captest.capdata.CapData.filter_custom

CapData.filter_custom(func, *args, custom_name=None, **kwargs)

Apply func to data_filtered as a row filter and record the step.

func is called as func(self.data_filtered, *args, **kwargs) and must return a DataFrame whose index is the rows to keep. Many pandas DataFrame methods qualify, e.g. pd.DataFrame.between_time or pd.DataFrame.dropna.

Parameters:

• func (callable) – Takes a DataFrame as the first argument and returns a DataFrame.

• *args – Forwarded to func.

• **kwargs – Forwarded to func.

• custom_name (str, default None) – Optional display label for the recorded filter step. Keyword-only so it cannot collide with positional args destined for func.

Notes

The class-based pipeline preserves the original column set: only the returned frame’s index is consumed. A function that drops or transforms columns will see its column changes discarded — pass column-transforming logic outside the filter pipeline.