FeaturesAutoTSConfig

class datarobotx.FeaturesAutoTSConfig(feature_derivation_window_start=None, feature_derivation_window_end=None, differencing_method=None, exponentially_weighted_moving_alpha=None, default_to_known_in_advance=None, default_to_do_not_derive=None, calendar_id=None, aggregation_type=None, cross_series_group_by_columns=None, segmentation_task_id=None, treat_as_exponential=None, use_supervised_feature_reduction=None, use_cross_series_features=None, periodicities=None, feature_settings=None)

Time series featurization configuration.

Parameters that default to ‘None’ (or are omitted by the user) are overridden to server-side defaults at runtime. Consult the DataRobot REST API and GUI documentation for additional information on each parameter.

Parameters:

• feature_derivation_window_start (int) – For time series projects only. How many timeUnits of the datetimePartitionColumn into the past relative to the forecast point the feature derivation window should begin.

• feature_derivation_window_end (int) – For time series projects only. How many timeUnits of the datetimePartitionColumn into the past relative to the forecast point the feature derivation window should end.

• differencing_method ({'auto', 'none', 'simple', 'seasonal'}) – For time series projects only. Used to specify which differencing method to apply if the data is stationary. For classification problems simple and seasonal are not allowed. Parameter periodicities must be specified if seasonal is chosen. Defaults to auto.

• exponentially_weighted_moving_alpha (float) – Discount factor (alpha) used for exponentially weighted moving features

• default_to_known_in_advance (bool) – For time series projects only. Sets whether all features default to being treated as known in advance features, which are features that are known into the future. Features marked as known in advance must be specified into the future when making predictions. The default is false, all features are not known in advance. Individual features can be set to a value different than the default using the featureSettings parameter. See the Time Series Overview for more context.

• default_to_do_not_derive (bool) – For time series projects only. Sets whether all features default to being treated as do-not-derive features, excluding them from feature derivation. Individual features can be set to a value different than the default by using the featureSettings parameter.

• calendar_id (str) – The ID of the calendar to be used in this project.

• aggregation_type ({'total', 'average'}) – For multiseries projects only. The aggregation type to apply when creating cross-series features.

• cross_series_group_by_columns (list of str) – For multiseries projects with cross-series features enabled only. List of columns (currently of length 1). Setting that indicates how to further split series into related groups. For example, if every series is sales of an individual product, the series group-by could be the product category with values like “men’s clothing”, “sports equipment”, etc.

• segmentation_task_id (str) – Specifies the SegmentationTask that will be used for dividing the project up into multiple segmented projects.

• treat_as_exponential ({'auto', 'never', 'always'}) – For time series projects only. Used to specify whether to treat data as exponential trend and apply transformations like log-transform. For classification problems always is not allowed.

• use_supervised_feature_reduction (bool) – When true, during feature generation DataRobot runs a supervised algorithm feature lists using only qualifying features. Setting false can severely impact autopilot duration, especially for datasets with many features.

• use_cross_series_features (bool) – Indicating if user wants to use cross-series features.

• periodicities (list of FeaturesTSPeriodicityConfig) – A list of periodicities for time series projects only. For classification problems periodicities are not allowed. If this is provided, parameter ‘differencing_method’ will default to ‘seasonal’ if not provided or ‘auto’.

• feature_settings (list of FeaturesTSFeatureSettingConfig) – An array specifying per feature settings. Features can be left unspecified.

See also

DRConfig

Configuration object for DataRobot project and autopilot settings, also includes detailed examples of usage

Attributes:

aggregation_type	For multiseries projects only.
calendar_id	The ID of the calendar to be used in this project.
cross_series_group_by_columns	For multiseries projects with cross-series features enabled only.
default_to_do_not_derive	For time series projects only.
default_to_known_in_advance	For time series projects only.
differencing_method	For time series projects only.
exponentially_weighted_moving_alpha	Discount factor (alpha) used for exponentially weighted moving features.
feature_derivation_window_end	For time series projects only.
feature_derivation_window_start	For time series projects only.
feature_settings	An array specifying per feature settings.
periodicities	A list of periodicities for time series projects only.
segmentation_task_id	Specifies the SegmentationTask that will be used for dividing the project up into multiple segmented projects.
treat_as_exponential	For time series projects only.
use_cross_series_features	Indicating if user wants to use cross-series features.
use_supervised_feature_reduction	When true, during feature generation DataRobot runs a supervised algorithm feature lists using only qualifying features.

Inherited methods:

keys()	rtype: Collection[str]
to_dict()	Return configuration as a dict.

property aggregation_type: str

For multiseries projects only. The aggregation type to apply when creating cross-series features.

Notes

aggregation_type : {‘total’, ‘average’}

property calendar_id: str

The ID of the calendar to be used in this project.

Notes

calendar_id : str

property cross_series_group_by_columns: List[str]

For multiseries projects with cross-series features enabled only. List of columns (currently of length 1). Setting that indicates how to further split series into related groups. For example, if every series is sales of an individual product, the series group-by could be the product category with values like “men’s clothing”, “sports equipment”, etc.

Notes

cross_series_group_by_columns : list of str

property default_to_do_not_derive: bool

For time series projects only. Sets whether all features default to being treated as do-not-derive features, excluding them from feature derivation. Individual features can be set to a value different than the default by using the featureSettings parameter.

Notes

default_to_do_not_derive : bool

property default_to_known_in_advance: bool

For time series projects only. Sets whether all features default to being treated as known in advance features, which are features that are known into the future. Features marked as known in advance must be specified into the future when making predictions. The default is false, all features are not known in advance. Individual features can be set to a value different than the default using the featureSettings parameter. See the Time Series Overview for more context.

Notes

default_to_known_in_advance : bool

property differencing_method: str

For time series projects only. Used to specify which differencing method to apply if the data is stationary. For classification problems simple and seasonal are not allowed. Parameter periodicities must be specified if seasonal is chosen. Defaults to auto.

Notes

differencing_method : {‘auto’, ‘none’, ‘simple’, ‘seasonal’}

property exponentially_weighted_moving_alpha: float

Discount factor (alpha) used for exponentially weighted moving features.

Notes

exponentially_weighted_moving_alpha : float

property feature_derivation_window_end: int

For time series projects only. How many timeUnits of the datetimePartitionColumn into the past relative to the forecast point the feature derivation window should end.

Notes

feature_derivation_window_end : int

property feature_derivation_window_start: int

For time series projects only. How many timeUnits of the datetimePartitionColumn into the past relative to the forecast point the feature derivation window should begin.

Notes

feature_derivation_window_start : int

property feature_settings: List[FeaturesTSFeatureSettingConfig]

An array specifying per feature settings. Features can be left unspecified.

Notes

feature_settings : list of FeaturesTSFeatureSettingConfig

property periodicities: List[FeaturesTSPeriodicityConfig]

A list of periodicities for time series projects only. For classification problems periodicities are not allowed. If this is provided, parameter ‘differencing_method’ will default to ‘seasonal’ if not provided or ‘auto’.

Notes

periodicities : list of FeaturesTSPeriodicityConfig

property segmentation_task_id: str

Specifies the SegmentationTask that will be used for dividing the project up into multiple segmented projects.

Notes

segmentation_task_id : str

to_dict()

Return configuration as a dict.

Return type:

Dict[str, Any]

property treat_as_exponential: str

For time series projects only. Used to specify whether to treat data as exponential trend and apply transformations like log-transform. For classification problems always is not allowed.

Notes

treat_as_exponential : {‘auto’, ‘never’, ‘always’}

property use_cross_series_features: bool

Indicating if user wants to use cross-series features.

Notes

use_cross_series_features : bool

property use_supervised_feature_reduction: bool

When true, during feature generation DataRobot runs a supervised algorithm feature lists using only qualifying features. Setting false can severely impact autopilot duration, especially for datasets with many features.

Notes

use_supervised_feature_reduction : bool