ts_benchmark.baselines.darts package

The following module is the adapter required for embedding the darts models into TFB.

ts_benchmark.baselines.darts.darts_models module

Functions

`darts_deep_model_adapter`(model_class)	Adapts a Darts deep model class to OTB protocol
`darts_regression_model_adapter`(model_class)	Adapts a Darts regression model class to OTB protocol
`darts_statistical_model_adapter`(model_class)	Adapts a Darts statistical model class to OTB protocol

Classes

`DartsConfig`(**kwargs)
`DartsModelAdapter`(model_class, model_args[, ...])	Darts model adapter class

class DartsConfig(**kwargs)[source]

Bases: object

get(key: str, default: Any | None = None) → Any[source]

get_darts_class_params() → dict[source]

class DartsModelAdapter(model_class: type, model_args: dict, model_name: str | None = None, allow_fit_on_eval: bool = False, supports_validation: bool = False, **kwargs)[source]

Bases: ModelBase

Darts model adapter class

Adapts Darts models to OTB forecasting interface.

forecast(horizon: int, series: pandas.DataFrame, **kwargs) → numpy.ndarray[source]

Use the adapted Darts model for prediction.

Parameters:

horizon – Forecast length.
series – Time series data to make inferences on.

Returns:

Forecast result.

forecast_fit(train_data: pandas.DataFrame, *, train_ratio_in_tv: float = 1.0, **kwargs) → ModelBase[source]

Fit a suitable Darts model on time series data.

Parameters:

train_data – Time series data.
train_ratio_in_tv – Represents the splitting ratio of the training set validation set. If it is equal to 1, it means that the validation set is not partitioned.

Returns:

The fitted model object.

property model_name: Returns the name of the model.

darts_deep_model_adapter(model_class: type) → Dict[source]

Adapts a Darts deep model class to OTB protocol

Parameters:: model_class – a class of deep forecasting model from Darts library.
Returns:: model factory that follows the OTB protocol.

darts_regression_model_adapter(model_class: type) → Dict[source]

Adapts a Darts regression model class to OTB protocol

Parameters:: model_class – a class of regression forecasting model from Darts library.
Returns:: model factory that follows the OTB protocol.

darts_statistical_model_adapter(model_class: type) → Dict[source]

Adapts a Darts statistical model class to OTB protocol

Parameters:: model_class – a class of statistical forecasting model from Darts library.
Returns:: model factory that follows the OTB protocol.