io.improbable.keanu.vertices.dbl.nonprobabilistic.diff

Class PartialDerivative

java.lang.Object

io.improbable.keanu.vertices.dbl.nonprobabilistic.diff.PartialDerivative

public class PartialDerivative
extends java.lang.Object

Field Summary

Fields

Modifier and Type	Field and Description
static PartialDerivative	EMPTY

Constructor Summary

Constructors

Constructor and Description
PartialDerivative(DoubleTensor partial)

Method Summary

Modifier and Type	Method and Description
PartialDerivative	add(PartialDerivative addition)
PartialDerivative	divideByAlongOfDimensions(DoubleTensor divisor)
PartialDerivative	divideByAlongOfDimensions(DoubleTensor divisor, int partialOfRank)
DoubleTensor	get()
long[]	getOfShape(long[] wrtShape)
long[]	getWrtShape(long[] ofShape)
boolean	isPresent()
static PartialDerivative	matrixMultiplyAlongOfDimensions(PartialDerivative partial, DoubleTensor multiplier, boolean partialIsLeft)
static PartialDerivative	matrixMultiplyAlongWrtDimensions(PartialDerivative partial, DoubleTensor multiplier, boolean partialIsLeft)
PartialDerivative	multiplyAlongOfDimensions(DoubleTensor multiplier) This method assumes the partial 'of' rank is the same as the multiplier.
PartialDerivative	multiplyAlongOfDimensions(DoubleTensor multiplier, int partialOfRank)
PartialDerivative	multiplyAlongWrtDimensions(DoubleTensor multiplier)
PartialDerivative	multiplyBy(double multiplier)
PartialDerivative	subtract(PartialDerivative subtraction)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

EMPTY

public static final PartialDerivative EMPTY

Constructor Detail

PartialDerivative

public PartialDerivative(DoubleTensor partial)

Method Detail

isPresent

public boolean isPresent()

get

public DoubleTensor get()

getOfShape

public long[] getOfShape(long[] wrtShape)

getWrtShape

public long[] getWrtShape(long[] ofShape)

add

public PartialDerivative add(PartialDerivative addition)

subtract

public PartialDerivative subtract(PartialDerivative subtraction)

multiplyBy

public PartialDerivative multiplyBy(double multiplier)

multiplyAlongOfDimensions

public PartialDerivative multiplyAlongOfDimensions(DoubleTensor multiplier)

This method assumes the partial 'of' rank is the same as the multiplier. This is usually the case except for some broadcast operations.

Parameters:

multiplier - the value to multiply by

Returns:

a partial derivative with of dimensions multiplied

multiplyAlongOfDimensions

public PartialDerivative multiplyAlongOfDimensions(DoubleTensor multiplier,
                                                   int partialOfRank)

Parameters:

multiplier - the value to multiply by

partialOfRank - the rank of the 'of' part of the partial. This is needed if it is different from the rank of the multiplier. This happens for rank changing broadcast ops.

Returns:

a partial derivative with of dimensions multiplied

divideByAlongOfDimensions

public PartialDerivative divideByAlongOfDimensions(DoubleTensor divisor)

divideByAlongOfDimensions

public PartialDerivative divideByAlongOfDimensions(DoubleTensor divisor,
                                                   int partialOfRank)

multiplyAlongWrtDimensions

public PartialDerivative multiplyAlongWrtDimensions(DoubleTensor multiplier)

matrixMultiplyAlongOfDimensions

public static PartialDerivative matrixMultiplyAlongOfDimensions(PartialDerivative partial,
                                                                DoubleTensor multiplier,
                                                                boolean partialIsLeft)

matrixMultiplyAlongWrtDimensions

public static PartialDerivative matrixMultiplyAlongWrtDimensions(PartialDerivative partial,
                                                                 DoubleTensor multiplier,
                                                                 boolean partialIsLeft)