gluonts.ext.statsforecast module#

class gluonts.ext.statsforecast.ADIDAPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the ADIDA model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of ADIDA

class gluonts.ext.statsforecast.AutoARIMAPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the AutoARIMA model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of AutoARIMA

class gluonts.ext.statsforecast.AutoCESPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the AutoCES model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of AutoCES

class gluonts.ext.statsforecast.AutoETSPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the AutoETS model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of AutoETS

class gluonts.ext.statsforecast.AutoThetaPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the AutoTheta model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of AutoTheta

class gluonts.ext.statsforecast.CrostonClassicPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the CrostonClassic model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of CrostonClassic

class gluonts.ext.statsforecast.CrostonOptimizedPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the CrostonOptimized model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of CrostonOptimized

class gluonts.ext.statsforecast.CrostonSBAPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the CrostonSBA model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of CrostonSBA

class gluonts.ext.statsforecast.DynamicOptimizedThetaPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the DynamicOptimizedTheta model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of DynamicOptimizedTheta

class gluonts.ext.statsforecast.DynamicThetaPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the DynamicTheta model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of DynamicTheta

class gluonts.ext.statsforecast.HistoricAveragePredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the HistoricAverage model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of HistoricAverage

class gluonts.ext.statsforecast.HoltPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the Holt model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of Holt

class gluonts.ext.statsforecast.HoltWintersPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the HoltWinters model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of HoltWinters

class gluonts.ext.statsforecast.IMAPAPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the IMAPA model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of IMAPA

class gluonts.ext.statsforecast.MSTLPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the MSTL model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of MSTL

class gluonts.ext.statsforecast.ModelConfig(quantile_levels: Optional[List[float]] = None)[source]#

Bases: object

forecast_keys: List[str]#
intervals: Optional[List[int]]#
quantile_levels: Optional[List[float]] = None#
statsforecast_keys: List[str]#
class gluonts.ext.statsforecast.NaivePredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the Naive model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of Naive

class gluonts.ext.statsforecast.OptimizedThetaPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the OptimizedTheta model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of OptimizedTheta

class gluonts.ext.statsforecast.RandomWalkWithDriftPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the RandomWalkWithDrift model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of RandomWalkWithDrift

class gluonts.ext.statsforecast.SeasonalExponentialSmoothingOptimizedPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the SeasonalExponentialSmoothingOptimized model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of SeasonalExponentialSmoothingOptimized

class gluonts.ext.statsforecast.SeasonalExponentialSmoothingPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the SeasonalExponentialSmoothing model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of SeasonalExponentialSmoothing

class gluonts.ext.statsforecast.SeasonalNaivePredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the SeasonalNaive model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of SeasonalNaive

class gluonts.ext.statsforecast.SeasonalWindowAveragePredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the SeasonalWindowAverage model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of SeasonalWindowAverage

class gluonts.ext.statsforecast.SimpleExponentialSmoothingOptimizedPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the SimpleExponentialSmoothingOptimized model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of SimpleExponentialSmoothingOptimized

class gluonts.ext.statsforecast.SimpleExponentialSmoothingPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the SimpleExponentialSmoothing model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of SimpleExponentialSmoothing

class gluonts.ext.statsforecast.StatsForecastPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: RepresentablePredictor

A predictor type that wraps models from the statsforecast package.

This class is used via subclassing and setting the ModelType class attribute to specify the statsforecast model type to use.

Parameters:
  • prediction_length – Prediction length for the model to use.

  • quantile_levels – Optional list of quantile levels that we want predictions for. Note: this is only supported by specific types of models, such as AutoARIMA. By default this is None, giving only the mean prediction.

  • **model_params – Keyword arguments to be passed to the model type for construction. The specific arguments accepted or required depend on the ModelType; please refer to the documentation of statsforecast for details.

ModelType: Type#
predict_item(entry: Dict[str, Any]) QuantileForecast[source]#
class gluonts.ext.statsforecast.TSBPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the TSB model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of TSB

class gluonts.ext.statsforecast.ThetaPredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the Theta model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of Theta

class gluonts.ext.statsforecast.WindowAveragePredictor(prediction_length: int, quantile_levels: Optional[List[float]] = None, **model_params)[source]#

Bases: StatsForecastPredictor

A predictor wrapping the WindowAverage model from statsforecast.

See StatsForecastPredictor for the list of arguments.

ModelType#

alias of WindowAverage