TargetAutoMLConfig

class datarobotx.TargetAutoMLConfig(positive_class=None, weights=None, offset=None, exposure=None, response_cap=None, events_count=None, preferable_target_value=None, class_mapping_aggregation_settings=None, model_group_id=None, model_regime_id=None, model_baselines=None, chunk_definition_id=None)

AutoML additional target options.

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:

• positive_class (str or int or float) – A value from the target column to use for the positive class. May only be specified for projects doing binary classification.If not specified, a positive class is selected automatically.

• weights (str) – The name of a column specifying row weights. The data in this column must be pure numeric (e.g. not currency, date, length, etc.) and without missing values

• offset (list of str) – An array of strings with names of a columns specifying row offsets.The data in this column must be pure numeric (e.g. not currency, date, length, etc.) and without missing values

• exposure (str) – The name of a column specifying row exposure.The data in this column must be pure numeric (e.g. not currency, date, length, etc.) and without missing values

• response_cap (float) – Used to cap the maximum response of a model

• events_count (str) – The name of a column specifying events count. The data in this column must be pure numeric and non negative without missing values

• preferable_target_value (str or int or float) – A target value that should be treated as a positive outcome for the prediction. For example if we want to check gender discrimination for giving a loan and our target named is_bad, then the positive outcome for the prediction would be No, which means that the loan is good and that’s what we treat as a preferable result for the loaner. Used and required only if Bias & Fairness in AutoML feature is enabled.

• class_mapping_aggregation_settings (TargetAggregationConfig) – Class mapping aggregation settings.

• model_group_id (str) – The name of a column specifying row model group id.

• model_regime_id (str) – The name of a column specifying row model regime id.

• model_baselines (list of str) – An array of strings with names of a columns specifying row offsets.The data in this column must be pure numeric (e.g. not currency, date, length, etc.).

• chunk_definition_id (string) – Unique definition for chunks needed to run automated incremental learning.

See also

DRConfig

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

Attributes:

chunk_definition_id	The Mongo ID that uniquely defines chunk definitions needed for incremental learning.
class_mapping_aggregation_settings	Class mapping aggregation settings.
events_count	The name of a column specifying events count.
exposure	The name of a column specifying row exposure.The data in this column must be pure numeric (e.g.
model_baselines	An array of strings with names of a columns specifying row model baselines.The data in this column must be pure numeric (e.g.
model_group_id	The name of a column specifying row model group id.
model_regime_id	The name of a column specifying row model regime id.
offset	An array of strings with names of a columns specifying row offsets.The data in this column must be pure numeric (e.g.
positive_class	A value from the target column to use for the positive class.
preferable_target_value	A target value that should be treated as a positive outcome for the prediction.
response_cap	Used to cap the maximum response of a model.
weights	The name of a column specifying row weights.

Inherited methods:

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

property chunk_definition_id: str

The Mongo ID that uniquely defines chunk definitions needed for incremental learning.

Notes

chunk_definition_id : str

property class_mapping_aggregation_settings: TargetAggregationConfig

Class mapping aggregation settings.

Notes

class_mapping_aggregation_settings : TargetAggregationConfig

property events_count: str

The name of a column specifying events count. The data in this column must be pure numeric and non negative without missing values.

Notes

events_count : str

property exposure: str

The name of a column specifying row exposure.The data in this column must be pure numeric (e.g. not currency, date, length, etc.) and without missing values.

Notes

exposure : str

property model_baselines: List[str]

An array of strings with names of a columns specifying row model baselines.The data in this column must be pure numeric (e.g. not currency, date, length, etc.).

Notes

model_baselines : list of str

property model_group_id: str

The name of a column specifying row model group id.

Notes

model_group_id : str

property model_regime_id: str

The name of a column specifying row model regime id.

Notes

model_regime_id : str

property offset: List[str]

An array of strings with names of a columns specifying row offsets.The data in this column must be pure numeric (e.g. not currency, date, length, etc.) and without missing values.

Notes

offset : list of str

property positive_class: str | int | float

A value from the target column to use for the positive class. May only be specified for projects doing binary classification.If not specified, a positive class is selected automatically.

Notes

positive_class : str or int or float

property preferable_target_value: str | int | float

A target value that should be treated as a positive outcome for the prediction. For example if we want to check gender discrimination for giving a loan and our target named is_bad, then the positive outcome for the prediction would be No, which means that the loan is good and that’s what we treat as a preferable result for the loaner. Used and required only if Bias & Fairness in AutoML feature is enabled.

Notes

preferable_target_value : str or int or float

property response_cap: float

Used to cap the maximum response of a model.

Notes

response_cap : float

to_dict()

Return configuration as a dict.

Return type:

Dict[str, Any]

property weights: str

The name of a column specifying row weights. The data in this column must be pure numeric (e.g. not currency, date, length, etc.) and without missing values.

Notes

weights : str