tempor.data.predictive module¶

Module defining the PredictiveTaskData class and its subclasses, which are used to store the data components relevant for different predictive tasks (e.g. prediction, time-to-event analysis, treatment effects).

class tempor.data.predictive.PredictiveTaskData(parent_dataset: PredictiveDataset, targets: Any, treatments: Any | None, **kwargs: Any)[source]¶

Bases: ABC

The predictive task data abstract base class.

Parameters:¶

parent_dataset : PredictiveDataset¶: The parent predictive dataset object.
targets : Any¶: The targets data.
treatments : Optional[Any]¶: The treatments data.
**kwargs : Any: Additional keyword arguments. Currently unused.

abstract property predictive_task : PredictiveTask¶

Return the predictive task enum value corresponding to the class.

Returns:¶: The predictive task enum value.
Return type:¶: data_typing.PredictiveTask

property targets : DataSamples | None¶

The property containing the targets data.

Returns:¶: The targets data.
Return type:¶: Optional[samples.DataSamples]

property treatments : DataSamples | None¶

The property containing the treatments data.

Returns:¶: The treatments data.
Return type:¶: Optional[samples.DataSamples]

class tempor.data.predictive.OneOffPredictionTaskData(parent_dataset: PredictiveDataset, targets: DataFrame | ndarray | None, **kwargs: Any)[source]¶

Bases: PredictiveTaskData

The one-off prediction task data class.

Parameters:¶

parent_dataset : PredictiveDataset¶: The parent predictive dataset object.
targets : Optional[data_typing.DataContainer]¶: The targets data.
**kwargs : Any: Additional keyword arguments. Currently unused.

property predictive_task : PredictiveTask¶

Return the predictive task enum value corresponding to the class. Here, ONE_OFF_PREDICTION.

Returns:¶: The predictive task enum value. Here, ONE_OFF_PREDICTION.
Return type:¶: data_typing.PredictiveTask

class tempor.data.predictive.TemporalPredictionTaskData(parent_dataset: PredictiveDataset, targets: DataFrame | ndarray | None, **kwargs: Any)[source]¶

Bases: PredictiveTaskData

The temporal prediction task data class.

Parameters:¶

parent_dataset : PredictiveDataset¶: The parent predictive dataset object.
targets : Optional[data_typing.DataContainer]¶: The targets data.
**kwargs : Any: Additional keyword arguments. Currently unused.

property predictive_task : PredictiveTask¶

Return the predictive task enum value corresponding to the class. Here, TEMPORAL_PREDICTION.

Returns:¶: The predictive task enum value. Here, TEMPORAL_PREDICTION.
Return type:¶: data_typing.PredictiveTask

class tempor.data.predictive.TimeToEventAnalysisTaskData(parent_dataset: PredictiveDataset, targets: DataFrame | ndarray | None, **kwargs: Any)[source]¶

Bases: PredictiveTaskData

The time-to-event analysis task data class.

Parameters:¶

parent_dataset : PredictiveDataset¶: The parent predictive dataset object.
targets : Optional[data_typing.DataContainer]¶: The targets data.
**kwargs : Any: Additional keyword arguments. Currently unused.

property predictive_task : PredictiveTask¶

Return the predictive task enum value corresponding to the class. Here, TIME_TO_EVENT_ANALYSIS.

Returns:¶: The predictive task enum value. Here, TIME_TO_EVENT_ANALYSIS.
Return type:¶: data_typing.PredictiveTask

class tempor.data.predictive.OneOffTreatmentEffectsTaskData(parent_dataset: PredictiveDataset, targets: DataFrame | ndarray | None, treatments: DataFrame | ndarray, **kwargs: Any)[source]¶

Bases: PredictiveTaskData

The one-off treatment effects task data class.

Parameters:¶

parent_dataset : PredictiveDataset¶: The parent predictive dataset object.
targets : Optional[data_typing.DataContainer]¶: The targets data.
treatments : data_typing.DataContainer¶: The treatments data.
**kwargs : Any: Additional keyword arguments. Currently unused.

property predictive_task : PredictiveTask¶

Return the predictive task enum value corresponding to the class. Here, ONE_OFF_TREATMENT_EFFECTS.

Returns:¶: The predictive task enum value. Here, ONE_OFF_TREATMENT_EFFECTS.
Return type:¶: data_typing.PredictiveTask

class tempor.data.predictive.TemporalTreatmentEffectsTaskData(parent_dataset: PredictiveDataset, targets: DataFrame | ndarray | None, treatments: DataFrame | ndarray, **kwargs: Any)[source]¶

Bases: PredictiveTaskData

The temporal treatment effects task data class.

Parameters:¶

parent_dataset : PredictiveDataset¶: The parent predictive dataset object.
targets : Optional[data_typing.DataContainer]¶: The targets data.
treatments : data_typing.DataContainer¶: The treatments data.
**kwargs : Any: Additional keyword arguments. Currently unused.

property predictive_task : PredictiveTask¶

Return the predictive task enum value corresponding to the class. Here, TEMPORAL_TREATMENT_EFFECTS.

Returns:¶: The predictive task enum value. Here, TEMPORAL_TREATMENT_EFFECTS.
Return type:¶: data_typing.PredictiveTask