io.github.timsetsfire.datarobot
class used to represent DateTime models.
class used to represent DateTime models. These should not be created directly.
How many timeUnits into the past relative to the forecast point the user needs to provide history for at prediction time. This can differ from the featureDerivationWindowStart set on the project due to the differencing method and period selected, or if the model is a time series native model such as ARIMA. Will be a negative integer.
an array of information on each backtesting fold of the model
– the duration spanned by the dates in the partition column for the data used to train the model
– either “duration”, “rowCount”, or “selectedDateRange”. Identifies which of trainingDuration, trainingRowCount, or trainingStartDate and train- ingEndDate define the training size of the model when making predictions and scoring.
This is the ID of the parent model. Otherwise Null.
– the status of the holdout score. Either “COMPLETED”, “INSUFFI- CIENT_DATA” or “HOLDOUT_BOUNDARIES_EXCEEDED”
Indicates which unit is the basis for the feature derivation window and the forecast window. Will be either detected time unit or “ROW”.
How many timeUnits into the future relative to the forecast point the forecast window should start. Will be a non-negative integer.
– an integer between 1 and 99, indicating the percentage of sampling within the time window. The points kept are determined by samplingMethod (random uniform by default). Will be null if no sampling was specified.
(string) – string (New in version 2.20). Either ‘random’ or ‘latest’, indicates sampling method used to select training data. For row-based project this is the way how requested number of rows are selected. For other projects (duration-based, start/end, project settings) - how specified percent of rows (timeWindowSamplePct) is selected from specified time window.
How many timeUnits into the past relative to the forecast point the feature derivation window should end. Will be a non-positive integer.
– json object describing the holdout and prediction training data as de- scribed below
How many timeUnits into the future relative to the forecast point the forecast window should end. Will be a non-negative integer.
- identifies the model, e.g. Nystroem Kernel SVM Regressor blueprintId – the blueprint used to construct the model - note this is not an ObjectId
–boolean,whether this model supports enforcing montonic constraints
–the ID of the model
– the ID of the project to which the model belongs
– boolean, indicating whether the model is frozen, i.e. uses tuning parameters from a parent model
– the ID of the featurelist used by the model
– the number of rows used to train the model
– the end date of the dates in the partition column for the data used to train the model
– always null for datetime models
–indicateswhatkindofmodelitis-willbeprimeforDataRobotPrime models, blend for blender models, and model for all other models
– the start date of the dates in the partition column for the data used to train the model
– the performance of the model according to ous metrics, see below modelType –
the ID of the featurelist that defines the set of features with a monotonically increasing relationship to the target. If null, no such constraints are enforced.
– the holdout score of the model according to the project metric, if the score is available and the holdout is unlocked
threshold used for binary classification in predictions.
– a json list of processes used by the model
– the name of the featurelist used by the model
indicates whether modification of a predictions threshold is forbidden. Threshold modification is forbidden once a model has had a deployment created or predictions made via the dedicated prediction API.
the ID of the featurelist that defines the set of features with a monotonically decreasing relationship to the target. If null, no such constraints are enforced.
the blueprint used to construct the model - note this is not an ObjectId
the blueprint used to construct the model - note this is not an ObjectId
the ID of the featurelist used by the model
the ID of the featurelist used by the model
the name of the featurelist used by the model
the name of the featurelist used by the model
Model ID
Model ID
boolean, indicating whether the model is frozen, i.e.
boolean, indicating whether the model is frozen, i.e. uses tuning parameters from a parent model
(bool) (New in version v2.13) whether the model has been starred
(bool) (New in version v2.13) whether the model has been starred
the performance of the model according to various metrics, see below
the performance of the model according to various metrics, see below
indicateswhatkindofmodelitis -willbeprimeforDataRobotPrime models, blend for blender models, scaleout for scaleout models, and model for all other models
indicateswhatkindofmodelitis -willbeprimeforDataRobotPrime models, blend for blender models, scaleout for scaleout models, and model for all other models
identifies the model, e.g.
identifies the model, e.g. Nystroem Kernel SVM Regressor
(new in v2.11) null or str, the ID of the featurelist that defines the set of features with a monotonically decreasing relationship to the target.
(new in v2.11) null or str, the ID of the featurelist that defines the set of features with a monotonically decreasing relationship to the target. If null, no such constraints are enforced.
(new in v2.11) null or str, the ID of the featurelist that defines the set of features with a monotonically increasing relationship to the target.
(new in v2.11) null or str, the ID of the featurelist that defines the set of features with a monotonically increasing relationship to the target. If null, no such constraints are enforced.
(float) (New in version v2.13) threshold used for binary classification in predictions.
(float) (New in version v2.13) threshold used for binary classification in predictions.
(boolean) (New in version v2.13) indicates whether modification of a predictions threshold is forbidden.
(boolean) (New in version v2.13) indicates whether modification of a predictions threshold is forbidden. Threshold modification is forbidden once a model has had a deployment created or predictions made via the dedicated prediction API.
a json list of processes used by the model
a json list of processes used by the model
the ID of the project to which the model belongs
the ID of the project to which the model belongs
the percentage of the dataset used in training the model
the percentage of the dataset used in training the model
(new in v2.11) boolean, whether this model supports enforcing montonic constraints
(new in v2.11) boolean, whether this model supports enforcing montonic constraints
the duration spanned by the dates in the partition column for the data used to train the model
the duration spanned by the dates in the partition column for the data used to train the model
the end date of the dates in the partition column for the data used to train the model
the end date of the dates in the partition column for the data used to train the model
the number of rows used to train the model
the number of rows used to train the model
the start date of the dates in the partition column for the data used to train the model
the start date of the dates in the partition column for the data used to train the model
(Since version ) see corresponding Javadoc for more information.
DateTimeModel object