DummyForecaster

class openstef_beam.backtesting.backtest_forecaster.dummy_forecaster.DummyForecaster(config: BacktestForecasterConfig | None = None, predict_quantiles: list[Quantile] | None = None) → None

Bases: BacktestForecasterMixin

Simple forecaster implementation for testing and development purposes.

Provides a minimal implementation of the backtesting forecaster interface without actual prediction logic. Useful for testing pipeline components, debugging, and as a reference implementation.

Variables:

• config – Configuration parameters for the forecasting interface.

• predict_quantiles – List of quantiles to return in predictions.

Parameters:

• config (BacktestForecasterConfig | None)

• predict_quantiles (list[Quantile] | None)

__init__(config: BacktestForecasterConfig | None = None, predict_quantiles: list[Quantile] | None = None) → None

Initialize the dummy forecaster with default configuration.

Parameters:

• config (BacktestForecasterConfig | None) – Forecaster configuration. If None, uses default no-training config.

• predict_quantiles (list[Quantile] | None) – Quantiles to include in predictions. If None, uses standard quantiles [0.05, 0.1, 0.3, 0.5, 0.7, 0.9, 0.95].

• config

• predict_quantiles

property quantiles: list[Quantile]

Return the list of quantiles for predictions.

predict(data: RestrictedHorizonVersionedTimeSeries) → TimeSeriesDataset | None

Placeholder prediction method that raises NotImplementedError.

Parameters:

• data (RestrictedHorizonVersionedTimeSeries) – Time series data with horizon restrictions for prediction.

• data

Returns:

None - this implementation does not provide actual predictions.

Raises:

NotImplementedError – Always raised as this is a dummy implementation.

Return type:

TimeSeriesDataset | None