ForeTiS.preprocess.FeatureAdder

Module Contents

Functions

add_cal_features(df, columns_for_counter, ...)

Function adding all calendar-based features

add_statistical_features(seasonal_periods, ...)

Function adding all statistical features
ForeTiS.preprocess.FeatureAdder.add_cal_features(df, columns_for_counter, resample_weekly, event_lags, values_for_counter)

Function adding all calendar-based features

Parameters:

  • df (pandas.DataFrame) – dataset used for adding features

  • resample_weekly (bool) – whether to resample weekly or not

  • columns_for_counter (list) –

  • event_lags (list) –

  • values_for_counter (list) –

ForeTiS.preprocess.FeatureAdder.add_statistical_features(seasonal_periods, windowsize_current_statistics, columns_for_lags, columns_for_lags_rolling_mean, columns_for_rolling_mean, windowsize_lagged_statistics, seasonal_lags, df)

Function adding all statistical features

Parameters:

  • seasonal_periods (int) – seasonality used for seasonal-based features

  • windowsize_current_statistics (int) – size of window used for feature statistics

  • windowsize_lagged_statistics (int) – size of window used for sales statistics

  • seasonal_lags (int) – seasonal lags to add of the features specified

  • df (pandas.DataFrame) – dataset used for adding features

  • columns_for_lags (list) –

  • columns_for_lags_rolling_mean (list) –

  • columns_for_rolling_mean (list) –