gluonts.model.seasonal_naive package

class gluonts.model.seasonal_naive.SeasonalNaivePredictor(prediction_length: int, season_length: Union[int, Callable], imputation_method: gluonts.transform.feature.MissingValueImputation = gluonts.transform.feature.LastValueImputation())

Bases: gluonts.model.predictor.RepresentablePredictor

Seasonal naïve forecaster.

For each time series $y$, this predictor produces a forecast $\tilde{y}(T+k)=y(T+k-h)$, where $T$ is the forecast time, $k=0,...,$ prediction_length - 1, and $h=$ season_length.

If prediction_length > season_length, then the season is repeated multiple times. If a time series is shorter than season_length, then the mean observed value is used as prediction.

Parameters

• prediction_length – Number of time points to predict.

• season_length – Seasonality used to make predictions. If this is an integer, then a fixed sesasonlity is applied; if this is a function, then it will be called on each given entry’s freq attribute of the "start" field, and will return the seasonality to use.

• imputation_method – The imputation method to use in case of missing values. Defaults to LastValueImputation which replaces each missing value with the last value that was not missing.

predict_item(item: Dict[str, Any]) → gluonts.model.forecast.Forecast