preprocessing.exog_providers.EnergyCrisisWindowProvider

preprocessing.exog_providers.EnergyCrisisWindowProvider(
    data_home=None,
    csv_path=None,
    column=None,
    max_gap=0,
    max_tail_gap=0,
    provider_window=None,
)

German energy-saving regulatory window provider.

Produces a single float32 column energy_saving_window whose value is 1.0 during the two regulatory energy-saving periods in force during the 2022–2023 European energy crisis and 0.0 otherwise.

The two bundled windows are:

• EnSikuMaV (Kurzfristenergieversorgungssicherungsmaßnahmenverordnung): in force 2022-09-01 00:00 CEST through 2023-04-15 23:00 CEST, including the February 2023 extension.
• EU Council Regulation 2022/1854: mandatory −5 % peak-hour consumption reduction 2022-12-01 00:00 CET through 2023-03-31 23:00 CEST.

The intensity values are 1.0 for both rows; overlapping hours (December 2022 through March 2023) therefore also yield 1.0.

Parameters

Name	Type	Description	Default
data_home	DataHome	Unused (kept for a uniform provider signature).	None
csv_path	Optional[Union[str, Path]]	Optional explicit path to the energy-saving CSV, overriding the bundled energy_saving_windows_de.csv.	None
column	Optional[str]	Output column name. Defaults to "energy_saving_window".	None
max_gap	int	Accepted for provider-factory API uniformity; has no effect.	0
max_tail_gap	int	Accepted for provider-factory API uniformity; has no effect.	0
provider_window	Optional[pd.DatetimeIndex]	Accepted for provider-factory API uniformity; has no effect.	None

Notes

An open-ended Ukraine-invasion step dummy (2022-02-24 onward) was considered and rejected: the load effect is gradual and non-permanent, and in recent training windows the signal would be near-constant and uninformative to tree-based models. The include_entsoe_day_ahead_price flag already captures the energy-price channel that the invasion primarily affected.

Examples

import pandas as pd
from spotforecast2_safe.preprocessing.exog_providers import (
    EnergyCrisisWindowProvider,
)

# Window is active: 2022-10-01 is inside ENSIKUMAV_DE
idx_in = pd.date_range("2022-10-01", periods=24, freq="h", tz="UTC")
out_in = EnergyCrisisWindowProvider().build(idx_in)
print(out_in.columns.tolist(), out_in.shape)
assert (out_in["energy_saving_window"] == 1.0).all()

# Window is inactive: 2022-07-01 predates both regulatory periods
idx_out = pd.date_range("2022-07-01", periods=24, freq="h", tz="UTC")
out_out = EnergyCrisisWindowProvider().build(idx_out)
assert (out_out["energy_saving_window"] == 0.0).all()

['energy_saving_window'] (24, 1)