incline.stl_decompose

incline.stl_decompose(df, value_column='value', period=None, seasonal_window_length=7, trend_window_length=None, robust=True)[source]

Seasonal-trend decomposition by LOESS.

An explicit STL request either returns an STL decomposition or raises. Automatic method selection may fall back to a moving average.

Parameters:

  • df (DataFrame) – Time series data.

  • value_column (str) – Column holding the values.

  • period (int | None) – Cycle length. Detected when None.

  • seasonal_window_length (int) – Length of the seasonal smoother; must be odd.

  • trend_window_length (int | None) – Length of the trend smoother; derived when None.

  • robust (bool) – Downweight outliers.

Returns:

The frame plus DECOMPOSITION_COLUMNS.

Raises:

ValueError – If STL cannot be fit with the requested configuration.

Return type:

DataFrame