generate_day_lags

openstef_models.transforms.time_domain.lags_adder.generate_day_lags(min_horizon: timedelta, max_day_lags: int) → list[timedelta]

Generate day-based lag features for capturing daily and weekly patterns.

Creates daily lags from the minimum required days up to the maximum allowed, useful for capturing day-of-week and weekly seasonality.

Parameters:

• min_horizon (timedelta) – Minimum forecast horizon - only lags >= this will be included.

• max_day_lags (int) – Maximum number of days to look back (typically 14 for two weekly cycles).

• min_horizon

• max_day_lags

Returns:

List of timedeltas representing valid day-based lags, sorted descending. Empty list if minimum required days exceeds max_day_lags.

Return type:

list[timedelta]