)
Full array zero fraction array reduction operation, optionally along specified dimensions: out = (count(x == 0) / length(x))
|
INDArray |
NDBase.zerosLike(INDArray input)
Return a variable of all 0s, with the same shape as the input variable.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
NDMath.abs(INDArray x)
Elementwise absolute value operation: out = abs(x)
|
INDArray |
NDLoss.absoluteDifference(INDArray label,
INDArray predictions,
INDArray weights)
Absolute difference loss:
sum_i abs( label[i] - predictions[i] ) |
INDArray |
NDLoss.absoluteDifference(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Absolute difference loss:
sum_i abs( label[i] - predictions[i] ) |
INDArray |
NDMath.acos(INDArray x)
Elementwise acos (arccosine, inverse cosine) operation: out = arccos(x)
|
INDArray |
NDMath.acosh(INDArray x)
Elementwise acosh (inverse hyperbolic cosine) function: out = acosh(x)
|
INDArray |
NDMath.add(INDArray x,
double value)
Scalar add operation, out = in + scalar
|
INDArray |
NDMath.add(INDArray x,
INDArray y)
Pairwise addition operation, out = x + y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDImage.adjustContrast(INDArray in,
double factor)
Adjusts contrast of RGB or grayscale images.
|
INDArray |
NDImage.adjustHue(INDArray in,
double delta)
Adjust hue of RGB image
|
INDArray |
NDImage.adjustSaturation(INDArray in,
double factor)
Adjust saturation of RGB images
|
INDArray |
NDBase.all(INDArray x,
int... dimensions)
Boolean and array reduction operation, optionally along specified dimensions
|
INDArray |
NDBitwise.and(INDArray x,
INDArray y)
Bitwise AND operation.
|
INDArray |
NDMath.and(INDArray x,
INDArray y)
Boolean AND operation: elementwise (x != 0) && (y != 0)
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. Returns an array with values 1 where condition is satisfied, or value 0 otherwise. |
INDArray |
NDBase.any(INDArray x,
int... dimensions)
Boolean or array reduction operation, optionally along specified dimensions
|
INDArray |
NDBase.argmax(INDArray in,
boolean keepDims,
int... dimensions)
Argmax array reduction operation, optionally along specified dimensions.
Output values are the index of the maximum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.argmax(INDArray in,
int... dimensions)
Argmax array reduction operation, optionally along specified dimensions.
Output values are the index of the maximum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.argmin(INDArray in,
boolean keepDims,
int... dimensions)
Argmin array reduction operation, optionally along specified dimensions.
Output values are the index of the minimum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.argmin(INDArray in,
int... dimensions)
Argmin array reduction operation, optionally along specified dimensions.
Output values are the index of the minimum value of each slice along the specified dimension. Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.asin(INDArray x)
Elementwise asin (arcsin, inverse sine) operation: out = arcsin(x)
|
INDArray |
NDMath.asinh(INDArray x)
Elementwise asinh (inverse hyperbolic sine) function: out = asinh(x)
|
INDArray |
NDBase.assign(INDArray x,
INDArray y)
Assign the contents of y to x.
Y must be broadcastable to x or the same shape. |
INDArray |
NDMath.asum(INDArray in,
boolean keepDims,
int... dimensions)
Absolute sum array reduction operation, optionally along specified dimensions: out = sum(abs(x))
|
INDArray |
NDMath.asum(INDArray in,
int... dimensions)
Absolute sum array reduction operation, optionally along specified dimensions: out = sum(abs(x))
|
INDArray |
NDMath.atan(INDArray x)
Elementwise atan (arctangent, inverse tangent) operation: out = arctangent(x)
|
INDArray |
NDMath.atan2(INDArray y,
INDArray x)
Elementwise atan (arctangent, inverse tangent) operation: out = atan2(x,y).
Similar to atan(y/x) but sigts of x and y are used to determine the location of the result |
INDArray |
NDMath.atanh(INDArray x)
Elementwise atanh (inverse hyperbolic tangent) function: out = atanh(x)
|
INDArray |
NDCNN.avgPooling2d(INDArray input,
Pooling2DConfig Pooling2DConfig)
2D Convolution layer operation - average pooling 2d
|
INDArray |
NDCNN.avgPooling3d(INDArray input,
Pooling3DConfig Pooling3DConfig)
3D convolution layer operation - average pooling 3d
|
INDArray[] |
NDBase.batchMmul(INDArray[] inputsA,
INDArray... inputsB)
Matrix multiply a batch of matrices.
|
INDArray[] |
NDBase.batchMmul(INDArray[] inputsA,
INDArray... inputsB)
Matrix multiply a batch of matrices.
|
INDArray[] |
NDBase.batchMmul(INDArray[] inputsA,
INDArray[] inputsB,
boolean transposeA,
boolean transposeB)
Matrix multiply a batch of matrices.
|
INDArray[] |
NDBase.batchMmul(INDArray[] inputsA,
INDArray[] inputsB,
boolean transposeA,
boolean transposeB)
Matrix multiply a batch of matrices.
|
INDArray |
NDNN.batchNorm(INDArray input,
INDArray mean,
INDArray variance,
INDArray gamma,
INDArray beta,
double epsilon,
int... axis)
|
INDArray |
NDCNN.batchToSpace(INDArray x,
int[] blocks,
int[] croppingTop,
int... croppingBottom)
Convolution 2d layer batch to space operation on 4d input.
Reduces input batch dimension by rearranging data into a larger spatial dimensions |
INDArray |
NDNN.biasAdd(INDArray input,
INDArray bias,
boolean nchw)
Bias addition operation: a special case of addition, typically used with CNN 4D activations and a 1D bias vector
|
INDArray |
NDBitwise.bitRotl(INDArray x,
INDArray shift)
Roll integer bits to the left, i.e.
|
INDArray |
NDBitwise.bitRotr(INDArray x,
INDArray shift)
Roll integer bits to the right, i.e.
|
INDArray |
NDBitwise.bitsHammingDistance(INDArray x,
INDArray y)
Bitwise Hamming distance reduction over all elements of both input arrays.
For example, if x=01100000 and y=1010000 then the bitwise Hamming distance is 2 (due to differences at positions 0 and 1) Inputs must satisfy the following constraints: Must be same types: isSameType(x, y) |
INDArray |
NDBitwise.bitShift(INDArray x,
INDArray shift)
Shift integer bits to the left, i.e.
|
INDArray |
NDMath.bitShift(INDArray x,
INDArray shift)
Bit shift operation
|
INDArray |
NDBitwise.bitShiftRight(INDArray x,
INDArray shift)
Shift integer bits to the right, i.e.
|
INDArray |
NDMath.bitShiftRight(INDArray x,
INDArray shift)
Right bit shift operation
|
INDArray |
NDMath.bitShiftRotl(INDArray x,
INDArray shift)
Cyclic bit shift operation
|
INDArray |
NDMath.bitShiftRotr(INDArray x,
INDArray shift)
Cyclic right shift operation
|
INDArray |
NDBase.castTo(INDArray arg,
DataType datatype)
Cast the array to a new datatype - for example, Integer -> Float
|
INDArray |
NDMath.ceil(INDArray x)
Element-wise ceiling function: out = ceil(x).
Rounds each value up to the nearest integer value (if not already an integer) |
INDArray |
NDLinalg.cholesky(INDArray input)
Computes the Cholesky decomposition of one or more square matrices.
|
INDArray |
NDMath.clipByAvgNorm(INDArray x,
double clipValue,
int... dimensions)
Clips tensor values to a maximum average L2-norm.
|
INDArray |
NDBase.clipByNorm(INDArray x,
double clipValue)
Returns a clipped ndarray such that the input is normalized so that its L2 norm
is <= the specified value. |
INDArray |
NDMath.clipByNorm(INDArray x,
double clipValue,
int... dimensions)
Clipping by L2 norm, optionally along dimension(s)
if l2Norm(x,dimension) < clipValue, then input is returned unmodifed Otherwise, out[i] = in[i] * clipValue / l2Norm(in, dimensions) where each value is clipped according to the corresponding l2Norm along the specified dimensions |
INDArray |
NDBase.clipByNorm(INDArray x,
INDArray clipValue,
INDArray dimensions)
Returns a clipped ndarray such that the input is normalized so that its L2 norm
is <= the specified value. |
INDArray |
NDBase.clipByValue(INDArray x,
double clipValueMin,
double clipValueMax)
Return the clipped ndarray containing values no smaller or larger than the given min and max.
|
INDArray |
NDMath.clipByValue(INDArray x,
double clipValueMin,
double clipValueMax)
Element-wise clipping function:
out[i] = in[i] if in[i] >= clipValueMin and in[i] <= clipValueMax out[i] = clipValueMin if in[i] < clipValueMin out[i] = clipValueMax if in[i] > clipValueMax |
INDArray |
NDBase.clipByValue(INDArray x,
INDArray clipValueMin,
INDArray clipValueMax)
Return the clipped ndarray containing values no smaller or larger than the given min and max.
|
INDArray |
NDCNN.col2Im(INDArray in,
Conv2DConfig Conv2DConfig)
col2im operation for use in 2D convolution operations.
|
INDArray |
NDBase.concat(int dimension,
INDArray... inputs)
Concatenate a set of inputs along the specified dimension.
Note that inputs must have identical rank and identical dimensions, other than the dimension to stack on. For example, if 2 inputs have shape [a, x, c] and [a, y, c] and dimension = 1, then the output has shape [a, x+y, c] Inputs must satisfy the following constraints: Input arrays must all be the same datatype: isSameType(inputs) |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
DataType dataType)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
INDArray weights)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
INDArray weights,
int numClasses)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. For example, if labels = [0, 1, 1], predicted = [0, 2, 1], numClasses = 4, and weights = [1, 2, 3] [1, 0, 0, 0] [0, 3, 2, 0] [0, 0, 0, 0] [0, 0, 0, 0] |
INDArray |
NDMath.confusionMatrix(INDArray labels,
INDArray pred,
int numClasses)
Compute the 2d confusion matrix of size [numClasses, numClasses] from a pair of labels and predictions, both of
which are represented as integer values. For example, if labels = [0, 1, 1], predicted = [0, 2, 1], and numClasses=4 then output is: [1, 0, 0, 0] [0, 1, 1, 0] [0, 0, 0, 0] [0, 0, 0, 0] |
INDArray |
NDCNN.conv1d(INDArray input,
INDArray weights,
Conv1DConfig Conv1DConfig)
Conv1d operation.
|
INDArray |
NDCNN.conv1d(INDArray input,
INDArray weights,
INDArray bias,
Conv1DConfig Conv1DConfig)
Conv1d operation.
|
INDArray |
NDCNN.conv2d(INDArray layerInput,
INDArray weights,
Conv2DConfig Conv2DConfig)
2D Convolution operation with optional bias
|
INDArray |
NDCNN.conv2d(INDArray layerInput,
INDArray weights,
INDArray bias,
Conv2DConfig Conv2DConfig)
2D Convolution operation with optional bias
|
INDArray |
NDCNN.conv3d(INDArray input,
INDArray weights,
Conv3DConfig Conv3DConfig)
Convolution 3D operation with optional bias
|
INDArray |
NDCNN.conv3d(INDArray input,
INDArray weights,
INDArray bias,
Conv3DConfig Conv3DConfig)
Convolution 3D operation with optional bias
|
INDArray |
NDMath.cos(INDArray x)
Elementwise cosine operation: out = cos(x)
|
INDArray |
NDMath.cosh(INDArray x)
Elementwise cosh (hyperbolic cosine) operation: out = cosh(x)
|
INDArray |
NDMath.cosineDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Cosine distance reduction operation.
|
INDArray |
NDLoss.cosineDistance(INDArray label,
INDArray predictions,
INDArray weights,
int dimension)
Cosine distance loss:
1 - cosineSimilarity(x,y) or 1 - sum_i label[i] * prediction[i], which isequivalent to cosine distance when both the predictions and labels are normalized. Note: This loss function assumes that both the predictions and labels are normalized to have unit l2 norm. If this is not the case, you should normalize them first by dividing by norm2(String, SDVariable, boolean, int...) along the cosine distance dimension (with keepDims=true). |
INDArray |
NDLoss.cosineDistance(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
int dimension)
Cosine distance loss:
1 - cosineSimilarity(x,y) or 1 - sum_i label[i] * prediction[i], which isequivalent to cosine distance when both the predictions and labels are normalized. Note: This loss function assumes that both the predictions and labels are normalized to have unit l2 norm. If this is not the case, you should normalize them first by dividing by norm2(String, SDVariable, boolean, int...) along the cosine distance dimension (with keepDims=true). |
INDArray |
NDMath.cosineDistance(INDArray x,
INDArray y,
int... dimensions)
Cosine distance reduction operation.
|
INDArray |
NDMath.cosineSimilarity(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Cosine similarity pairwise reduction operation.
|
INDArray |
NDMath.cosineSimilarity(INDArray x,
INDArray y,
int... dimensions)
Cosine similarity pairwise reduction operation.
|
INDArray |
NDMath.countNonZero(INDArray in,
boolean keepDims,
int... dimensions)
Count non zero array reduction operation, optionally along specified dimensions: out = count(x != 0)
|
INDArray |
NDMath.countNonZero(INDArray in,
int... dimensions)
Count non zero array reduction operation, optionally along specified dimensions: out = count(x != 0)
|
INDArray |
NDMath.countZero(INDArray in,
boolean keepDims,
int... dimensions)
Count zero array reduction operation, optionally along specified dimensions: out = count(x == 0)
|
INDArray |
NDMath.countZero(INDArray in,
int... dimensions)
Count zero array reduction operation, optionally along specified dimensions: out = count(x == 0)
|
INDArray |
NDBase.create(INDArray shape,
DataType dataType)
Return a newly created variable, with the specified shape and data type.
|
INDArray |
NDBase.create(INDArray shape,
DataType dataType,
String order,
boolean initialize)
Return a newly created variable, with the specified shape and data type.
|
INDArray |
NDNN.cReLU(INDArray x)
Concatenates a ReLU which selects only the positive part of the activation with a ReLU which selects only the negative part of the activation.
|
INDArray |
NDImage.cropAndResize(INDArray image,
INDArray cropBoxes,
INDArray boxIndices,
INDArray cropOutSize)
Given an input image and some crop boxes, extract out the image subsets and resize them to the specified size.
|
INDArray |
NDImage.cropAndResize(INDArray image,
INDArray cropBoxes,
INDArray boxIndices,
INDArray cropOutSize,
double extrapolationValue)
Given an input image and some crop boxes, extract out the image subsets and resize them to the specified size.
|
INDArray |
NDLinalg.cross(INDArray a,
INDArray b)
Computes pairwise cross product.
|
INDArray |
NDMath.cross(INDArray a,
INDArray b)
Returns the pair-wise cross product of equal size arrays a and b: a x b = ||a||x||b|| sin(theta).
Can take rank 1 or above inputs (of equal shapes), but note that the last dimension must have dimension 3 |
INDArray |
NDLoss.ctcLoss(INDArray targetLabels,
INDArray logitInput,
INDArray targetLabelLengths,
INDArray logitInputLengths)
CTC Loss: Connectionist Temporal Classification Loss.
|
INDArray |
NDMath.cube(INDArray x)
Element-wise cube function: out = x^3
|
INDArray |
NDBase.cumprod(INDArray in,
boolean exclusive,
boolean reverse,
int... axis)
Cumulative product operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a*b, a*b*c] exclusive=true, reverse=false, [0, a, a*b] exclusive=false, reverse=true: [a*b*c, b*c, c] exclusive=true, reverse=true: [b*c, c, 0] |
INDArray |
NDBase.cumprod(INDArray in,
int... axis)
Cumulative product operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a*b, a*b*c] exclusive=true, reverse=false, [0, a, a*b] exclusive=false, reverse=true: [a*b*c, b*c, c] exclusive=true, reverse=true: [b*c, c, 0] |
INDArray |
NDBase.cumsum(INDArray in,
boolean exclusive,
boolean reverse,
int... axis)
Cumulative sum operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a+b, a+b+c] exclusive=true, reverse=false, [0, a, a+b] exclusive=false, reverse=true: [a+b+c, b+c, c] exclusive=true, reverse=true: [b+c, c, 0] |
INDArray |
NDBase.cumsum(INDArray in,
int... axis)
Cumulative sum operation.
For input: [ a, b, c], output is: exclusive=false, reverse=false: [a, a+b, a+b+c] exclusive=true, reverse=false, [0, a, a+b] exclusive=false, reverse=true: [a+b+c, b+c, c] exclusive=true, reverse=true: [b+c, c, 0] |
INDArray |
NDCNN.deconv2d(INDArray layerInput,
INDArray weights,
DeConv2DConfig DeConv2DConfig)
2D deconvolution operation with optional bias
|
INDArray |
NDCNN.deconv2d(INDArray layerInput,
INDArray weights,
INDArray bias,
DeConv2DConfig DeConv2DConfig)
2D deconvolution operation with optional bias
|
INDArray |
NDCNN.deconv3d(INDArray input,
INDArray weights,
DeConv3DConfig DeConv3DConfig)
3D CNN deconvolution operation with or without optional bias
|
INDArray |
NDCNN.deconv3d(INDArray input,
INDArray weights,
INDArray bias,
DeConv3DConfig DeConv3DConfig)
3D CNN deconvolution operation with or without optional bias
|
INDArray |
NDCNN.depthToSpace(INDArray x,
int blockSize,
DataFormat dataFormat)
Convolution 2d layer batch to space operation on 4d input.
Reduces input channels dimension by rearranging data into a larger spatial dimensions Example: if input has shape [mb, 8, 2, 2] and block size is 2, then output size is [mb, 8/(2*2), 2*2, 2*2] = [mb, 2, 4, 4] |
INDArray |
NDCNN.depthWiseConv2d(INDArray layerInput,
INDArray depthWeights,
Conv2DConfig Conv2DConfig)
Depth-wise 2D convolution operation with optional bias
|
INDArray |
NDCNN.depthWiseConv2d(INDArray layerInput,
INDArray depthWeights,
INDArray bias,
Conv2DConfig Conv2DConfig)
Depth-wise 2D convolution operation with optional bias
|
INDArray |
NDLinalg.diag_part(INDArray input)
Calculates diagonal tensor.
|
INDArray |
NDLinalg.diag(INDArray input)
Calculates diagonal tensor.
|
INDArray |
NDMath.diag(INDArray x)
Returns an output variable with diagonal values equal to the specified values; off-diagonal values will be set to 0
For example, if input = [1,2,3], then output is given by: [ 1, 0, 0] [ 0, 2, 0] [ 0, 0, 3] Higher input ranks are also supported: if input has shape [a,...,R-1] then output[i,...,k,i,...,k] = input[i,...,k]. i.e., for input rank R, output has rank 2R |
INDArray |
NDMath.diagPart(INDArray x)
Extract the diagonal part from the input array.
If input is [ 1, 0, 0] [ 0, 2, 0] [ 0, 0, 3] then output is [1, 2, 3]. Supports higher dimensions: in general, out[i,...,k] = in[i,...,k,i,...,k] |
INDArray |
NDCNN.dilation2D(INDArray df,
INDArray weights,
int[] strides,
int[] rates,
boolean isSameMode)
TODO doc string
|
INDArray |
NDMath.div(INDArray x,
double value)
Scalar division operation, out = in / scalar
|
INDArray |
NDMath.div(INDArray x,
INDArray y)
Pairwise division operation, out = x / y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.dot(INDArray x,
INDArray y,
int... dimensions)
Pairwise dot product reduction along dimension
output = sum(i=0 ... |
INDArray |
NDNN.dotProductAttention(INDArray queries,
INDArray keys,
INDArray values,
INDArray mask,
boolean scaled)
This operation performs dot product attention on the given timeseries input with the given queries
out = sum(similarity(k_i, q) * v_i) similarity(k, q) = softmax(k * q) where x * q is the dot product of x and q Optionally with normalization step: similarity(k, q) = softmax(k * q / sqrt(size(q)) See also "Attention is all you need" (https://arxiv.org/abs/1706.03762, p. |
INDArray |
NDNN.dropout(INDArray input,
double inputRetainProbability)
Dropout operation
|
INDArray |
NDNN.dropoutInverted(INDArray input,
double p)
Dropout inverted operation.
|
INDArray[] |
NDBase.dynamicPartition(INDArray x,
INDArray partitions,
int numPartitions)
Dynamically partition the input variable values into the specified number of paritions, using the indices.
Example: |
INDArray |
NDBase.dynamicStitch(INDArray[] indices,
INDArray... x)
Dynamically merge the specified input arrays into a single array, using the specified indices
|
INDArray |
NDBase.dynamicStitch(INDArray[] indices,
INDArray... x)
Dynamically merge the specified input arrays into a single array, using the specified indices
|
INDArray |
NDNN.elu(INDArray x)
Element-wise exponential linear unit (ELU) function:
out = x if x > 0 out = a * (exp(x) - 1) if x <= 0 with constant a = 1.0 |
INDArray |
NDMath.embeddingLookup(INDArray x,
INDArray indices,
PartitionMode PartitionMode)
Looks up ids in a list of embedding tensors.
|
INDArray |
NDMath.entropy(INDArray in,
boolean keepDims,
int... dimensions)
Entropy reduction: -sum(x * log(x))
|
INDArray |
NDMath.entropy(INDArray in,
int... dimensions)
Entropy reduction: -sum(x * log(x))
|
INDArray |
NDBase.eq(INDArray x,
double y)
Equals operation: elementwise x == y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.eq(INDArray x,
INDArray y)
Equal to operation: elementwise x == y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDMath.erf(INDArray x)
Element-wise Gaussian error function - out = erf(in)
|
INDArray |
NDMath.erfc(INDArray x)
Element-wise complementary Gaussian error function - out = erfc(in) = 1 - erf(in)
|
INDArray |
NDMath.euclideanDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Euclidean distance (l2 norm, l2 distance) reduction operation.
|
INDArray |
NDMath.euclideanDistance(INDArray x,
INDArray y,
int... dimensions)
Euclidean distance (l2 norm, l2 distance) reduction operation.
|
INDArray |
NDMath.exp(INDArray x)
Elementwise exponent function: out = exp(x) = 2.71828...^x
|
INDArray |
NDBase.expandDims(INDArray x,
int axis)
Reshape the input by adding a 1 at the specified location.
For example, if input has shape [a, b], then output shape is: axis = 0: [1, a, b] axis = 1: [a, 1, b] axis = 2: [a, b, 1] |
INDArray |
NDMath.expm1(INDArray x)
Elementwise 1.0 - exponent function: out = 1.0 - exp(x) = 1.0 - 2.71828...^x
|
INDArray |
NDImage.extractImagePatches(INDArray image,
int[] kSizes,
int[] strides,
int[] rates,
boolean sameMode)
Given an input image, extract out image patches (of size kSizes - h x w) and place them in the depth dimension.
|
INDArray |
NDCNN.extractImagePatches(INDArray input,
int kH,
int kW,
int sH,
int sW,
int rH,
int rW,
boolean sameMode)
Extract image patches
|
INDArray |
NDMath.eye(INDArray rows)
As per eye(String, int) but with the number of rows specified as a scalar INDArray
|
INDArray |
NDMath.eye(INDArray rows,
INDArray cols)
As per eye(int, int) bit with the number of rows/columns specified as scalar INDArrays
|
INDArray |
NDBase.fill(INDArray shape,
DataType dataType,
double value)
Generate an output variable with the specified (dynamic) shape with all elements set to the specified value
|
INDArray |
NDMath.firstIndex(INDArray in,
Condition condition,
boolean keepDims,
int... dimensions)
First index reduction operation.
Returns a variable that contains the index of the first element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.firstIndex(INDArray in,
Condition condition,
int... dimensions)
First index reduction operation.
Returns a variable that contains the index of the first element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.floor(INDArray x)
Element-wise floor function: out = floor(x).
Rounds each value down to the nearest integer value (if not already an integer) |
INDArray |
NDMath.floorDiv(INDArray x,
INDArray y)
Pairwise floor division operation, out = floor(x / y)
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.floorMod(INDArray x,
double value)
Scalar floor modulus operation
|
INDArray |
NDMath.floorMod(INDArray x,
INDArray y)
Pairwise Modulus division operation
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.gather(INDArray df,
INDArray indices,
int axis)
Gather slices from the input variable where the indices are specified as dynamic array values.
Output shape is same as input shape, except for axis dimension, which has size equal to indices.length. |
INDArray |
NDBase.gather(INDArray df,
int[] indices,
int axis)
Gather slices from the input variable where the indices are specified as fixed int[] values.
Output shape is same as input shape, except for axis dimension, which has size equal to indices.length. |
INDArray |
NDBase.gatherNd(INDArray df,
INDArray indices)
Gather slices from df with shape specified by indices.
|
INDArray |
NDNN.gelu(INDArray x)
GELU activation function - Gaussian Error Linear Units
For more details, see Gaussian Error Linear Units (GELUs) - https://arxiv.org/abs/1606.08415 This method uses the sigmoid approximation |
INDArray |
NDRNN.gru(INDArray x,
INDArray hLast,
INDArray Wx,
INDArray Wh,
INDArray biases)
The GRU operation.
|
INDArray[] |
NDRNN.gruCell(INDArray x,
INDArray hLast,
GRUWeights GRUWeights)
The GRU cell.
|
INDArray |
NDBase.gt(INDArray x,
double y)
Greater than operation: elementwise x > y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.gt(INDArray x,
INDArray y)
Greater than operation: elementwise x > y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.gte(INDArray x,
double y)
Greater than or equals operation: elementwise x >= y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.gte(INDArray x,
INDArray y)
Greater than or equal to operation: elementwise x >= y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDMath.hammingDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Hamming distance reduction operation.
|
INDArray |
NDMath.hammingDistance(INDArray x,
INDArray y,
int... dimensions)
Hamming distance reduction operation.
|
INDArray |
NDNN.hardSigmoid(INDArray x)
Element-wise hard sigmoid function:
out[i] = 0 if in[i] <= -2.5 out[1] = 0.2*in[i]+0.5 if -2.5 < in[i] < 2.5 out[i] = 1 if in[i] >= 2.5 |
INDArray |
NDNN.hardTanh(INDArray x)
Element-wise hard tanh function:
out[i] = -1 if in[i] <= -1 out[1] = in[i] if -1 < in[i] < 1 out[i] = 1 if in[i] >= 1 |
INDArray |
NDNN.hardTanhDerivative(INDArray x)
Derivative (dOut/dIn) of the element-wise hard Tanh function - hardTanh(INDArray)
|
INDArray |
NDLoss.hingeLoss(INDArray label,
INDArray predictions,
INDArray weights)
Hinge loss: a loss function used for training classifiers.
Implements L = max(0, 1 - t * predictions) where t is the label values after internally converting to {-1,1}from the user specified {0,1}. |
INDArray |
NDLoss.hingeLoss(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Hinge loss: a loss function used for training classifiers.
Implements L = max(0, 1 - t * predictions) where t is the label values after internally converting to {-1,1}from the user specified {0,1}. |
INDArray |
NDImage.hsvToRgb(INDArray input)
Converting image from HSV to RGB format
|
INDArray |
NDLoss.huberLoss(INDArray label,
INDArray predictions,
INDArray weights,
double delta)
Huber loss function, used for robust regression.
|
INDArray |
NDLoss.huberLoss(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double delta)
Huber loss function, used for robust regression.
|
INDArray |
NDMath.iamax(INDArray in,
boolean keepDims,
int... dimensions)
Index of the max absolute value: argmax(abs(in))
see argmax(String, INDArray, boolean, int...) |
INDArray |
NDMath.iamax(INDArray in,
int... dimensions)
Index of the max absolute value: argmax(abs(in))
see argmax(String, INDArray, boolean, int...) |
INDArray |
NDMath.iamin(INDArray in,
boolean keepDims,
int... dimensions)
Index of the min absolute value: argmin(abs(in))
see argmin(String, INDArray, boolean, int...) |
INDArray |
NDMath.iamin(INDArray in,
int... dimensions)
Index of the min absolute value: argmin(abs(in))
see argmin(String, INDArray, boolean, int...) |
INDArray |
NDBase.identity(INDArray input)
Elementwise identity operation: out = x
|
INDArray |
NDCNN.im2Col(INDArray in,
Conv2DConfig Conv2DConfig)
im2col operation for use in 2D convolution operations.
|
INDArray |
NDImage.imageResize(INDArray input,
INDArray size,
boolean preserveAspectRatio,
boolean antialias,
ImageResizeMethod ImageResizeMethod)
Resize images to size using the specified method.
|
INDArray |
NDImage.imageResize(INDArray input,
INDArray size,
ImageResizeMethod ImageResizeMethod)
Resize images to size using the specified method.
|
INDArray |
NDBase.invertPermutation(INDArray input)
Compute the inverse permutation indices for a permutation operation
Example: if input is [2, 0, 1] then output is [1, 2, 0] The idea is that x.permute(input).permute(invertPermutation(input)) == x |
INDArray |
NDMath.isFinite(INDArray x)
Is finite operation: elementwise isFinite(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isInfinite(INDArray x)
Is infinite operation: elementwise isInfinite(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isMax(INDArray x)
Is maximum operation: elementwise x == max(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isNaN(INDArray x)
Is Not a Number operation: elementwise isNaN(x)
Returns an array with the same shape/size as the input, with values 1 where condition is satisfied, or value 0 otherwise |
INDArray |
NDMath.isNonDecreasing(INDArray x)
Is the array non decreasing?
An array is non-decreasing if for every valid i, x[i] <= x[i+1]. |
INDArray |
NDBase.isNumericTensor(INDArray x)
Is the director a numeric tensor? In the current version of ND4J/SameDiff, this always returns true/1
|
INDArray |
NDMath.isStrictlyIncreasing(INDArray x)
Is the array strictly increasing?
An array is strictly increasing if for every valid i, x[i] < x[i+1]. |
INDArray |
NDMath.jaccardDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Jaccard similarity reduction operation.
|
INDArray |
NDMath.jaccardDistance(INDArray x,
INDArray y,
int... dimensions)
Jaccard similarity reduction operation.
|
INDArray |
NDLoss.l2Loss(INDArray var)
L2 loss: 1/2 * sum(x^2)
|
INDArray |
NDMath.lastIndex(INDArray in,
Condition condition,
boolean keepDims,
int... dimensions)
Last index reduction operation.
Returns a variable that contains the index of the last element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.lastIndex(INDArray in,
Condition condition,
int... dimensions)
Last index reduction operation.
Returns a variable that contains the index of the last element that matches the specified condition (for each slice along the specified dimensions) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDNN.layerNorm(INDArray input,
INDArray gain,
boolean channelsFirst,
int... dimensions)
Apply Layer Normalization
y = gain * standardize(x) + bias |
INDArray |
NDNN.layerNorm(INDArray input,
INDArray gain,
INDArray bias,
boolean channelsFirst,
int... dimensions)
Apply Layer Normalization
y = gain * standardize(x) + bias |
INDArray |
NDNN.leakyRelu(INDArray x,
double alpha)
Element-wise leaky ReLU function:
out = x if x >= 0.0 out = alpha * x if x < cutoff Alpha value is most commonly set to 0.01 |
INDArray |
NDNN.leakyReluDerivative(INDArray x,
double alpha)
Leaky ReLU derivative: dOut/dIn given input.
|
INDArray |
NDBitwise.leftShift(INDArray x,
INDArray y)
Bitwise left shift operation.
|
INDArray |
NDBitwise.leftShiftCyclic(INDArray x,
INDArray y)
Bitwise left cyclical shift operation.
|
INDArray |
NDNN.linear(INDArray input,
INDArray weights,
INDArray bias)
Linear layer operation: out = mmul(in,w) + bias
Note that bias array is optional |
INDArray |
NDBase.linspace(INDArray start,
INDArray stop,
INDArray number,
DataType dataType)
Create a new 1d array with values evenly spaced between values 'start' and 'stop'
For example, linspace(start=3.0, stop=4.0, number=3) will generate [3.0, 3.5, 4.0] |
INDArray[] |
NDMath.listDiff(INDArray x,
INDArray y)
Calculates difference between inputs X and Y.
|
INDArray |
NDCNN.localResponseNormalization(INDArray input,
LocalResponseNormalizationConfig LocalResponseNormalizationConfig)
2D convolution layer operation - local response normalization
|
INDArray |
NDMath.log(INDArray x)
Element-wise logarithm function (base e - natural logarithm): out = log(x)
|
INDArray |
NDMath.log(INDArray x,
double base)
Element-wise logarithm function (with specified base): out = log_{base}(x)
|
INDArray |
NDMath.log1p(INDArray x)
Elementwise natural logarithm function: out = log_e (1 + x)
|
INDArray |
NDLinalg.logdet(INDArray input)
Calculates log of determinant.
|
INDArray |
NDMath.logEntropy(INDArray in,
boolean keepDims,
int... dimensions)
Log entropy reduction: log(-sum(x * log(x)))
|
INDArray |
NDMath.logEntropy(INDArray in,
int... dimensions)
Log entropy reduction: log(-sum(x * log(x)))
|
INDArray |
NDLoss.logLoss(INDArray label,
INDArray predictions)
Log loss, i.e., binary cross entropy loss, usually used for binary multi-label classification.
|
INDArray |
NDLoss.logLoss(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
double epsilon)
Log loss, i.e., binary cross entropy loss, usually used for binary multi-label classification.
|
INDArray |
NDLoss.logPoisson(INDArray label,
INDArray predictions,
INDArray weights,
boolean full)
Log poisson loss: a loss function used for training classifiers.
Implements L = exp(c) - z * c where c is log(predictions) and z is labels. |
INDArray |
NDLoss.logPoisson(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce,
boolean full)
Log poisson loss: a loss function used for training classifiers.
Implements L = exp(c) - z * c where c is log(predictions) and z is labels. |
INDArray |
NDNN.logSigmoid(INDArray x)
Element-wise sigmoid function: out[i] = log(sigmoid(in[i]))
|
INDArray |
NDNN.logSoftmax(INDArray x)
Log softmax activation
|
INDArray |
NDNN.logSoftmax(INDArray x,
int dimension)
Log softmax activation
|
INDArray |
NDMath.logSumExp(INDArray input,
int... dimensions)
Log-sum-exp reduction (optionally along dimension).
Computes log(sum(exp(x)) |
INDArray |
NDRNN.lstmblock(INDArray maxTSLength,
INDArray x,
INDArray cLast,
INDArray yLast,
LSTMWeights LSTMWeights,
LSTMConfiguration LSTMConfiguration)
The LSTM block
|
INDArray |
NDRNN.lstmblock(INDArray x,
LSTMWeights LSTMWeights,
LSTMConfiguration LSTMConfiguration)
The LSTM block
|
INDArray[] |
NDRNN.lstmCell(INDArray x,
INDArray cLast,
INDArray yLast,
LSTMWeights LSTMWeights,
LSTMConfiguration LSTMConfiguration)
The LSTM cell.
|
INDArray[] |
NDRNN.lstmLayer(INDArray x,
INDArray cLast,
INDArray yLast,
INDArray maxTSLength,
LSTMLayerWeights LSTMLayerWeights,
LSTMLayerConfig LSTMLayerConfig)
Long Short-Term Memory layer - Hochreiter 1997.
SUPPORTS following data formats: for unidirectional: TNS: shapes [timeLength, numExamples, inOutSize] NST: shapes [numExamples, inOutSize, timeLength] NTS: shapes [numExamples, timeLength, inOutSize] for bidirectional: T2NS: shapes [timeLength, 2, numExamples, inOutSize] (for ONNX) SUPPORTS following direction modes: FWD: forward BWD: backward BIDIR_SUM: bidirectional sum BIDIR_CONCAT: bidirectional concat BIDIR_EXTRA_DIM: bidirectional extra output dim (in conjunction with format dataFormat - T2NS) You may use different gate configurations: specify gate/cell/out aplha/beta and numbers of activations for gate/cell/out described in activations enum ("RELU","SIGMOID","AFFINE","LEAKY_RELU","THRESHHOLD_RELU","SCALED_TAHN","HARD_SIGMOID","ELU","SOFTSIGN","SOFTPLUS") Also this layer supports MKLDNN (DNNL) and cuDNN acceleration |
INDArray[] |
NDRNN.lstmLayer(INDArray x,
LSTMLayerWeights LSTMLayerWeights,
LSTMLayerConfig LSTMLayerConfig)
Long Short-Term Memory layer - Hochreiter 1997.
SUPPORTS following data formats: for unidirectional: TNS: shapes [timeLength, numExamples, inOutSize] NST: shapes [numExamples, inOutSize, timeLength] NTS: shapes [numExamples, timeLength, inOutSize] for bidirectional: T2NS: shapes [timeLength, 2, numExamples, inOutSize] (for ONNX) SUPPORTS following direction modes: FWD: forward BWD: backward BIDIR_SUM: bidirectional sum BIDIR_CONCAT: bidirectional concat BIDIR_EXTRA_DIM: bidirectional extra output dim (in conjunction with format dataFormat - T2NS) You may use different gate configurations: specify gate/cell/out aplha/beta and numbers of activations for gate/cell/out described in activations enum ("RELU","SIGMOID","AFFINE","LEAKY_RELU","THRESHHOLD_RELU","SCALED_TAHN","HARD_SIGMOID","ELU","SOFTSIGN","SOFTPLUS") Also this layer supports MKLDNN (DNNL) and cuDNN acceleration |
INDArray |
NDLinalg.lstsq(INDArray matrix,
INDArray rhs,
double l2_reguralizer)
Solver for linear squares problems.
|
INDArray |
NDLinalg.lstsq(INDArray matrix,
INDArray rhs,
double l2_reguralizer,
boolean fast)
Solver for linear squares problems.
|
INDArray |
NDBase.lt(INDArray x,
double y)
Less than operation: elementwise x < y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.lt(INDArray x,
INDArray y)
Less than operation: elementwise x < y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.lte(INDArray x,
double y)
Less than or equals operation: elementwise x <= y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.lte(INDArray x,
INDArray y)
Less than or equal to operation: elementwise x <= y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDLinalg.lu(INDArray input)
Computes LU decomposition.
|
INDArray |
NDMath.manhattanDistance(INDArray x,
INDArray y,
boolean keepDims,
boolean isComplex,
int... dimensions)
Manhattan distance (l1 norm, l1 distance) reduction operation.
|
INDArray |
NDMath.manhattanDistance(INDArray x,
INDArray y,
int... dimensions)
Manhattan distance (l1 norm, l1 distance) reduction operation.
|
INDArray |
NDBase.matchCondition(INDArray in,
Condition condition)
Returns a boolean mask of equal shape to the input, where the condition is satisfied - value 1 where satisfied, 0 otherwise
|
INDArray |
NDBase.matchConditionCount(INDArray in,
Condition condition)
Returns a count of the number of elements that satisfy the condition
|
INDArray |
NDBase.matchConditionCount(INDArray in,
Condition condition,
boolean keepDim,
int... dimensions)
Returns a count of the number of elements that satisfy the condition (for each slice along the specified dimensions)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.matchConditionCount(INDArray in,
Condition condition,
int... dimensions)
Returns a count of the number of elements that satisfy the condition (for each slice along the specified dimensions)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDLinalg.matmul(INDArray a,
INDArray b)
Performs matrix mutiplication on input tensors.
|
INDArray[] |
NDLinalg.matrixBandPart(INDArray input,
int minLower,
int maxUpper)
Copy a tensor setting outside a central band in each innermost matrix.
|
INDArray |
NDMath.matrixDeterminant(INDArray in)
Matrix determinant op.
|
INDArray |
NDMath.matrixInverse(INDArray in)
Matrix inverse op.
|
INDArray |
NDBase.max(INDArray x,
boolean keepDims,
int... dimensions)
Max array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.max(INDArray first,
INDArray second)
Element-wise maximum operation: out[i] = max(first[i], second[i])
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.max(INDArray x,
INDArray y)
Pairwise max operation, out = max(x, y)
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.max(INDArray x,
int... dimensions)
Max array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDCNN.maxPooling2d(INDArray input,
Pooling2DConfig Pooling2DConfig)
2D Convolution layer operation - max pooling 2d
|
INDArray |
NDCNN.maxPooling3d(INDArray input,
Pooling3DConfig Pooling3DConfig)
3D convolution layer operation - max pooling 3d operation.
|
INDArray[] |
NDCNN.maxPoolWithArgmax(INDArray input,
Pooling2DConfig Pooling2DConfig)
2D Convolution layer operation - Max pooling on the input and outputs both max values and indices
|
INDArray |
NDBase.mean(INDArray x,
boolean keepDims,
int... dimensions)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
boolean keepDims,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.mean(INDArray x,
INDArray dimensions)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
INDArray dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.mean(INDArray x,
INDArray dimensions,
boolean keepDims)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
INDArray dimensions,
boolean keepDims)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.mean(INDArray x,
int... dimensions)
Mean (average) array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.mean(INDArray in,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDLoss.meanPairwiseSquaredError(INDArray label,
INDArray predictions,
INDArray weights)
Mean pairwise squared error.
MPWSE loss calculates the difference between pairs of consecutive elements in the predictions and labels arrays. For example, if predictions = [p0, p1, p2] and labels are [l0, l1, l2] then MPWSE is: [((p0-p1) - (l0-l1))^2 + ((p0-p2) - (l0-l2))^2 + ((p1-p2) - (l1-l2))^2] / 3 |
INDArray |
NDLoss.meanPairwiseSquaredError(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Mean pairwise squared error.
MPWSE loss calculates the difference between pairs of consecutive elements in the predictions and labels arrays. For example, if predictions = [p0, p1, p2] and labels are [l0, l1, l2] then MPWSE is: [((p0-p1) - (l0-l1))^2 + ((p0-p2) - (l0-l2))^2 + ((p1-p2) - (l1-l2))^2] / 3 |
INDArray |
NDLoss.meanSquaredError(INDArray label,
INDArray predictions,
INDArray weights)
Mean squared error loss function.
|
INDArray |
NDLoss.meanSquaredError(INDArray label,
INDArray predictions,
INDArray weights,
LossReduce lossReduce)
Mean squared error loss function.
|
INDArray |
NDBase.merge(INDArray x,
INDArray y)
The merge operation is a control operation that forwards the either of the inputs to the output, when
the first of them becomes available. |
INDArray |
NDMath.mergeAdd(INDArray... inputs)
Merge add function: merges an arbitrary number of equal shaped arrays using element-wise addition:
out = sum_i in[i] |
INDArray |
NDMath.mergeAvg(INDArray... inputs)
Merge average function: merges an arbitrary number of equal shaped arrays using element-wise mean operation:
out = mean_i in[i] |
INDArray |
NDMath.mergeMax(INDArray... inputs)
Merge max function: merges an arbitrary number of equal shaped arrays using element-wise maximum operation:
out = max_i in[i] |
INDArray |
NDMath.mergeMaxIndex(INDArray... x)
Return array of max elements indices with along tensor dimensions
|
INDArray |
NDMath.mergeMaxIndex(INDArray[] x,
DataType dataType)
Return array of max elements indices with along tensor dimensions
|
INDArray[] |
NDMath.meshgrid(INDArray[] inputs,
boolean cartesian)
Broadcasts parameters for evaluation on an N-D grid.
|
INDArray |
NDBase.min(INDArray x,
boolean keepDims,
int... dimensions)
Minimum array reduction operation, optionally along specified dimensions.
|
INDArray |
NDBase.min(INDArray first,
INDArray second)
Element-wise minimum operation: out[i] = min(first[i], second[i])
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.min(INDArray x,
INDArray y)
Pairwise max operation, out = min(x, y)
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.min(INDArray x,
int... dimensions)
Minimum array reduction operation, optionally along specified dimensions.
|
INDArray |
NDLinalg.mmul(INDArray x,
INDArray y)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDBase.mmul(INDArray x,
INDArray y)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDLinalg.mmul(INDArray x,
INDArray y,
boolean transposeX,
boolean transposeY,
boolean transposeZ)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDBase.mmul(INDArray x,
INDArray y,
boolean transposeX,
boolean transposeY,
boolean transposeZ)
Matrix multiplication: out = mmul(x,y)
Supports specifying transpose argument to perform operation such as mmul(a^T, b), etc. |
INDArray |
NDMath.mod(INDArray x,
INDArray y)
Pairwise modulus (remainder) operation, out = x % y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray[] |
NDMath.moments(INDArray input,
INDArray axes,
boolean keepDims)
Calculate the mean and (population) variance for the input variable, for the specified axis
|
INDArray[] |
NDMath.moments(INDArray input,
int[] axes,
boolean keepDims)
Calculate the mean and (population) variance for the input variable, for the specified axis
|
INDArray |
NDMath.mul(INDArray x,
double value)
Scalar multiplication operation, out = in * scalar
|
INDArray |
NDMath.mul(INDArray x,
INDArray y)
Pairwise multiplication operation, out = x * y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDNN.multiHeadDotProductAttention(INDArray queries,
INDArray keys,
INDArray values,
INDArray Wq,
INDArray Wk,
INDArray Wv,
INDArray Wo,
INDArray mask,
boolean scaled)
This performs multi-headed dot product attention on the given timeseries input
out = concat(head_1, head_2, ..., head_n) * Wo head_i = dot_product_attention(Wq_i*q, Wk_i*k, Wv_i*v) Optionally with normalization when calculating the attention for each head. See also "Attention is all you need" (https://arxiv.org/abs/1706.03762, pp. |
INDArray |
NDMath.neg(INDArray x)
Elementwise negative operation: out = -x
|
INDArray |
NDBase.neq(INDArray x,
double y)
Not equals operation: elementwise x != y
Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDBase.neq(INDArray x,
INDArray y)
Not equal to operation: elementwise x != y
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html Return boolean array with values true where satisfied, or false otherwise. |
INDArray |
NDImage.nonMaxSuppression(INDArray boxes,
INDArray scores,
int maxOutSize,
double iouThreshold,
double scoreThreshold)
Greedily selects a subset of bounding boxes in descending order of score
|
INDArray |
NDBase.norm1(INDArray x,
boolean keepDims,
int... dimensions)
Norm1 (L1 norm) reduction operation: The output contains the L1 norm for each tensor/subset along the specified dimensions:
out = sum_i abs(x[i]) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm1(INDArray in,
boolean keepDims,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.norm1(INDArray in,
INDArray dimensions)
Sum of absolute differences.
|
INDArray |
NDMath.norm1(INDArray in,
INDArray dimensions,
boolean keepDims)
Sum of absolute differences.
|
INDArray |
NDBase.norm1(INDArray x,
int... dimensions)
Norm1 (L1 norm) reduction operation: The output contains the L1 norm for each tensor/subset along the specified dimensions:
out = sum_i abs(x[i]) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm1(INDArray in,
int... dimensions)
Mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDBase.norm2(INDArray x,
boolean keepDims,
int... dimensions)
Norm2 (L2 norm) reduction operation: The output contains the L2 norm for each tensor/subset along the specified dimensions:
out = sqrt(sum_i x[i]^2) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm2(INDArray in,
boolean keepDims,
int... dimensions)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray |
NDMath.norm2(INDArray in,
INDArray dimensions)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray |
NDMath.norm2(INDArray in,
INDArray dimensions,
boolean keepDims)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray |
NDBase.norm2(INDArray x,
int... dimensions)
Norm2 (L2 norm) reduction operation: The output contains the L2 norm for each tensor/subset along the specified dimensions:
out = sqrt(sum_i x[i]^2) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.norm2(INDArray in,
int... dimensions)
Euclidean norm: euclidean distance of a vector from the origin
|
INDArray[] |
NDMath.normalizeMoments(INDArray counts,
INDArray means,
INDArray variances,
double shift)
Calculate the mean and variance from the sufficient statistics
|
INDArray |
NDBase.normmax(INDArray x,
boolean keepDims,
int... dimensions)
Max norm (infinity norm) reduction operation: The output contains the max norm for each tensor/subset along the
specified dimensions: out = max(abs(x[i])) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.normMax(INDArray in,
boolean keepDims,
int... dimensions)
Differences between max absolute value
|
INDArray |
NDMath.normMax(INDArray in,
INDArray dimensions)
Differences between max absolute value
|
INDArray |
NDMath.normMax(INDArray in,
INDArray dimensions,
boolean keepDims)
Differences between max absolute value
|
INDArray |
NDBase.normmax(INDArray x,
int... dimensions)
Max norm (infinity norm) reduction operation: The output contains the max norm for each tensor/subset along the
specified dimensions: out = max(abs(x[i])) Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.normMax(INDArray in,
int... dimensions)
Differences between max absolute value
|
INDArray |
NDBase.oneHot(INDArray indices,
int depth)
Convert the array to a one-hot array with values 0 and 1 for each entry
If input has shape [ a, ..., n] then output has shape [ a, ..., n, depth], with out[i, ..., j, in[i,...,j]] = 1 with other values being set to 0 see oneHot(SDVariable, int, int, double, double) |
INDArray |
NDBase.oneHot(INDArray indices,
int depth,
int axis,
double on,
double off)
Convert the array to a one-hot array with values and for each entry
If input has shape [ a, ..., n] then output has shape [ a, ..., n, depth], with {out[i, ..., j, in[i,...,j]] with other values being set to |
INDArray |
NDBase.oneHot(INDArray indices,
int depth,
int axis,
double on,
double off,
DataType dataType)
Convert the array to a one-hot array with values and for each entry
If input has shape [ a, ..., n] then output has shape [ a, ..., n, depth], with {out[i, ..., j, in[i,...,j]] with other values being set to |
INDArray |
NDBase.onesLike(INDArray input)
Return a variable of all 1s, with the same shape as the input variable.
|
INDArray |
NDBase.onesLike(INDArray input,
DataType dataType)
As per onesLike(String, SDVariable) but the output datatype may be specified
|
INDArray |
NDBitwise.or(INDArray x,
INDArray y)
Bitwise OR operation.
|
INDArray |
NDMath.or(INDArray x,
INDArray y)
Boolean OR operation: elementwise (x != 0) || (y != 0)
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. Returns an array with values 1 where condition is satisfied, or value 0 otherwise. |
INDArray |
NDNN.pad(INDArray input,
INDArray padding,
double constant)
Padding operation
|
INDArray |
NDImage.pad(INDArray input,
INDArray padding,
Mode Mode,
double padValue)
Pads an image according to the given padding type
|
INDArray |
NDNN.pad(INDArray input,
INDArray padding,
PadMode PadMode,
double constant)
Padding operation
|
INDArray |
NDBase.permute(INDArray x,
INDArray dimensions)
Array permutation operation: permute the dimensions according to the specified permutation indices.
Example: if input has shape [a,b,c] and dimensions = [2,0,1] the output has shape [c,a,b] |
INDArray |
NDBase.permute(INDArray x,
int... dimensions)
Array permutation operation: permute the dimensions according to the specified permutation indices.
Example: if input has shape [a,b,c] and dimensions = [2,0,1] the output has shape [c,a,b] |
INDArray |
NDMath.pow(INDArray x,
double value)
Element-wise power function: out = x^value
|
INDArray |
NDMath.pow(INDArray x,
INDArray y)
Element-wise (broadcastable) power function: out = x[i]^y[i]
|
INDArray |
NDNN.preciseGelu(INDArray x)
GELU activation function - Gaussian Error Linear Units
For more details, see Gaussian Error Linear Units (GELUs) - https://arxiv.org/abs/1606.08415 This method uses the precise method |
INDArray |
NDNN.prelu(INDArray input,
INDArray alpha,
int... sharedAxes)
PReLU (Parameterized Rectified Linear Unit) operation.
|
INDArray |
NDBase.prod(INDArray x,
boolean keepDims,
int... dimensions)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
boolean keepDims,
int... dimensions)
The max of an array along each dimension
|
INDArray |
NDBase.prod(INDArray x,
INDArray dimensions)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
INDArray dimensions)
The product of an array long each dimension
|
INDArray |
NDBase.prod(INDArray x,
INDArray dimensions,
boolean keepDims)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
INDArray dimensions,
boolean keepDims)
The product of an array long each dimension
|
INDArray |
NDBase.prod(INDArray x,
int... dimensions)
Product array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.prod(INDArray in,
int... dimensions)
The max of an array along each dimension
|
INDArray[] |
NDLinalg.qr(INDArray input)
Computes the QR decompositions of input matrix.
|
INDArray[] |
NDLinalg.qr(INDArray input,
boolean full)
Computes the QR decompositions of input matrix.
|
INDArray |
NDImage.randomCrop(INDArray input,
INDArray shape)
Randomly crops image
|
INDArray |
NDBase.range(INDArray from,
INDArray to,
INDArray step,
DataType dataType)
Create a new variable with a 1d array, where the values start at from and increment by step
up to (but not including) limit. For example, range(1.0, 3.0, 0.5) will return [1.0, 1.5, 2.0, 2.5] |
INDArray |
NDBase.rank(INDArray in)
Returns the rank (number of dimensions, i.e., length(shape)) of the specified INDArray as a 0D scalar variable
|
INDArray |
NDMath.rationalTanh(INDArray x)
Rational Tanh Approximation elementwise function, as described in the paper:
Compact Convolutional Neural Network Cascade for Face Detection This is a faster Tanh approximation |
INDArray |
NDMath.rdiv(INDArray x,
double value)
Scalar reverse division operation, out = scalar / in
|
INDArray |
NDMath.rdiv(INDArray x,
INDArray y)
Pairwise reverse division operation, out = y / x
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDMath.reciprocal(INDArray x)
Element-wise reciprocal (inverse) function: out[i] = 1 / in[i]
|
INDArray |
NDMath.rectifiedTanh(INDArray x)
Rectified tanh operation: max(0, tanh(in))
|
INDArray |
NDMath.reduceAMax(INDArray in,
boolean keepDims,
int... dimensions)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAMax(INDArray in,
INDArray dimensions)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAMax(INDArray in,
INDArray dimensions,
boolean keepDims)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAMax(INDArray in,
int... dimensions)
Absolute max array reduction operation, optionally along specified dimensions: out = max(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
boolean keepDims,
int... dimensions)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
INDArray dimensions)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
INDArray dimensions,
boolean keepDims)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmean(INDArray in,
int... dimensions)
Absolute mean array reduction operation, optionally along specified dimensions: out = mean(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
boolean keepDims,
int... dimensions)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
INDArray dimensions)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
INDArray dimensions,
boolean keepDims)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceAmin(INDArray in,
int... dimensions)
Absolute min array reduction operation, optionally along specified dimensions: out = min(abs(x))
|
INDArray |
NDMath.reduceMax(INDArray in,
boolean keepDims,
int... dimensions)
The max of an array along each dimension
|
INDArray |
NDMath.reduceMax(INDArray in,
INDArray dimensions)
The max of an array long each dimension
|
INDArray |
NDMath.reduceMax(INDArray in,
INDArray dimensions,
boolean keepDims)
The max of an array long each dimension
|
INDArray |
NDMath.reduceMax(INDArray in,
int... dimensions)
The max of an array along each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
boolean keepDims,
int... dimensions)
The minimum of an array along each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
INDArray dimensions)
The minimum of an array long each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
INDArray dimensions,
boolean keepDims)
The minimum of an array long each dimension
|
INDArray |
NDMath.reduceMin(INDArray in,
int... dimensions)
The minimum of an array along each dimension
|
INDArray |
NDNN.relu(INDArray x,
double cutoff)
Element-wise rectified linear function with specified cutoff:
out[i] = in[i] if in[i] >= cutoff out[i] = 0 otherwise |
INDArray |
NDNN.relu6(INDArray x,
double cutoff)
Element-wise "rectified linear 6" function with specified cutoff:
out[i] = min(max(in, cutoff), 6) |
INDArray |
NDNN.reluLayer(INDArray input,
INDArray weights,
INDArray bias)
ReLU (Rectified Linear Unit) layer operation: out = relu(mmul(in,w) + bias)
Note that bias array is optional |
INDArray |
NDBase.repeat(INDArray input,
INDArray repeats,
int axis)
A tensor with the shape of input minus the specified axis with elements repeated along the specified axis.
|
INDArray |
NDBase.replaceWhere(INDArray update,
double value,
Condition condition)
Element-wise replace where condition:
out[i] = value if condition(update[i]) is satisfied, or out[i] = update[i] if condition(update[i]) is NOT satisfied |
INDArray |
NDBase.replaceWhere(INDArray update,
INDArray from,
Condition condition)
Element-wise replace where condition:
out[i] = from[i] if condition(update[i]) is satisfied, or out[i] = update[i] if condition(update[i]) is NOT satisfied |
INDArray |
NDBase.reshape(INDArray x,
INDArray shape)
Reshape the input variable to the specified (fixed) shape.
|
INDArray |
NDBase.reshape(INDArray x,
long... shape)
Reshape the input variable to the specified (fixed) shape.
|
INDArray |
NDImage.resizeBiCubic(INDArray input,
INDArray size,
boolean alignCorners,
boolean alignPixelCenters)
Resize images to size using the specified method.
|
INDArray |
NDImage.resizeBiLinear(INDArray input,
int height,
int width,
boolean alignCorners,
boolean halfPixelCenters)
Resize images to size using the specified method.
|
INDArray |
NDBase.reverse(INDArray x,
int... dimensions)
Reverse the values of an array for the specified dimensions
If input is: [ 1, 2, 3] [ 4, 5, 6] then reverse(in, 0): [3, 2, 1] [6, 5, 4] reverse(in, 1): [4, 5, 6] [1, 2 3] |
INDArray |
NDBase.reverseSequence(INDArray x,
INDArray seq_lengths)
Reverse sequence op: for each slice along dimension seqDimension, the first seqLength values are reversed
|
INDArray |
NDBase.reverseSequence(INDArray x,
INDArray seq_lengths,
int seqDim,
int batchDim)
Reverse sequence op: for each slice along dimension seqDimension, the first seqLength values are reversed
|
INDArray |
NDImage.rgbToHsv(INDArray input)
Converting array from HSV to RGB format
|
INDArray |
NDImage.rgbToYiq(INDArray input)
Converting array from RGB to YIQ format
|
INDArray |
NDImage.rgbToYuv(INDArray input)
Converting array from RGB to YUV format
|
INDArray |
NDBitwise.rightShift(INDArray x,
INDArray y)
Bitwise right shift operation.
|
INDArray |
NDBitwise.rightShiftCyclic(INDArray x,
INDArray y)
Bitwise right cyclical shift operation.
|
INDArray |
NDMath.round(INDArray x)
Element-wise round function: out = round(x).
Rounds (up or down depending on value) to the nearest integer value. |
INDArray |
NDMath.rsqrt(INDArray x)
Element-wise reciprocal (inverse) of square root: out = 1.0 / sqrt(x)
|
INDArray |
NDMath.rsub(INDArray x,
double value)
Scalar reverse subtraction operation, out = scalar - in
|
INDArray |
NDMath.rsub(INDArray x,
INDArray y)
Pairwise reverse subtraction operation, out = y - x
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.scalarFloorMod(INDArray in,
double value)
Element-wise scalar floor modulus operation: out = floorMod(in, value).
i.e., returns the remainder after division by 'value' |
INDArray |
NDBase.scalarMax(INDArray in,
double value)
Element-wise scalar maximum operation: out = max(in, value)
|
INDArray |
NDBase.scalarMin(INDArray in,
double value)
Element-wise scalar minimum operation: out = min(in, value)
|
INDArray |
NDBase.scalarSet(INDArray in,
double set)
Return a variable with equal shape to the input, but all elements set to value 'set'
|
INDArray |
NDBase.scatterAdd(INDArray ref,
INDArray indices,
INDArray updates)
Scatter addition operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterDiv(INDArray ref,
INDArray indices,
INDArray updates)
Scatter division operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterMax(INDArray ref,
INDArray indices,
INDArray updates)
Scatter max operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterMin(INDArray ref,
INDArray indices,
INDArray updates)
Scatter min operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterMul(INDArray ref,
INDArray indices,
INDArray updates)
Scatter multiplication operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterSub(INDArray ref,
INDArray indices,
INDArray updates)
Scatter subtraction operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.scatterUpdate(INDArray ref,
INDArray indices,
INDArray updates)
Scatter update operation.
If indices is rank 0 (a scalar), then out[index, ...] = out[index, ...] + op(updates[...]) If indices is rank 1 (a vector), then for each position i, out[indices[i], ...] = out[indices[i], ...] + op(updates[i, ...]) If indices is rank 2+, then for each position (i,...,k), out[indices[i], ..., indices[k], ...] = out[indices[i], ..., indices[k], ...] + op(updates[i, ..., k, ...]) Note that if multiple indices refer to the same location, the contributions from each is handled correctly. |
INDArray |
NDBase.segmentMax(INDArray data,
INDArray segmentIds)
Segment max operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentMean(INDArray data,
INDArray segmentIds)
Segment mean operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentMin(INDArray data,
INDArray segmentIds)
Segment min operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentProd(INDArray data,
INDArray segmentIds)
Segment product operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDBase.segmentSum(INDArray data,
INDArray segmentIds)
Segment sum operation.
If data = [3, 6, 1, 4, 9, 2, 8] segmentIds = [0, 0, 1, 1, 1, 2, 2] then output = [6, 9, 8] = [op(3,6), op(1,4,9), op(2,8)] Note that the segment IDs must be sorted from smallest to largest segment. See {unsortedSegment (String, SDVariable, SDVariable, int) ops for the same op without this sorted requirement |
INDArray |
NDNN.selu(INDArray x)
Element-wise SeLU function - Scaled exponential Lineal Unit: see Self-Normalizing Neural Networks
out[i] = scale * alpha * (exp(in[i])-1) if in[i]>0, or 0 if in[i] <= 0 Uses default scale and alpha values. |
INDArray |
NDCNN.separableConv2d(INDArray layerInput,
INDArray depthWeights,
INDArray pointWeights,
Conv2DConfig Conv2DConfig)
Separable 2D convolution operation with optional bias
|
INDArray |
NDCNN.separableConv2d(INDArray layerInput,
INDArray depthWeights,
INDArray pointWeights,
INDArray bias,
Conv2DConfig Conv2DConfig)
Separable 2D convolution operation with optional bias
|
INDArray |
NDBase.sequenceMask(INDArray lengths,
DataType dataType)
see sequenceMask(String, SDVariable, SDVariable, DataType)
|
INDArray |
NDBase.sequenceMask(INDArray lengths,
INDArray maxLen,
DataType dataType)
Generate a sequence mask (with values 0 or 1) based on the specified lengths
Specifically, out[i, ..., k, j] = (j < lengths[i, ..., k] ? 1.0 : 0.0) |
INDArray |
NDBase.sequenceMask(INDArray lengths,
int maxLen,
DataType dataType)
Generate a sequence mask (with values 0 or 1) based on the specified lengths
Specifically, out[i, ..., k, j] = (j < lengths[i, ..., k] ? 1.0 : 0.0) |
INDArray |
NDMath.setDiag(INDArray in,
INDArray diag)
Set the diagonal value to the specified values
If input is [ a, b, c] [ d, e, f] [ g, h, i] and diag = [ 1, 2, 3] then output is [ 1, b, c] [ d, 2, f] [ g, h, 3] |
INDArray[] |
NDBase.setShape(INDArray input,
INDArray shape)
Sets an inplace shape on the passed in input.
|
INDArray |
NDMath.shannonEntropy(INDArray in,
boolean keepDims,
int... dimensions)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDMath.shannonEntropy(INDArray in,
INDArray dimensions)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDMath.shannonEntropy(INDArray in,
INDArray dimensions,
boolean keepDims)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDMath.shannonEntropy(INDArray in,
int... dimensions)
Shannon Entropy reduction: -sum(x * log2(x))
|
INDArray |
NDBase.shape(INDArray input)
Returns the shape of the specified INDArray as a 1D INDArray
|
INDArray |
NDNN.sigmoid(INDArray x)
Element-wise sigmoid function: out[i] = 1.0/(1+exp(-in[i]))
|
INDArray |
NDLoss.sigmoidCrossEntropy(INDArray label,
INDArray predictionLogits,
INDArray weights)
Sigmoid cross entropy: applies the sigmoid activation function on the input logits (input "pre-sigmoid preductions")
and implements the binary cross entropy loss function. |
INDArray |
NDLoss.sigmoidCrossEntropy(INDArray label,
INDArray predictionLogits,
INDArray weights,
LossReduce lossReduce,
double labelSmoothing)
Sigmoid cross entropy: applies the sigmoid activation function on the input logits (input "pre-sigmoid preductions")
and implements the binary cross entropy loss function. |
INDArray |
NDNN.sigmoidDerivative(INDArray x,
INDArray wrt)
Element-wise sigmoid function derivative: dL/dIn given input and dL/dOut
|
INDArray |
NDMath.sign(INDArray x)
Element-wise sign (signum) function:
out = -1 if in < 0 out = 0 if in = 0 out = 1 if in > 0 |
INDArray |
NDMath.sin(INDArray x)
Elementwise sine operation: out = sin(x)
|
INDArray |
NDMath.sinh(INDArray x)
Elementwise sinh (hyperbolic sine) operation: out = sinh(x)
|
INDArray |
NDBase.size(INDArray in)
Returns the size (number of elements, i.e., prod(shape)) of the specified INDArray as a 0D scalar variable
|
INDArray |
NDBase.sizeAt(INDArray in,
int dimension)
Returns a rank 0 (scalar) variable for the size of the specified dimension.
For example, if X has shape [10,20,30] then sizeAt(X,1)=20. |
INDArray |
NDBase.slice(INDArray input,
INDArray begin,
INDArray size)
Get a subset of the specified input, by specifying the first element and the size of the array.
For example, if input is: [a, b, c] [d, e, f] then slice(input, begin=[0,1], size=[2,1] will return: [b] [e] Note that for each dimension i, begin[i] + size[i] <= input.size(i) |
INDArray |
NDBase.slice(INDArray input,
int[] begin,
int... size)
Get a subset of the specified input, by specifying the first element and the size of the array.
For example, if input is: [a, b, c] [d, e, f] then slice(input, begin=[0,1], size=[2,1] will return: [b] [e] Note that for each dimension i, begin[i] + size[i] <= input.size(i) |
INDArray |
NDNN.softmax(INDArray x)
Softmax activation, along the specified dimension
|
INDArray |
NDNN.softmax(INDArray x,
int dimension)
Softmax activation, along the specified dimension
|
INDArray |
NDLoss.softmaxCrossEntropy(INDArray oneHotLabels,
INDArray logitPredictions,
INDArray weights)
Applies the softmax activation function to the input, then implement multi-class cross entropy:
-sum_classes label[i] * log(p[c]) where p = softmax(logits)If LossReduce#NONE is used, returned shape is [numExamples] out for [numExamples, numClasses] predicitons/labels; otherwise, the output is a scalar. |
INDArray |
NDLoss.softmaxCrossEntropy(INDArray oneHotLabels,
INDArray logitPredictions,
INDArray weights,
LossReduce lossReduce,
double labelSmoothing)
Applies the softmax activation function to the input, then implement multi-class cross entropy:
-sum_classes label[i] * log(p[c]) where p = softmax(logits)If LossReduce#NONE is used, returned shape is [numExamples] out for [numExamples, numClasses] predicitons/labels; otherwise, the output is a scalar. |
INDArray |
NDNN.softmaxDerivative(INDArray x,
INDArray wrt,
int dimension)
Softmax derivative function
|
INDArray |
NDNN.softplus(INDArray x)
Element-wise softplus function: out = log(exp(x) + 1)
|
INDArray |
NDNN.softsign(INDArray x)
Element-wise softsign function: out = x / (abs(x) + 1)
|
INDArray |
NDNN.softsignDerivative(INDArray x)
Element-wise derivative (dOut/dIn) of the softsign function softsign(INDArray)
|
INDArray |
NDLinalg.solve(INDArray matrix,
INDArray rhs)
Solver for systems of linear equations.
|
INDArray |
NDLinalg.solve(INDArray matrix,
INDArray rhs,
boolean adjoint)
Solver for systems of linear equations.
|
INDArray |
NDCNN.spaceToBatch(INDArray x,
int[] blocks,
int[] paddingTop,
int... paddingBottom)
Convolution 2d layer space to batch operation on 4d input.
Increases input batch dimension by rearranging data from spatial dimensions into batch dimension |
INDArray |
NDCNN.spaceToDepth(INDArray x,
int blockSize,
DataFormat dataFormat)
Convolution 2d layer space to depth operation on 4d input.
Increases input channels (reduced spatial dimensions) by rearranging data into a larger channels dimension Example: if input has shape [mb, 2, 4, 4] and block size is 2, then output size is [mb, 8/(2*2), 2*2, 2*2] = [mb, 2, 4, 4] |
INDArray |
NDLoss.sparseSoftmaxCrossEntropy(INDArray logits,
INDArray labels)
As per softmaxCrossEntropy(String, SDVariable, SDVariable, LossReduce) but the labels variable
is represented as an integer array instead of the equivalent one-hot array. i.e., if logits are rank N, then labels have rank N-1 |
INDArray |
NDBase.sparseToDense(INDArray indices,
INDArray shape,
INDArray values)
Create a dense matrix equivalent of a sparse matrix based on the given input.
|
INDArray |
NDBase.sparseToDense(INDArray indices,
INDArray shape,
INDArray values,
INDArray defaultValue)
Create a dense matrix equivalent of a sparse matrix based on the given input.
|
INDArray[] |
NDBase.split(INDArray input,
INDArray numSplit,
int splitDim)
Split a value in to a list of ndarrays.
|
INDArray[] |
NDBase.split(INDArray input,
int numSplit,
int splitDim)
Split a value in to a list of ndarrays.
|
INDArray[] |
NDBase.splitV(INDArray input,
INDArray sizes,
int numSplit,
int splitDim)
Split a value in to a list of ndarrays with varying sizes
according to the sizes parameter. |
INDArray |
NDMath.sqrt(INDArray x)
Element-wise square root function: out = sqrt(x)
|
INDArray |
NDMath.square(INDArray x)
Element-wise square function: out = x^2
|
INDArray |
NDMath.squaredDifference(INDArray x,
INDArray y)
Pairwise squared difference operation.
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.squaredNorm(INDArray x,
boolean keepDims,
int... dimensions)
Squared L2 norm: see norm2(String, SDVariable, boolean, int...)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.squaredNorm(INDArray in,
boolean keepDims,
int... dimensions)
Sum of squared differences.
|
INDArray |
NDMath.squaredNorm(INDArray in,
INDArray dimensions)
Sum of squared differences.
|
INDArray |
NDMath.squaredNorm(INDArray in,
INDArray dimensions,
boolean keepDims)
Sum of squared differences.
|
INDArray |
NDBase.squaredNorm(INDArray x,
int... dimensions)
Squared L2 norm: see norm2(String, SDVariable, boolean, int...)
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.squaredNorm(INDArray in,
int... dimensions)
Sum of squared differences.
|
INDArray |
NDBase.squeeze(INDArray x,
int axis)
Remove a single dimension of size 1.
For example, if input has shape [a,b,1,c] then squeeze(input, 2) returns an array of shape [a,b,c] |
INDArray |
NDRNN.sru(INDArray x,
INDArray initialC,
INDArray mask,
SRUWeights SRUWeights)
The SRU layer.
|
INDArray |
NDRNN.sru(INDArray x,
INDArray initialC,
SRUWeights SRUWeights)
The SRU layer.
|
INDArray |
NDRNN.sruCell(INDArray x,
INDArray cLast,
SRUWeights SRUWeights)
The SRU layer.
|
INDArray |
NDBase.stack(int axis,
INDArray... values)
Stack a set of N INDArray of rank X into one rank X+1 variable.
If inputs have shape [a,b,c] then output has shape: axis = 0: [N,a,b,c] axis = 1: [a,N,b,c] axis = 2: [a,b,N,c] axis = 3: [a,b,c,N] see unstack(String[], SDVariable, int, int) |
INDArray |
NDBase.standardDeviation(INDArray x,
boolean biasCorrected,
boolean keepDims,
int... dimensions)
Standard deviation array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.standardDeviation(INDArray x,
boolean biasCorrected,
int... dimensions)
Standard deviation array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.standardize(INDArray x,
int... dimensions)
Standardize input variable along given axis
|
INDArray |
NDMath.step(INDArray x,
double value)
Elementwise step function:
out(x) = 1 if x >= cutoff out(x) = 0 otherwise |
INDArray |
NDBase.stridedSlice(INDArray in,
INDArray begin,
INDArray end,
INDArray strides)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDBase.stridedSlice(INDArray in,
INDArray begin,
INDArray end,
INDArray strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDBase.stridedSlice(INDArray in,
long[] begin,
long[] end,
long... strides)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDBase.stridedSlice(INDArray in,
long[] begin,
long[] end,
long[] strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask)
Get a subset of the specified input, by specifying the first element, last element, and the strides.
For example, if input is: [a, b, c] [d, e, f] [g, h, i] then stridedSlice(input, begin=[0,1], end=[2,2], strides=[2,1], all masks = 0) will return: [b, c] [h, i] |
INDArray |
NDMath.sub(INDArray x,
double value)
Scalar subtraction operation, out = in - scalar
|
INDArray |
NDMath.sub(INDArray x,
INDArray y)
Pairwise subtraction operation, out = x - y
Note: supports broadcasting if x and y have different shapes and are broadcastable. For example, if X has shape [1,10] and Y has shape [5,10] then op(X,Y) has output shape [5,10] Broadcast rules are the same as NumPy: https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html |
INDArray |
NDBase.sum(INDArray x,
boolean keepDims,
int... dimensions)
Sum array reduction operation, optionally along specified dimensions.
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.sum(INDArray in,
boolean keepDims,
int... dimensions)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDMath.sum(INDArray in,
INDArray dimensions)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDMath.sum(INDArray in,
INDArray dimensions,
boolean keepDims)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDBase.sum(INDArray x,
int... dimensions)
Sum array reduction operation, optionally along specified dimensions.
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDMath.sum(INDArray in,
int... dimensions)
Sum of an array, optionally along specified dimensions: out = sum(x))
|
INDArray |
NDLinalg.svd(INDArray input,
boolean fullUV,
boolean computeUV)
Calculates singular value decomposition.
|
INDArray |
NDLinalg.svd(INDArray input,
boolean fullUV,
boolean computeUV,
int switchNum)
Calculates singular value decomposition.
|
INDArray |
NDNN.swish(INDArray x)
Element-wise "swish" function: out = x * sigmoid(b*x) with b=1.0
See: https://arxiv.org/abs/1710.05941 |
INDArray[] |
NDBase.switchOp(INDArray x,
INDArray predicate)
Switch operation
Predicate - if false, values are output to left (first) branch/output; if true, to right (second) branch/output |
INDArray |
NDMath.tan(INDArray x)
Elementwise tangent operation: out = tan(x)
|
INDArray |
NDNN.tanh(INDArray x)
Elementwise tanh (hyperbolic tangent) operation: out = tanh(x)
|
INDArray |
NDMath.tanh(INDArray x)
Elementwise tanh (hyperbolic tangent) operation: out = tanh(x)
|
INDArray |
NDBase.tensorMmul(INDArray x,
INDArray y,
int[] dimensionsX,
int... dimensionsY)
//TODO: Ops must be documented.
|
INDArray |
NDBase.tensorMmul(INDArray x,
INDArray y,
int[] dimensionsX,
int[] dimensionsY,
boolean transposeX,
boolean transposeY,
boolean transposeZ)
//TODO: Ops must be documented.
|
INDArray |
NDBase.tile(INDArray x,
INDArray repeat)
Repeat (tile) the input tensor the specified number of times.
For example, if input is [1, 2] [3, 4] and repeat is [2, 3] then output is [1, 2, 1, 2, 1, 2] [3, 4, 3, 4, 3, 4] [1, 2, 1, 2, 1, 2] [3, 4, 3, 4, 3, 4] |
INDArray |
NDBase.tile(INDArray x,
int... repeat)
see tile(String, SDVariable, int...)
|
INDArray |
NDMath.trace(INDArray in)
Matrix trace operation
For rank 2 matrices, the output is a scalar vith the trace - i.e., sum of the main diagonal. For higher rank inputs, output[a,b,c] = trace(in[a,b,c,:,:]) |
INDArray |
NDBase.transpose(INDArray x)
Matrix transpose operation: If input has shape [a,b] output has shape [b,a]
|
INDArray |
NDLinalg.triangularSolve(INDArray matrix,
INDArray rhs,
boolean lower,
boolean adjoint)
Solver for systems of linear questions.
|
INDArray |
NDLinalg.triu(INDArray input)
Upper triangle of an array.
|
INDArray |
NDLinalg.triu(INDArray input,
int diag)
Upper triangle of an array.
|
INDArray |
NDBase.unsortedSegmentMax(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment max operation.
|
INDArray |
NDBase.unsortedSegmentMean(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment mean operation.
|
INDArray |
NDBase.unsortedSegmentMin(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment min operation.
|
INDArray |
NDBase.unsortedSegmentProd(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment product operation.
|
INDArray |
NDBase.unsortedSegmentSqrtN(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment sqrtN operation.
|
INDArray |
NDBase.unsortedSegmentSum(INDArray data,
INDArray segmentIds,
int numSegments)
Unsorted segment sum operation.
|
INDArray[] |
NDBase.unstack(INDArray value,
int axis,
int num)
Unstack a variable of rank X into N rank X-1 variables by taking slices along the specified axis.
If input has shape [a,b,c] then output has shape: axis = 0: [b,c] axis = 1: [a,c] axis = 2: [a,b] |
INDArray |
NDCNN.upsampling2d(INDArray input,
int scale)
Upsampling layer for 2D inputs.
scale is used for both height and width dimensions. |
INDArray |
NDCNN.upsampling2d(INDArray input,
int scaleH,
int scaleW,
boolean nchw)
2D Convolution layer operation - Upsampling 2d
|
INDArray |
NDCNN.upsampling3d(INDArray input,
boolean ncdhw,
int scaleD,
int scaleH,
int scaleW)
3D Convolution layer operation - Upsampling 3d
|
INDArray |
NDBase.variance(INDArray x,
boolean biasCorrected,
boolean keepDims,
int... dimensions)
Variance array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.variance(INDArray x,
boolean biasCorrected,
int... dimensions)
Variance array reduction operation, optionally along specified dimensions
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.where(INDArray condition)
Returns elements that are true from the given condition array
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.where(INDArray x,
INDArray condition)
Similar to numpy where, takes elements from x or y depending on whether the condition at a given element is true or false
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.where(INDArray x,
INDArray y,
INDArray condition)
Similar to numpy where, takes elements from x or y depending on whether the condition at a given element is true or false
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBase.whereNumpy(INDArray x,
INDArray y,
INDArray condition)
As implemented in numpy, Return elements chosen from x or y depending on condition.
Note that if keepDims = true, the output variable has the same rank as the input variable, with the reduced dimensions having size 1. |
INDArray |
NDBitwise.xor(INDArray x,
INDArray y)
Bitwise XOR operation (exclusive OR).
|
INDArray |
NDMath.xor(INDArray x,
INDArray y)
Boolean XOR (exclusive OR) operation: elementwise (x != 0) XOR (y != 0)
If x and y arrays have equal shape, the output shape is the same as these inputs. Note: supports broadcasting if x and y have different shapes and are broadcastable. Returns an array with values 1 where condition is satisfied, or value 0 otherwise. |
INDArray |
NDImage.yiqToRgb(INDArray input)
Converting image from YIQ to RGB format
|
INDArray |
NDImage.yuvToRgb(INDArray input)
Converting image from YUV to RGB format
|
INDArray |
NDMath.zeroFraction(INDArray input)
Full array zero fraction array reduction operation, optionally along specified dimensions: out = (count(x == 0) / length(x))
|
INDArray |
NDBase.zerosLike(INDArray input)
Return a variable of all 0s, with the same shape as the input variable.
|
| Modifier and Type | Method and Description |
|---|---|
static Task |
TaskUtils.buildTask(INDArray array) |
static Task |
TaskUtils.buildTask(INDArray[] array,
INDArray[] labels) |
static Task |
TaskUtils.buildTask(INDArray[] array,
INDArray[] labels) |
static Task |
TaskUtils.buildTask(INDArray array,
INDArray labels) |
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
BooleanIndexing.applyMask(INDArray arr,
INDArray mask)
A minor shortcut for applying a bitmask to
a matrix
|
static INDArray |
BooleanIndexing.applyMaskInPlace(INDArray arr,
INDArray mask)
A minor shortcut for applying a bitmask to
a matrix
|
static INDArray |
BooleanIndexing.chooseFrom(@NonNull INDArray[] input,
@NonNull Condition condition)
Choose from the inputs based on the given condition.
|
static INDArray |
BooleanIndexing.chooseFrom(@NonNull INDArray[] input,
@NonNull List<Double> tArgs,
@NonNull List<Integer> iArgs,
@NonNull Condition condition)
Choose from the inputs based on the given condition.
|
static INDArray |
BooleanIndexing.firstIndex(INDArray array,
Condition condition)
This method returns first index matching given condition
PLEASE NOTE: This method will return -1 value if condition wasn't met
|
static INDArray |
BooleanIndexing.firstIndex(INDArray array,
Condition condition,
int... dimension)
This method returns first index matching given condition along given dimensions
PLEASE NOTE: This method will return -1 values for missing conditions
|
static INDArray |
BooleanIndexing.lastIndex(INDArray array,
Condition condition)
This method returns last index matching given condition
PLEASE NOTE: This method will return -1 value if condition wasn't met
|
static INDArray |
BooleanIndexing.lastIndex(INDArray array,
Condition condition,
int... dimension)
This method returns first index matching given condition along given dimensions
PLEASE NOTE: This method will return -1 values for missing conditions
|
| Modifier and Type | Method and Description |
|---|---|
static INDArrayIndex[] |
NDArrayIndex.allFor(INDArray arr)
Generate an all index
equal to the rank of the given array
|
static boolean |
BooleanIndexing.and(INDArray n,
Condition cond)
And over the whole ndarray given some condition
|
static boolean[] |
BooleanIndexing.and(INDArray n,
Condition condition,
int... dimension)
And over the whole ndarray given some condition, with respect to dimensions
|
static INDArray |
BooleanIndexing.applyMask(INDArray arr,
INDArray mask)
A minor shortcut for applying a bitmask to
a matrix
|
static INDArray |
BooleanIndexing.applyMaskInPlace(INDArray arr,
INDArray mask)
A minor shortcut for applying a bitmask to
a matrix
|
static void |
BooleanIndexing.assignIf(@NonNull INDArray to,
@NonNull INDArray from,
@NonNull Condition condition)
This method does element-wise comparison
for 2 equal-sized matrices, for each element that matches Condition.
|
static void |
BooleanIndexing.assignIf(@NonNull INDArray to,
@NonNull INDArray from,
@NonNull Condition condition)
This method does element-wise comparison
for 2 equal-sized matrices, for each element that matches Condition.
|
static INDArray |
BooleanIndexing.chooseFrom(@NonNull INDArray[] input,
@NonNull Condition condition)
Choose from the inputs based on the given condition.
|
static INDArray |
BooleanIndexing.chooseFrom(@NonNull INDArray[] input,
@NonNull List<Double> tArgs,
@NonNull List<Integer> iArgs,
@NonNull Condition condition)
Choose from the inputs based on the given condition.
|
static INDArrayIndex[] |
Indices.createFromStartAndEnd(INDArray start,
INDArray end)
Create an n dimensional index
based on the given interval indices.
|
static INDArrayIndex[] |
Indices.createFromStartAndEnd(INDArray start,
INDArray end,
boolean inclusive)
Create indices representing intervals
along each dimension
|
static INDArray |
BooleanIndexing.firstIndex(INDArray array,
Condition condition)
This method returns first index matching given condition
PLEASE NOTE: This method will return -1 value if condition wasn't met
|
static INDArray |
BooleanIndexing.firstIndex(INDArray array,
Condition condition,
int... dimension)
This method returns first index matching given condition along given dimensions
PLEASE NOTE: This method will return -1 values for missing conditions
|
void |
NewAxis.init(INDArray arr,
int dimension) |
void |
PointIndex.init(INDArray arr,
int dimension) |
void |
SpecifiedIndex.init(INDArray arr,
int dimension) |
void |
NDArrayIndex.init(INDArray arr,
int dimension) |
void |
INDArrayIndex.init(INDArray arr,
int dimension)
Init the index wrt
the dimension and the given nd array
|
void |
IntervalIndex.init(INDArray arr,
int dimension) |
void |
NewAxis.init(INDArray arr,
long begin,
int dimension) |
void |
PointIndex.init(INDArray arr,
long begin,
int dimension) |
void |
SpecifiedIndex.init(INDArray arr,
long begin,
int dimension) |
void |
NDArrayIndex.init(INDArray arr,
long begin,
int dimension) |
void |
NDArrayIndexAll.init(INDArray arr,
long begin,
int dimension) |
void |
INDArrayIndex.init(INDArray arr,
long begin,
int dimension)
Init the index wrt
the dimension and the given nd array
|
void |
IntervalIndex.init(INDArray arr,
long begin,
int dimension) |
static boolean |
Indices.isScalar(INDArray indexOver,
INDArrayIndex... indexes)
Check if the given indexes
over the specified array
are searching for a scalar
|
static INDArray |
BooleanIndexing.lastIndex(INDArray array,
Condition condition)
This method returns last index matching given condition
PLEASE NOTE: This method will return -1 value if condition wasn't met
|
static INDArray |
BooleanIndexing.lastIndex(INDArray array,
Condition condition,
int... dimension)
This method returns first index matching given condition along given dimensions
PLEASE NOTE: This method will return -1 values for missing conditions
|
static long |
Indices.linearOffset(int index,
INDArray arr)
Compute the linear offset
for an index in an ndarray.
|
static long |
NDArrayIndex.offset(INDArray arr,
INDArrayIndex... indices)
Compute the offset given an array of offsets.
|
static long |
NDArrayIndex.offset(INDArray arr,
long... offsets)
Compute the offset given an array of offsets.
|
static boolean |
BooleanIndexing.or(INDArray n,
Condition cond)
Or over the whole ndarray given some condition
|
static boolean[] |
BooleanIndexing.or(INDArray n,
Condition condition,
int... dimension)
Or over the whole ndarray given some condition, with respect to dimensions
|
static void |
BooleanIndexing.replaceWhere(@NonNull INDArray to,
@NonNull INDArray from,
@NonNull Condition condition)
This method does element-wise comparison for 2 equal-sized matrices, for each element that matches Condition
|
static void |
BooleanIndexing.replaceWhere(@NonNull INDArray to,
@NonNull INDArray from,
@NonNull Condition condition)
This method does element-wise comparison for 2 equal-sized matrices, for each element that matches Condition
|
static void |
BooleanIndexing.replaceWhere(@NonNull INDArray to,
@NonNull Number set,
@NonNull Condition condition)
This method does element-wise assessing for 2 equal-sized matrices, for each element that matches Condition
|
static INDArrayIndex[] |
NDArrayIndex.resolve(INDArray arr,
INDArrayIndex... intendedIndexes)
Given an all index and
the intended indexes, return an
index array containing a combination of all elements
for slicing and overriding particular indexes where necessary
|
static int |
Indices.rowNumber(int index,
INDArray arr)
Compute the linear offset
for an index in an ndarray.
|
static int[] |
Indices.stride(INDArray arr,
INDArrayIndex[] indexes,
int... shape)
Return the stride to be used for indexing
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
InvertMatrix.invert(INDArray arr,
boolean inPlace)
Inverts a matrix
|
static INDArray |
InvertMatrix.pinvert(INDArray arr,
boolean inPlace)
Calculates pseudo inverse of a matrix using QR decomposition
|
static INDArray |
InvertMatrix.pLeftInvert(INDArray arr,
boolean inPlace)
Compute the left pseudo inverse.
|
static INDArray |
InvertMatrix.pRightInvert(INDArray arr,
boolean inPlace)
Compute the right pseudo inverse.
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
InvertMatrix.invert(INDArray arr,
boolean inPlace)
Inverts a matrix
|
static INDArray |
InvertMatrix.pinvert(INDArray arr,
boolean inPlace)
Calculates pseudo inverse of a matrix using QR decomposition
|
static INDArray |
InvertMatrix.pLeftInvert(INDArray arr,
boolean inPlace)
Compute the left pseudo inverse.
|
static INDArray |
InvertMatrix.pRightInvert(INDArray arr,
boolean inPlace)
Compute the right pseudo inverse.
|
| Modifier and Type | Field and Description |
|---|---|
INDArray |
AdaGradUpdater.historicalGradient |
| Modifier and Type | Method and Description |
|---|---|
Map<String,INDArray> |
NesterovsUpdater.getState() |
Map<String,INDArray> |
AdaGradUpdater.getState() |
Map<String,INDArray> |
AdaBeliefUpdater.getState() |
Map<String,INDArray> |
AdaDeltaUpdater.getState() |
Map<String,INDArray> |
AdamUpdater.getState() |
Map<String,INDArray> |
RmsPropUpdater.getState() |
Map<String,INDArray> |
AMSGradUpdater.getState() |
Map<String,INDArray> |
NoOpUpdater.getState() |
Map<String,INDArray> |
SgdUpdater.getState() |
Map<String,INDArray> |
AdaMaxUpdater.getState() |
Map<String,INDArray> |
GradientUpdater.getState() |
Map<String,INDArray> |
NadamUpdater.getState() |
| Modifier and Type | Method and Description |
|---|---|
void |
NesterovsUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Get the nesterov update
|
void |
AdaGradUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Gets feature specific learning rates
Adagrad keeps a history of gradients being passed in.
|
void |
AdaBeliefUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Calculate the update based on the given gradient
|
void |
AdaDeltaUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Get the updated gradient for the given gradient
and also update the state of ada delta.
|
void |
AdamUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Calculate the update based on the given gradient
|
void |
RmsPropUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch) |
void |
AMSGradUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch) |
void |
NoOpUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch) |
void |
SgdUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch) |
void |
AdaMaxUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Calculate the update based on the given gradient
|
void |
GradientUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Modify the gradient to be an update.
|
void |
NadamUpdater.applyUpdater(INDArray gradient,
int iteration,
int epoch)
Calculate the update based on the given gradient
|
void |
NesterovsUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
AdaGradUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
AdaBeliefUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
AdaDeltaUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
AdamUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
RmsPropUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
AMSGradUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
NoOpUpdater.setStateViewArray(INDArray viewArray,
long[] shape,
char order,
boolean initialize) |
void |
SgdUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
AdaMaxUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
GradientUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize)
For the internal updater state (if any): set this to use the provided array.
|
void |
NadamUpdater.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
| Modifier and Type | Method and Description |
|---|---|
void |
AdaGradUpdater.setState(Map<String,INDArray> stateMap,
boolean initialize) |
void |
AdaDeltaUpdater.setState(Map<String,INDArray> stateMap,
boolean initialize) |
void |
NoOpUpdater.setState(Map<String,INDArray> stateMap,
boolean initialize) |
void |
SgdUpdater.setState(Map<String,INDArray> stateMap,
boolean initialize) |
void |
GradientUpdater.setState(Map<String,INDArray> stateMap,
boolean initialize) |
| Modifier and Type | Method and Description |
|---|---|
GradientUpdater |
AdaBelief.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
AdaGrad.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
AdaDelta.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
RmsProp.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
AMSGrad.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
Adam.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
IUpdater.instantiate(INDArray viewArray,
boolean initializeViewArray)
Create a new gradient updater
|
GradientUpdater |
Nadam.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
Nesterovs.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
NoOp.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
AdaMax.instantiate(INDArray viewArray,
boolean initializeViewArray) |
GradientUpdater |
Sgd.instantiate(INDArray viewArray,
boolean initializeViewArray) |
| Modifier and Type | Method and Description |
|---|---|
GradientUpdater |
AdaBelief.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
AdaGrad.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
AdaDelta.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
RmsProp.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
AMSGrad.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
Adam.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
IUpdater.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
Nadam.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
Nesterovs.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
NoOp.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
AdaMax.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
GradientUpdater |
Sgd.instantiate(Map<String,INDArray> updaterState,
boolean initializeStateArrays) |
| Modifier and Type | Field and Description |
|---|---|
INDArray |
AdaGrad.historicalGradient |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
AdaGrad.getGradient(INDArray gradient,
int iteration)
Gets feature specific learning rates
Adagrad keeps a history of gradients being passed in.
|
INDArray |
AdaGrad.getGradient(INDArray gradient,
int slice,
long[] shape) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
AdaGrad.getGradient(INDArray gradient,
int iteration)
Gets feature specific learning rates
Adagrad keeps a history of gradients being passed in.
|
INDArray |
AdaGrad.getGradient(INDArray gradient,
int slice,
long[] shape) |
void |
AdaGrad.setStateViewArray(INDArray viewArray,
int[] gradientShape,
char gradientOrder,
boolean initialize) |
void |
AdaGrad.setStateViewArray(INDArray viewArray,
long[] gradientShape,
char gradientOrder,
boolean initialize) |
| Modifier and Type | Method and Description |
|---|---|
void |
L2Regularization.apply(INDArray param,
INDArray gradView,
double lr,
int iteration,
int epoch) |
void |
L1Regularization.apply(INDArray param,
INDArray gradView,
double lr,
int iteration,
int epoch) |
void |
Regularization.apply(INDArray param,
INDArray gradView,
double lr,
int iteration,
int epoch)
Apply the regularization by modifying the gradient array in-place
|
void |
WeightDecay.apply(INDArray param,
INDArray gradView,
double lr,
int iteration,
int epoch) |
double |
L2Regularization.score(INDArray param,
int iteration,
int epoch) |
double |
L1Regularization.score(INDArray param,
int iteration,
int epoch) |
double |
Regularization.score(INDArray param,
int iteration,
int epoch)
Calculate the loss function score component for the regularization.
For example, in L2 regularization, this would return L = 0.5 * sum_i param[i]^2For regularization types that don't have a score component, this method can return 0. |
double |
WeightDecay.score(INDArray param,
int iteration,
int epoch) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
ILossFunction.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the gradient of the loss function with respect to the inputs: dL/dOutput
|
INDArray |
SameDiffLoss.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the gradient of the loss function with respect to the inputs: dL/dOutput
|
INDArray |
ILossFunction.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the score (loss function value) for each example individually.
|
INDArray |
SameDiffLoss.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the score (loss function value) for each example individually.
|
| Modifier and Type | Method and Description |
|---|---|
Pair<Double,INDArray> |
ILossFunction.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average)
Compute both the score (loss function value) and gradient.
|
Pair<Double,INDArray> |
SameDiffLoss.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average)
Compute both the score (loss function value) and gradient.
|
| Modifier and Type | Method and Description |
|---|---|
static void |
LossUtil.applyMask(INDArray to,
INDArray mask) |
INDArray |
ILossFunction.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the gradient of the loss function with respect to the inputs: dL/dOutput
|
INDArray |
SameDiffLoss.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the gradient of the loss function with respect to the inputs: dL/dOutput
|
Pair<Double,INDArray> |
ILossFunction.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average)
Compute both the score (loss function value) and gradient.
|
Pair<Double,INDArray> |
SameDiffLoss.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average)
Compute both the score (loss function value) and gradient.
|
double |
ILossFunction.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average)
Compute the score (loss function value) for the given inputs.
|
double |
SameDiffLoss.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average)
Compute the score (loss function value) for the given inputs.
|
INDArray |
ILossFunction.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the score (loss function value) for each example individually.
|
INDArray |
SameDiffLoss.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
Compute the score (loss function value) for each example individually.
|
static boolean |
LossUtil.isPerOutputMasking(INDArray to,
INDArray mask) |
| Modifier and Type | Field and Description |
|---|---|
protected INDArray |
LossL1.weights |
protected INDArray |
LossL2.weights |
protected INDArray |
LossMCXENT.weights |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
LossMixtureDensity.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
This method returns the gradient of the cost function with respect to the
output from the previous layer.
|
INDArray |
LossWasserstein.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSquaredHinge.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAPE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossKLD.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSparseMCXENT.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL1.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossBinaryXENT.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL2.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMultiLabel.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossPoisson.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossCosineProximity.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossHinge.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMCXENT.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSLE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossFMeasure.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMixtureDensity.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
This method returns the score for each of the given outputs against the
given set of labels.
|
INDArray |
LossWasserstein.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSquaredHinge.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAPE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossKLD.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSparseMCXENT.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL1.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossBinaryXENT.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL2.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMultiLabel.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossPoisson.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossCosineProximity.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossHinge.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMCXENT.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSLE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossFMeasure.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSquaredHinge.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAPE.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL1.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
protected INDArray |
LossL2.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMultiLabel.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossPoisson.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossCosineProximity.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossHinge.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
protected INDArray |
LossMCXENT.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSLE.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
protected INDArray |
LossSparseMCXENT.sparseScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
| Modifier and Type | Method and Description |
|---|---|
Pair<Double,INDArray> |
LossMixtureDensity.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossWasserstein.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossSquaredHinge.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMAPE.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossKLD.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossSparseMCXENT.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossL1.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossBinaryXENT.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossL2.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMultiLabel.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossPoisson.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossCosineProximity.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossHinge.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMCXENT.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMSLE.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossFMeasure.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
LossMixtureDensity.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
This method returns the gradient of the cost function with respect to the
output from the previous layer.
|
INDArray |
LossWasserstein.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSquaredHinge.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAPE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossKLD.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSparseMCXENT.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL1.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossBinaryXENT.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL2.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMultiLabel.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossPoisson.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossCosineProximity.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossHinge.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMCXENT.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSLE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossFMeasure.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSE.computeGradient(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
Pair<Double,INDArray> |
LossMixtureDensity.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossWasserstein.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossSquaredHinge.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMAPE.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossKLD.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossSparseMCXENT.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossL1.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossBinaryXENT.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossL2.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMultiLabel.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossPoisson.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossCosineProximity.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossHinge.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMCXENT.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossMSLE.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
Pair<Double,INDArray> |
LossFMeasure.computeGradientAndScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossMixtureDensity.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average)
Computes the aggregate score as a sum of all of the individual scores of
each of the labels against each of the outputs of the network.
|
double |
LossWasserstein.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossSquaredHinge.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossMAPE.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossKLD.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossSparseMCXENT.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossL1.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossBinaryXENT.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossL2.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossMAE.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossMultiLabel.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossPoisson.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossCosineProximity.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossHinge.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossMCXENT.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossMSLE.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossFMeasure.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
double |
LossMSE.computeScore(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask,
boolean average) |
INDArray |
LossMixtureDensity.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask)
This method returns the score for each of the given outputs against the
given set of labels.
|
INDArray |
LossWasserstein.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSquaredHinge.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAPE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossKLD.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossSparseMCXENT.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL1.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossBinaryXENT.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL2.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMultiLabel.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossPoisson.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossCosineProximity.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossHinge.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMCXENT.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSLE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossFMeasure.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSE.computeScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
LossMixtureDensity.MixtureDensityComponents |
LossMixtureDensity.extractComponents(INDArray output) |
INDArray |
LossSquaredHinge.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMAPE.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossL1.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
protected INDArray |
LossL2.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMultiLabel.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossPoisson.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossCosineProximity.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossHinge.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
protected INDArray |
LossMCXENT.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
INDArray |
LossMSLE.scoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
protected INDArray |
LossSparseMCXENT.sparseScoreArray(INDArray labels,
INDArray preOutput,
IActivation activationFn,
INDArray mask) |
| Constructor and Description |
|---|
LossBinaryXENT(double clipEps,
INDArray weights)
Binary cross entropy where each the output is
(optionally) weighted/scaled by a fixed scalar value.
|
LossBinaryXENT(INDArray weights)
Binary cross entropy where each the output is
(optionally) weighted/scaled by a fixed scalar value.
|
LossL1(INDArray weights)
L1 loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
LossL2(INDArray weights)
L2 loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
LossMAE(INDArray weights)
Mean Absolute Error loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
LossMAPE(INDArray weights)
Mean Absolute Percentage Error loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
LossMCXENT(double softmaxClipEps,
INDArray weights)
Multi-Class Cross Entropy loss function where each the output is (optionally) weighted/scaled by a fixed scalar value.
|
LossMCXENT(INDArray weights)
Multi-Class Cross Entropy loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
LossMSE(INDArray weights)
Mean Squared Error loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
LossMSLE(INDArray weights)
Mean Squared Logarithmic Error loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
LossNegativeLogLikelihood(INDArray weights) |
LossSparseMCXENT(double softmaxClipEps,
INDArray weights)
Multi-Class Cross Entropy loss function where each the output is (optionally) weighted/scaled by a fixed scalar value.
|
LossSparseMCXENT(INDArray weights)
Multi-Class Cross Entropy loss function where each the output is (optionally) weighted/scaled by a flags scalar value.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
RowVectorDeserializer.deserialize(org.nd4j.shade.jackson.core.JsonParser jsonParser,
org.nd4j.shade.jackson.databind.DeserializationContext deserializationContext)
Deprecated.
|
| Modifier and Type | Method and Description |
|---|---|
void |
RowVectorSerializer.serialize(INDArray array,
org.nd4j.shade.jackson.core.JsonGenerator jsonGenerator,
org.nd4j.shade.jackson.databind.SerializerProvider serializerProvider)
Deprecated.
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Transforms.abs(INDArray ndArray)
Abs function
|
static INDArray |
Transforms.abs(INDArray ndArray,
boolean dup)
Abs function
|
static INDArray |
Transforms.acos(INDArray arr) |
static INDArray |
Transforms.acos(INDArray in,
boolean copy) |
static INDArray |
Transforms.allCosineDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allCosineSimilarities(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allEuclideanDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allManhattanDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.and(INDArray x,
INDArray y) |
static INDArray |
Transforms.asin(INDArray arr) |
static INDArray |
Transforms.asin(INDArray in,
boolean copy) |
static INDArray |
Transforms.atan(INDArray arr) |
static INDArray |
Transforms.atan(INDArray in,
boolean copy) |
static INDArray |
Transforms.atan2(@NonNull INDArray x,
@NonNull INDArray y)
Atan2 operation, new INDArray instance will be returned
Note the order of x and y parameters is opposite to that of
Math.atan2(double, double) |
static INDArray |
Transforms.atanh(INDArray in)
Sin function
|
static INDArray |
Transforms.atanh(INDArray in,
boolean copy)
Sin function
|
static INDArray |
Transforms.ceil(INDArray arr) |
static INDArray |
Transforms.ceil(INDArray in,
boolean copy) |
static INDArray |
Transforms.ceiling(INDArray ndArray)
Binary matrix of whether the number at a given index is greater than
|
static INDArray |
Transforms.ceiling(INDArray ndArray,
boolean copyOnOps)
Ceiling function
|
static INDArray |
Transforms.cos(INDArray in) |
static INDArray |
Transforms.cos(INDArray in,
boolean copy) |
static INDArray |
Transforms.cosh(INDArray in) |
static INDArray |
Transforms.cosh(INDArray in,
boolean copy) |
static INDArray |
Transforms.cross(INDArray x,
INDArray y) |
static INDArray |
Transforms.dot(INDArray x,
INDArray y)
Dot product, new INDArray instance will be returned.
Note that the Nd4J design is different from Numpy. |
static INDArray |
Transforms.elu(INDArray arr) |
static INDArray |
Transforms.elu(INDArray in,
boolean copy) |
static INDArray |
Transforms.eluDerivative(INDArray arr,
INDArray grad) |
static INDArray |
Transforms.eluDerivative(INDArray in,
INDArray grad,
boolean copy) |
static INDArray |
Transforms.eps(INDArray ndArray) |
static INDArray |
Transforms.exp(INDArray ndArray)
Run the exp operation
|
static INDArray |
Transforms.exp(INDArray ndArray,
boolean dup)
Exp function
|
static INDArray |
Transforms.expm1(INDArray ndArray,
boolean dup)
Elementwise exponential - 1 function
|
static INDArray |
Transforms.floor(INDArray ndArray)
Binary matrix of whether the number at a given index is greater than
|
static INDArray |
Transforms.floor(INDArray ndArray,
boolean dup)
Floor function
|
static INDArray |
Transforms.greaterThanOrEqual(INDArray first,
INDArray ndArray)
1 if greater than or equal to 0 otherwise (at each element)
|
static INDArray |
Transforms.greaterThanOrEqual(INDArray first,
INDArray ndArray,
boolean dup)
Eps function
|
static INDArray |
Transforms.hardSigmoid(INDArray arr,
boolean dup) |
static INDArray |
Transforms.hardTanh(INDArray ndArray) |
static INDArray |
Transforms.hardTanh(INDArray ndArray,
boolean dup)
Hard tanh
|
static INDArray |
Transforms.hardTanhDerivative(INDArray ndArray) |
static INDArray |
Transforms.hardTanhDerivative(INDArray ndArray,
boolean dup)
Hard tanh
|
static INDArray |
Transforms.identity(INDArray ndArray) |
static INDArray |
Transforms.identity(INDArray ndArray,
boolean dup)
Identity function
|
static INDArray |
Transforms.isMax(INDArray input) |
static INDArray |
Transforms.isMax(INDArray input,
DataType dataType) |
static INDArray |
Transforms.isMax(INDArray input,
INDArray output) |
static INDArray |
Transforms.leakyRelu(INDArray arr) |
static INDArray |
Transforms.leakyRelu(INDArray in,
boolean copy) |
static INDArray |
Transforms.leakyRelu(INDArray arr,
double cutoff) |
static INDArray |
Transforms.leakyRelu(INDArray in,
double cutoff,
boolean copy) |
static INDArray |
Transforms.leakyReluDerivative(INDArray arr,
double cutoff) |
static INDArray |
Transforms.leakyReluDerivative(INDArray in,
double cutoff,
boolean copy) |
static INDArray |
Transforms.lessThanOrEqual(INDArray first,
INDArray ndArray)
1 if less than or equal to 0 otherwise (at each element)
|
static INDArray |
Transforms.lessThanOrEqual(INDArray first,
INDArray ndArray,
boolean dup)
Eps function
|
static INDArray |
Transforms.log(INDArray ndArray) |
static INDArray |
Transforms.log(INDArray ndArray,
boolean dup)
Log function
|
static INDArray |
Transforms.log(INDArray ndArray,
double base)
Log on arbitrary base
|
static INDArray |
Transforms.log(INDArray ndArray,
double base,
boolean duplicate)
Log on arbitrary base
|
static INDArray |
Transforms.log1p(INDArray ndArray,
boolean dup)
Log of x + 1 function
|
static INDArray |
Transforms.max(INDArray ndArray,
double k)
Maximum function with a scalar
|
static INDArray |
Transforms.max(INDArray ndArray,
double k,
boolean dup)
Maximum function with a scalar
|
static INDArray |
Transforms.max(INDArray first,
INDArray second)
Element wise maximum function between 2 INDArrays
|
static INDArray |
Transforms.max(INDArray first,
INDArray second,
boolean dup)
Element wise maximum function between 2 INDArrays
|
static INDArray |
Transforms.min(INDArray ndArray,
double k)
Maximum function with a scalar
|
static INDArray |
Transforms.min(INDArray ndArray,
double k,
boolean dup)
Minimum function with a scalar
|
static INDArray |
Transforms.min(INDArray first,
INDArray second)
Element wise minimum function between 2 INDArrays
|
static INDArray |
Transforms.min(INDArray first,
INDArray second,
boolean dup)
Element wise minimum function between 2 INDArrays
|
static INDArray |
Transforms.mpow(INDArray in,
int n,
boolean dup)
Raises a square matrix to a power n, which can be positive, negative, or zero.
|
static INDArray |
Transforms.neg(INDArray ndArray)
Returns the negative of an ndarray
|
static INDArray |
Transforms.neg(INDArray ndArray,
boolean dup)
Negative
|
static INDArray |
Transforms.normalizeZeroMeanAndUnitVariance(INDArray toNormalize)
Normalize data to zero mean and unit variance
substract by the mean and divide by the standard deviation
|
static INDArray |
Transforms.not(INDArray x) |
static INDArray |
Transforms.or(INDArray x,
INDArray y) |
static INDArray |
Transforms.pow(INDArray ndArray,
INDArray power)
Element-wise power function - x^y, performed element-wise.
|
static INDArray |
Transforms.pow(INDArray ndArray,
INDArray power,
boolean dup)
Element-wise power function - x^y, performed element-wise
|
static INDArray |
Transforms.pow(INDArray ndArray,
Number power)
Pow function
|
static INDArray |
Transforms.pow(INDArray ndArray,
Number power,
boolean dup)
Pow function
|
static INDArray |
Transforms.relu(INDArray arr) |
static INDArray |
Transforms.relu(INDArray in,
boolean copy) |
static INDArray |
Transforms.relu6(INDArray arr) |
static INDArray |
Transforms.relu6(INDArray in,
boolean copy) |
static INDArray |
Transforms.reverse(INDArray x,
boolean dup) |
static INDArray |
Transforms.round(INDArray ndArray)
Rounding function
|
static INDArray |
Transforms.round(INDArray ndArray,
boolean dup)
Rounding function
|
static INDArray |
Transforms.sigmoid(INDArray ndArray)
Sigmoid function
|
static INDArray |
Transforms.sigmoid(INDArray ndArray,
boolean dup)
Sigmoid function
|
static INDArray |
Transforms.sigmoidDerivative(INDArray ndArray)
Sigmoid function
|
static INDArray |
Transforms.sigmoidDerivative(INDArray ndArray,
boolean dup)
Sigmoid function
|
static INDArray |
Transforms.sign(INDArray toSign)
Signum function of this ndarray
|
static INDArray |
Transforms.sign(INDArray toSign,
boolean dup)
Signum function of this ndarray
|
static INDArray |
Transforms.sin(INDArray in)
Sin function
|
static INDArray |
Transforms.sin(INDArray in,
boolean copy)
Sin function
|
static INDArray |
Transforms.sinh(INDArray in)
Sinh function
|
static INDArray |
Transforms.sinh(INDArray in,
boolean copy)
Sinh function
|
static INDArray |
Transforms.softmax(INDArray arr) |
static INDArray |
Transforms.softmax(INDArray in,
boolean copy) |
static INDArray |
Transforms.softPlus(INDArray arr) |
static INDArray |
Transforms.softPlus(INDArray in,
boolean copy) |
static INDArray |
Transforms.softsign(INDArray arr) |
static INDArray |
Transforms.softsign(INDArray in,
boolean copy) |
static INDArray |
Transforms.softsignDerivative(INDArray arr) |
static INDArray |
Transforms.softsignDerivative(INDArray in,
boolean copy) |
static INDArray |
Transforms.sqrt(INDArray ndArray)
Sqrt function
|
static INDArray |
Transforms.sqrt(INDArray ndArray,
boolean dup)
Sqrt function
|
static INDArray |
Transforms.stabilize(INDArray ndArray,
double k) |
static INDArray |
Transforms.stabilize(INDArray ndArray,
double k,
boolean dup)
Stabilize to be within a range of k
|
static INDArray |
Transforms.step(INDArray arr) |
static INDArray |
Transforms.step(INDArray in,
boolean copy) |
static INDArray |
Transforms.tan(INDArray ndArray)
Element-wise tan function.
|
static INDArray |
Transforms.tan(INDArray ndArray,
boolean dup)
Element-wise tan function.
|
static INDArray |
Transforms.tanh(INDArray ndArray)
Tanh function
|
static INDArray |
Transforms.tanh(INDArray ndArray,
boolean dup)
Tanh function
|
static INDArray |
Transforms.timesOneMinus(INDArray in,
boolean copy)
out = in * (1-in)
|
static INDArray |
Transforms.unitVec(INDArray toScale)
Scale by 1 / norm2 of the matrix
|
static INDArray |
Transforms.xor(INDArray x,
INDArray y) |
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Transforms.abs(INDArray ndArray)
Abs function
|
static INDArray |
Transforms.abs(INDArray ndArray,
boolean dup)
Abs function
|
static INDArray |
Transforms.acos(INDArray arr) |
static INDArray |
Transforms.acos(INDArray in,
boolean copy) |
static INDArray |
Transforms.allCosineDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allCosineDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allCosineSimilarities(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allCosineSimilarities(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allEuclideanDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allEuclideanDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allManhattanDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.allManhattanDistances(@NonNull INDArray d1,
@NonNull INDArray d2,
int... dimensions) |
static INDArray |
Transforms.and(INDArray x,
INDArray y) |
static INDArray |
Transforms.asin(INDArray arr) |
static INDArray |
Transforms.asin(INDArray in,
boolean copy) |
static INDArray |
Transforms.atan(INDArray arr) |
static INDArray |
Transforms.atan(INDArray in,
boolean copy) |
static INDArray |
Transforms.atan2(@NonNull INDArray x,
@NonNull INDArray y)
Atan2 operation, new INDArray instance will be returned
Note the order of x and y parameters is opposite to that of
Math.atan2(double, double) |
static INDArray |
Transforms.atan2(@NonNull INDArray x,
@NonNull INDArray y)
Atan2 operation, new INDArray instance will be returned
Note the order of x and y parameters is opposite to that of
Math.atan2(double, double) |
static INDArray |
Transforms.atanh(INDArray in)
Sin function
|
static INDArray |
Transforms.atanh(INDArray in,
boolean copy)
Sin function
|
protected static long[] |
Transforms.broadcastResultShape(INDArray first,
INDArray second) |
static INDArray |
Transforms.ceil(INDArray arr) |
static INDArray |
Transforms.ceil(INDArray in,
boolean copy) |
static INDArray |
Transforms.ceiling(INDArray ndArray)
Binary matrix of whether the number at a given index is greater than
|
static INDArray |
Transforms.ceiling(INDArray ndArray,
boolean copyOnOps)
Ceiling function
|
static INDArray |
Transforms.cos(INDArray in) |
static INDArray |
Transforms.cos(INDArray in,
boolean copy) |
static INDArray |
Transforms.cosh(INDArray in) |
static INDArray |
Transforms.cosh(INDArray in,
boolean copy) |
static double |
Transforms.cosineDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static double |
Transforms.cosineDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static double |
Transforms.cosineSim(@NonNull INDArray d1,
@NonNull INDArray d2)
Cosine similarity
|
static double |
Transforms.cosineSim(@NonNull INDArray d1,
@NonNull INDArray d2)
Cosine similarity
|
static INDArray |
Transforms.cross(INDArray x,
INDArray y) |
static INDArray |
Transforms.dot(INDArray x,
INDArray y)
Dot product, new INDArray instance will be returned.
Note that the Nd4J design is different from Numpy. |
static INDArray |
Transforms.elu(INDArray arr) |
static INDArray |
Transforms.elu(INDArray in,
boolean copy) |
static INDArray |
Transforms.eluDerivative(INDArray arr,
INDArray grad) |
static INDArray |
Transforms.eluDerivative(INDArray in,
INDArray grad,
boolean copy) |
static INDArray |
Transforms.eps(INDArray ndArray) |
static double |
Transforms.euclideanDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static double |
Transforms.euclideanDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static INDArray |
Transforms.exp(INDArray ndArray)
Run the exp operation
|
static INDArray |
Transforms.exp(INDArray ndArray,
boolean dup)
Exp function
|
static INDArray |
Transforms.expm1(INDArray ndArray,
boolean dup)
Elementwise exponential - 1 function
|
static INDArray |
Transforms.floor(INDArray ndArray)
Binary matrix of whether the number at a given index is greater than
|
static INDArray |
Transforms.floor(INDArray ndArray,
boolean dup)
Floor function
|
static INDArray |
Transforms.greaterThanOrEqual(INDArray first,
INDArray ndArray)
1 if greater than or equal to 0 otherwise (at each element)
|
static INDArray |
Transforms.greaterThanOrEqual(INDArray first,
INDArray ndArray,
boolean dup)
Eps function
|
static double |
Transforms.hammingDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static double |
Transforms.hammingDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static INDArray |
Transforms.hardSigmoid(INDArray arr,
boolean dup) |
static INDArray |
Transforms.hardTanh(INDArray ndArray) |
static INDArray |
Transforms.hardTanh(INDArray ndArray,
boolean dup)
Hard tanh
|
static INDArray |
Transforms.hardTanhDerivative(INDArray ndArray) |
static INDArray |
Transforms.hardTanhDerivative(INDArray ndArray,
boolean dup)
Hard tanh
|
static INDArray |
Transforms.identity(INDArray ndArray) |
static INDArray |
Transforms.identity(INDArray ndArray,
boolean dup)
Identity function
|
static INDArray |
Transforms.isMax(INDArray input) |
static INDArray |
Transforms.isMax(INDArray input,
DataType dataType) |
static INDArray |
Transforms.isMax(INDArray input,
INDArray output) |
static double |
Transforms.jaccardDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static double |
Transforms.jaccardDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static INDArray |
Transforms.leakyRelu(INDArray arr) |
static INDArray |
Transforms.leakyRelu(INDArray in,
boolean copy) |
static INDArray |
Transforms.leakyRelu(INDArray arr,
double cutoff) |
static INDArray |
Transforms.leakyRelu(INDArray in,
double cutoff,
boolean copy) |
static INDArray |
Transforms.leakyReluDerivative(INDArray arr,
double cutoff) |
static INDArray |
Transforms.leakyReluDerivative(INDArray in,
double cutoff,
boolean copy) |
static INDArray |
Transforms.lessThanOrEqual(INDArray first,
INDArray ndArray)
1 if less than or equal to 0 otherwise (at each element)
|
static INDArray |
Transforms.lessThanOrEqual(INDArray first,
INDArray ndArray,
boolean dup)
Eps function
|
static INDArray |
Transforms.log(INDArray ndArray) |
static INDArray |
Transforms.log(INDArray ndArray,
boolean dup)
Log function
|
static INDArray |
Transforms.log(INDArray ndArray,
double base)
Log on arbitrary base
|
static INDArray |
Transforms.log(INDArray ndArray,
double base,
boolean duplicate)
Log on arbitrary base
|
static INDArray |
Transforms.log1p(INDArray ndArray,
boolean dup)
Log of x + 1 function
|
static double |
Transforms.manhattanDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static double |
Transforms.manhattanDistance(@NonNull INDArray d1,
@NonNull INDArray d2) |
static INDArray |
Transforms.max(INDArray ndArray,
double k)
Maximum function with a scalar
|
static INDArray |
Transforms.max(INDArray ndArray,
double k,
boolean dup)
Maximum function with a scalar
|
static INDArray |
Transforms.max(INDArray first,
INDArray second)
Element wise maximum function between 2 INDArrays
|
static INDArray |
Transforms.max(INDArray first,
INDArray second,
boolean dup)
Element wise maximum function between 2 INDArrays
|
static INDArray |
Transforms.min(INDArray ndArray,
double k)
Maximum function with a scalar
|
static INDArray |
Transforms.min(INDArray ndArray,
double k,
boolean dup)
Minimum function with a scalar
|
static INDArray |
Transforms.min(INDArray first,
INDArray second)
Element wise minimum function between 2 INDArrays
|
static INDArray |
Transforms.min(INDArray first,
INDArray second,
boolean dup)
Element wise minimum function between 2 INDArrays
|
static INDArray |
Transforms.mpow(INDArray in,
int n,
boolean dup)
Raises a square matrix to a power n, which can be positive, negative, or zero.
|
static INDArray |
Transforms.neg(INDArray ndArray)
Returns the negative of an ndarray
|
static INDArray |
Transforms.neg(INDArray ndArray,
boolean dup)
Negative
|
static INDArray |
Transforms.normalizeZeroMeanAndUnitVariance(INDArray toNormalize)
Normalize data to zero mean and unit variance
substract by the mean and divide by the standard deviation
|
static INDArray |
Transforms.not(INDArray x) |
static INDArray |
Transforms.or(INDArray x,
INDArray y) |
static INDArray |
Transforms.pow(INDArray ndArray,
INDArray power)
Element-wise power function - x^y, performed element-wise.
|
static INDArray |
Transforms.pow(INDArray ndArray,
INDArray power,
boolean dup)
Element-wise power function - x^y, performed element-wise
|
static INDArray |
Transforms.pow(INDArray ndArray,
Number power)
Pow function
|
static INDArray |
Transforms.pow(INDArray ndArray,
Number power,
boolean dup)
Pow function
|
static INDArray |
Transforms.relu(INDArray arr) |
static INDArray |
Transforms.relu(INDArray in,
boolean copy) |
static INDArray |
Transforms.relu6(INDArray arr) |
static INDArray |
Transforms.relu6(INDArray in,
boolean copy) |
static INDArray |
Transforms.reverse(INDArray x,
boolean dup) |
static INDArray |
Transforms.round(INDArray ndArray)
Rounding function
|
static INDArray |
Transforms.round(INDArray ndArray,
boolean dup)
Rounding function
|
static INDArray |
Transforms.sigmoid(INDArray ndArray)
Sigmoid function
|
static INDArray |
Transforms.sigmoid(INDArray ndArray,
boolean dup)
Sigmoid function
|
static INDArray |
Transforms.sigmoidDerivative(INDArray ndArray)
Sigmoid function
|
static INDArray |
Transforms.sigmoidDerivative(INDArray ndArray,
boolean dup)
Sigmoid function
|
static INDArray |
Transforms.sign(INDArray toSign)
Signum function of this ndarray
|
static INDArray |
Transforms.sign(INDArray toSign,
boolean dup)
Signum function of this ndarray
|
static INDArray |
Transforms.sin(INDArray in)
Sin function
|
static INDArray |
Transforms.sin(INDArray in,
boolean copy)
Sin function
|
static INDArray |
Transforms.sinh(INDArray in)
Sinh function
|
static INDArray |
Transforms.sinh(INDArray in,
boolean copy)
Sinh function
|
static INDArray |
Transforms.softmax(INDArray arr) |
static INDArray |
Transforms.softmax(INDArray in,
boolean copy) |
static INDArray |
Transforms.softPlus(INDArray arr) |
static INDArray |
Transforms.softPlus(INDArray in,
boolean copy) |
static INDArray |
Transforms.softsign(INDArray arr) |
static INDArray |
Transforms.softsign(INDArray in,
boolean copy) |
static INDArray |
Transforms.softsignDerivative(INDArray arr) |
static INDArray |
Transforms.softsignDerivative(INDArray in,
boolean copy) |
static INDArray |
Transforms.sqrt(INDArray ndArray)
Sqrt function
|
static INDArray |
Transforms.sqrt(INDArray ndArray,
boolean dup)
Sqrt function
|
static INDArray |
Transforms.stabilize(INDArray ndArray,
double k) |
static INDArray |
Transforms.stabilize(INDArray ndArray,
double k,
boolean dup)
Stabilize to be within a range of k
|
static INDArray |
Transforms.step(INDArray arr) |
static INDArray |
Transforms.step(INDArray in,
boolean copy) |
static INDArray |
Transforms.tan(INDArray ndArray)
Element-wise tan function.
|
static INDArray |
Transforms.tan(INDArray ndArray,
boolean dup)
Element-wise tan function.
|
static INDArray |
Transforms.tanh(INDArray ndArray)
Tanh function
|
static INDArray |
Transforms.tanh(INDArray ndArray,
boolean dup)
Tanh function
|
static INDArray |
Transforms.timesOneMinus(INDArray in,
boolean copy)
out = in * (1-in)
|
static INDArray |
Transforms.unitVec(INDArray toScale)
Scale by 1 / norm2 of the matrix
|
static INDArray |
Transforms.xor(INDArray x,
INDArray y) |
| Modifier and Type | Method and Description |
|---|---|
void |
OpProfiler.processBlasCall(boolean isGemm,
INDArray... operands) |
OpProfiler.PenaltyCause[] |
OpProfiler.processOperands(INDArray... operands)
This method checks for something somewhere
|
OpProfiler.PenaltyCause[] |
OpProfiler.processOperands(INDArray x,
INDArray y) |
OpProfiler.PenaltyCause[] |
OpProfiler.processOperands(INDArray x,
INDArray y,
INDArray z) |
String |
OpProfiler.processOrders(INDArray... operands) |
| Modifier and Type | Method and Description |
|---|---|
String |
NDArrayStrings.format(INDArray arr) |
String |
NDArrayStrings.format(INDArray arr,
boolean summarize)
Format the given ndarray as a string
|
| Modifier and Type | Field and Description |
|---|---|
protected INDArray |
DeviceLocal.delayedArray |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
DeviceLocalNDArray.get()
This method returns object local to current deviceId
|
static INDArray |
NDArrayUtil.toNDArray(int[] nums)
Deprecated.
|
static INDArray |
NDArrayUtil.toNDArray(int[][] nums)
Deprecated.
|
static INDArray |
NDArrayUtil.toNDArray(long[] nums)
Deprecated.
|
static INDArray |
FeatureUtil.toOutcomeMatrix(int[] index,
long numOutcomes)
Creates an out come vector from the specified inputs
|
static INDArray |
FeatureUtil.toOutcomeVector(long index,
long numOutcomes)
Creates an out come vector from the specified inputs
|
| Modifier and Type | Method and Description |
|---|---|
Boolean |
ND4JTestUtils.EqualsFn.apply(INDArray i1,
INDArray i2) |
Boolean |
ND4JTestUtils.EqualsWithEpsFn.apply(INDArray i1,
INDArray i2) |
static void |
LinAlgExceptions.assertColumns(INDArray n,
INDArray n2) |
static void |
LinAlgExceptions.assertMatrix(INDArray... arr) |
static void |
LinAlgExceptions.assertMatrix(INDArray arr) |
static void |
LinAlgExceptions.assertMultiplies(INDArray nd1,
INDArray nd2)
Asserts matrix multiply rules (columns of left == rows of right or rows of left == columns of right)
|
static void |
LinAlgExceptions.assertRows(INDArray n,
INDArray n2) |
static void |
LinAlgExceptions.assertSameLength(INDArray x,
INDArray z)
Asserts both arrays be the same length
|
static void |
LinAlgExceptions.assertSameLength(INDArray x,
INDArray y,
INDArray z) |
static void |
LinAlgExceptions.assertSameShape(INDArray n,
INDArray n2) |
static void |
LinAlgExceptions.assertSameShape(INDArray x,
INDArray y,
INDArray z) |
static void |
LinAlgExceptions.assertValidNum(INDArray n) |
static void |
LinAlgExceptions.assertVector(INDArray... arr) |
static void |
LinAlgExceptions.assertVector(INDArray arr) |
void |
DeviceLocalNDArray.broadcast(INDArray array)
This method duplicates array, and stores it to all devices
PLEASE NOTE: this method is NOT atomic, so you must be sure no other threads are using this instance during the update
|
static long |
NDArrayMath.lengthPerSlice(INDArray arr)
Return the length of a slice
|
static long |
NDArrayMath.lengthPerSlice(INDArray arr,
int... dimension)
The number of elements in a slice
along a set of dimensions
|
static int |
NDArrayMath.mapIndexOntoTensor(int index,
INDArray arr,
int... rank)
This maps an index of a vector
on to a vector in the matrix that can be used
for indexing in to a tensor
|
static long |
NDArrayMath.mapIndexOntoVector(int index,
INDArray arr)
This maps an index of a vector
on to a vector in the matrix that can be used
for indexing in to a tensor
|
static long |
NDArrayMath.matricesPerSlice(INDArray arr)
The number of vectors
in each slice of an ndarray.
|
static void |
FeatureUtil.normalizeMatrix(INDArray toNormalize) |
static long |
NDArrayMath.numVectors(INDArray arr)
Return the number of vectors for an array
the number of vectors for an array
|
static long |
NDArrayMath.offsetForSlice(INDArray arr,
int slice)
Compute the offset for a given slice
|
static void |
FeatureUtil.scaleByMax(INDArray toScale)
Divides each row by its max
|
static void |
FeatureUtil.scaleMinMax(double min,
double max,
INDArray toScale)
Scales the ndarray columns
to the given min/max values
|
void |
DataSetUtils.showINDArray(int mtLv,
String itemCode,
INDArray INDA,
int digits,
int r_End_I,
int c_End_I)
showINDArray
public void showINDArray( int mtLv, String itemCode, INDArray INDA, int digits, int r_End_I, int c_End_I ) Shows content of INDArray. Shows first rows and than columns. |
void |
DataSetUtils.showINDArray(int mtLv,
String itemCode,
INDArray INDA,
int digits,
int r_End_I,
int c_End_I,
boolean turned)
showINDArray
public void showINDArray( int mtLv, String itemCode, INDArray INDA, int digits, int r_End_I, int c_End_I, boolean turned ) Shows content of INDArray. If turned is false shows first rows and than columns. If turned is true shows first columns and than rows. |
static long |
NDArrayMath.sliceOffsetForTensor(int index,
INDArray arr,
int[] tensorShape)
calculates the offset for a tensor
|
static long |
NDArrayMath.sliceOffsetForTensor(int index,
INDArray arr,
long[] tensorShape) |
static long |
NDArrayMath.tensorsPerSlice(INDArray arr,
int[] tensorShape)
Computes the tensors per slice
given a tensor shape and array
|
static int[] |
NDArrayUtil.toInts(INDArray n)
Deprecated.
|
static long[] |
NDArrayUtil.toLongs(INDArray n)
Deprecated.
|
void |
DeviceLocalNDArray.update(@NonNull INDArray array)
This method updates
PLEASE NOTE: this method is NOT atomic, so you must be sure no other threads are using this instance during the update
|
static long |
NDArrayMath.vectorsPerSlice(INDArray arr)
The number of vectors
in each slice of an ndarray.
|
static long |
NDArrayMath.vectorsPerSlice(INDArray arr,
int... rank)
The number of vectors
in each slice of an ndarray.
|
| Modifier and Type | Method and Description |
|---|---|
static ND4JTestUtils.ComparisonResult |
ND4JTestUtils.validateSerializedArrays(File dir1,
File dir2,
boolean recursive,
BiFunction<INDArray,INDArray,Boolean> evalFn)
Scan the specified directories for matching files (i.e., same path relative to their respective root directories)
and compare the contents using a provided function.
Assumes the saved files represent INDArrays saved with Nd4j.saveBinary(INDArray, File) |
static ND4JTestUtils.ComparisonResult |
ND4JTestUtils.validateSerializedArrays(File dir1,
File dir2,
boolean recursive,
BiFunction<INDArray,INDArray,Boolean> evalFn)
Scan the specified directories for matching files (i.e., same path relative to their respective root directories)
and compare the contents using a provided function.
Assumes the saved files represent INDArrays saved with Nd4j.saveBinary(INDArray, File) |
| Constructor and Description |
|---|
DeviceLocalNDArray(INDArray array) |
DeviceLocalNDArray(INDArray array,
boolean delayedMode) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BaseWorkspaceMgr.castTo(T arrayType,
@NonNull DataType dataType,
@NonNull INDArray toCast,
boolean dupIfCorrectType) |
INDArray |
WorkspaceMgr.castTo(T arrayType,
@NonNull DataType dataType,
@NonNull INDArray toCast,
boolean dupIfCorrectType)
Cast the specified array to the specified datatype.
If the array is already the correct type, the bahaviour depends on the 'dupIfCorrectType' argument. dupIfCorrectType = false && toCast.dataType() == dataType: return input array as-is (unless workspace is wrong) dupIfCorrectType = true && toCast.dataType() == dataType: duplicate the array into the specified workspace |
INDArray |
BaseWorkspaceMgr.create(T arrayType,
@NonNull DataType dataType,
long... shape) |
INDArray |
WorkspaceMgr.create(T arrayType,
DataType dataType,
long... shape)
Create an array in the specified array type's workspace (or detached if none is specified).
|
INDArray |
BaseWorkspaceMgr.create(T arrayType,
@NonNull DataType dataType,
@NonNull long[] shape,
@NonNull char order) |
INDArray |
WorkspaceMgr.create(T arrayType,
DataType dataType,
long[] shape,
char ordering)
Create an array in the specified array type's workspace (or detached if none is specified).
|
INDArray |
BaseWorkspaceMgr.createUninitialized(T arrayType,
DataType dataType,
long... shape) |
INDArray |
WorkspaceMgr.createUninitialized(T arrayType,
DataType dataType,
long... shape)
Create an uninitialized array in the specified array type's workspace (or detached if none is specified).
|
INDArray |
BaseWorkspaceMgr.createUninitialized(T arrayType,
@NonNull DataType dataType,
@NonNull long[] shape,
char order) |
INDArray |
WorkspaceMgr.createUninitialized(T arrayType,
DataType dataType,
long[] shape,
char order)
Create an uninitialized array in the specified array type's workspace (or detached if none is specified).
|
INDArray |
BaseWorkspaceMgr.dup(T arrayType,
@NonNull INDArray toDup) |
INDArray |
WorkspaceMgr.dup(T arrayType,
@NonNull INDArray toDup)
Duplicate the array, where the array is put into the specified array type's workspace (if applicable)
|
INDArray |
BaseWorkspaceMgr.dup(T arrayType,
@NonNull INDArray toDup,
char order) |
INDArray |
WorkspaceMgr.dup(T arrayType,
INDArray toDup,
char order)
Duplicate the array, where the array is put into the specified array type's workspace (if applicable)
|
INDArray |
BaseWorkspaceMgr.leverageTo(T arrayType,
@NonNull INDArray array) |
INDArray |
WorkspaceMgr.leverageTo(T toWorkspace,
INDArray array)
Leverage the array to the specified array type's workspace (or detach if required).
|
INDArray |
BaseWorkspaceMgr.validateArrayLocation(T arrayType,
@NonNull INDArray array,
boolean migrateIfInvalid,
boolean exceptionIfDetached) |
INDArray |
WorkspaceMgr.validateArrayLocation(T arrayType,
INDArray array,
boolean migrateIfInvalid,
boolean exceptionIfDetached)
Validate that the specified array type is actually in the workspace it's supposed to be in
|
| Modifier and Type | Method and Description |
|---|---|
static void |
WorkspaceUtils.assertValidArray(INDArray array,
String msg)
Assert that the specified array is valid, in terms of workspaces: i.e., if it is attached (and not in a circular
workspace), assert that the workspace is open, and that the data is not from an old generation.
|
INDArray |
BaseWorkspaceMgr.castTo(T arrayType,
@NonNull DataType dataType,
@NonNull INDArray toCast,
boolean dupIfCorrectType) |
INDArray |
WorkspaceMgr.castTo(T arrayType,
@NonNull DataType dataType,
@NonNull INDArray toCast,
boolean dupIfCorrectType)
Cast the specified array to the specified datatype.
If the array is already the correct type, the bahaviour depends on the 'dupIfCorrectType' argument. dupIfCorrectType = false && toCast.dataType() == dataType: return input array as-is (unless workspace is wrong) dupIfCorrectType = true && toCast.dataType() == dataType: duplicate the array into the specified workspace |
INDArray |
BaseWorkspaceMgr.dup(T arrayType,
@NonNull INDArray toDup) |
INDArray |
WorkspaceMgr.dup(T arrayType,
@NonNull INDArray toDup)
Duplicate the array, where the array is put into the specified array type's workspace (if applicable)
|
INDArray |
BaseWorkspaceMgr.dup(T arrayType,
@NonNull INDArray toDup,
char order) |
INDArray |
WorkspaceMgr.dup(T arrayType,
INDArray toDup,
char order)
Duplicate the array, where the array is put into the specified array type's workspace (if applicable)
|
INDArray |
BaseWorkspaceMgr.leverageTo(T arrayType,
@NonNull INDArray array) |
INDArray |
WorkspaceMgr.leverageTo(T toWorkspace,
INDArray array)
Leverage the array to the specified array type's workspace (or detach if required).
|
INDArray |
BaseWorkspaceMgr.validateArrayLocation(T arrayType,
@NonNull INDArray array,
boolean migrateIfInvalid,
boolean exceptionIfDetached) |
INDArray |
WorkspaceMgr.validateArrayLocation(T arrayType,
INDArray array,
boolean migrateIfInvalid,
boolean exceptionIfDetached)
Validate that the specified array type is actually in the workspace it's supposed to be in
|
| Modifier and Type | Field and Description |
|---|---|
protected INDArray |
BaseNDArrayList.container |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
NDArrayList.array()
Get a view of the underlying array
relative to the size of the actual array.
|
INDArray |
BaseNDArrayList.array()
Get a view of the underlying array
relative to the size of the actual array.
|
| Constructor and Description |
|---|
NDArrayList(@NonNull INDArray container)
Specify the underlying ndarray for this list.
|
NDArrayList(@NonNull INDArray container,
int size)
Specify the underlying ndarray for this list.
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
Nd4jBase64.fromBase64(String base64)
Create an ndarray from a base 64
representation
|
| Modifier and Type | Method and Description |
|---|---|
static String |
Nd4jBase64.base64String(INDArray arr)
Returns an ndarray
as base 64
|
| Modifier and Type | Method and Description |
|---|---|
static INDArray |
BinarySerde.readFromDisk(File readFrom)
Read an ndarray from disk
|
static INDArray |
BinarySerde.toArray(ByteBuffer buffer)
Create an ndarray
from the unsafe buffer
|
static INDArray |
BinarySerde.toArray(ByteBuffer buffer,
int offset)
Create an ndarray
from the unsafe buffer
|
| Modifier and Type | Method and Description |
|---|---|
protected static Pair<INDArray,ByteBuffer> |
BinarySerde.toArrayAndByteBuffer(ByteBuffer buffer,
int offset)
Create an ndarray and existing bytebuffer
|
| Modifier and Type | Method and Description |
|---|---|
static int |
BinarySerde.byteBufferSizeFor(INDArray arr)
Returns the byte buffer size for the given
ndarray.
|
static void |
BinarySerde.doByteBufferPutCompressed(INDArray arr,
ByteBuffer allocated,
boolean rewind)
Setup the given byte buffer
for serialization (note that this is for compressed INDArrays)
4 bytes for rank
4 bytes for data opType
shape information
codec information
data opType
|
static void |
BinarySerde.doByteBufferPutUnCompressed(INDArray arr,
ByteBuffer allocated,
boolean rewind)
Setup the given byte buffer
for serialization (note that this is for uncompressed INDArrays)
4 bytes int for rank
4 bytes for data opType
shape buffer
data buffer
|
static ByteBuffer |
BinarySerde.toByteBuffer(INDArray arr)
Convert an ndarray to an unsafe buffer
for use by aeron
|
static void |
BinarySerde.writeArrayToDisk(INDArray arr,
File toWrite)
Write an ndarray to disk in
binary format
|
static void |
BinarySerde.writeArrayToOutputStream(INDArray arr,
OutputStream outputStream)
Write an array to an output stream.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
NDArrayTextDeSerializer.deserialize(org.nd4j.shade.jackson.databind.JsonNode n) |
INDArray |
NDArrayTextDeSerializer.deserialize(org.nd4j.shade.jackson.core.JsonParser jp,
org.nd4j.shade.jackson.databind.DeserializationContext deserializationContext) |
INDArray |
NDArrayDeSerializer.deserialize(org.nd4j.shade.jackson.core.JsonParser jp,
org.nd4j.shade.jackson.databind.DeserializationContext deserializationContext) |
| Modifier and Type | Method and Description |
|---|---|
void |
NDArrayTextSerializer.serialize(INDArray arr,
org.nd4j.shade.jackson.core.JsonGenerator jg,
org.nd4j.shade.jackson.databind.SerializerProvider serializerProvider) |
void |
NDArraySerializer.serialize(INDArray indArray,
org.nd4j.shade.jackson.core.JsonGenerator jsonGenerator,
org.nd4j.shade.jackson.databind.SerializerProvider serializerProvider) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BaseWeightInitScheme.create(DataType dataType,
long... shape) |
INDArray |
WeightInitScheme.create(DataType dataType,
long... shape)
Create the array
|
INDArray |
BaseWeightInitScheme.create(long[] shape,
INDArray paramsView) |
INDArray |
WeightInitScheme.create(long[] shape,
INDArray paramsView)
Create the array
|
abstract INDArray |
BaseWeightInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
protected INDArray |
BaseWeightInitScheme.handleParamsView(INDArray outputArray,
INDArray paramView) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
BaseWeightInitScheme.create(long[] shape,
INDArray paramsView) |
INDArray |
WeightInitScheme.create(long[] shape,
INDArray paramsView)
Create the array
|
abstract INDArray |
BaseWeightInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
protected INDArray |
BaseWeightInitScheme.handleParamsView(INDArray outputArray,
INDArray paramView) |
| Modifier and Type | Method and Description |
|---|---|
INDArray |
NDArraySupplierInitScheme.create(DataType dataType,
long[] shape) |
INDArray |
NDArraySupplierInitScheme.create(long[] shape,
INDArray paramsView) |
INDArray |
XavierUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ConstantInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalFanInInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ZeroInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
IdentityInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
XavierInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
DistributionInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingUniformFanAvgInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
XavierFanInInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ReluUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
UniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalUniformFanInInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
LecunUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalFanAvgInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalFanOutInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalUniformFanOutInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ReluInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
SigmoidUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
OneInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
NDArraySupplierInitScheme.NDArraySupplier.getArr()
An array proxy method.
|
| Modifier and Type | Method and Description |
|---|---|
INDArray |
NDArraySupplierInitScheme.create(long[] shape,
INDArray paramsView) |
INDArray |
XavierUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ConstantInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalFanInInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ZeroInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
IdentityInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
XavierInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
DistributionInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingUniformFanAvgInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
XavierFanInInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ReluUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
UniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalUniformFanInInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
LecunUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalFanAvgInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalFanOutInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
VarScalingNormalUniformFanOutInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
ReluInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
SigmoidUniformInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
INDArray |
OneInitScheme.doCreate(DataType dataType,
long[] shape,
INDArray paramsView) |
| Constructor and Description |
|---|
NDArraySupplierInitScheme(INDArray arr) |
Copyright © 2022. All rights reserved.