qis.generate_rebalancing_indicators

qis.generate_rebalancing_indicators(df=None, index=None, freq='ME', include_start_date=False, include_end_date=False, num_warmup_periods=None, return_true_only=False)[source]

boolean rebalancing flags aligned to an observation index.

Builds a schedule at freq over the span of the index, then marks each observation date on or after a scheduled date. The result is the is_rebalancing input the backtester expects, so rebalancing happens on dates the data actually has rather than on calendar dates that may fall on a holiday.

Parameters:

  • df (DataFrame | Series) – frame or series whose index defines the observation dates; ignored if index is given

  • index (DatetimeIndex) – observation dates, as an alternative to df

  • freq (str) – rebalancing frequency, passed to generate_dates_schedule()

  • include_start_date (bool) – mark the first observation as a rebalancing date

  • include_end_date (bool) – mark the last observation as a rebalancing date

  • num_warmup_periods (int | None) – number of leading observations forced to False, so an estimator has data before the first rebalancing

  • return_true_only (bool) – return only the True entries rather than a flag per observation

Returns:

boolean series indexed as the observations, or only its True entries when return_true_only

Raises:
Return type:

Series