featuretools.make_temporal_cutoffs#

featuretools.make_temporal_cutoffs(instance_ids, cutoffs, window_size=None, num_windows=None, start=None)[source]#

Makes a set of equally spaced cutoff times prior to a set of input cutoffs and instance ids.

If window_size and num_windows are provided, then num_windows of size window_size will be created prior to each cutoff time

If window_size and a start list is provided, then a variable number of windows will be created prior to each cutoff time, with the corresponding start time as the first cutoff.

If num_windows and a start list is provided, then num_windows of variable size will be created prior to each cutoff time, with the corresponding start time as the first cutoff

Parameters:

  • instance_ids (list, np.ndarray, or pd.Series) – list of instance ids. This function will make a new datetime series of multiple cutoff times for each value in this array.

  • cutoffs (list, np.ndarray, or pd.Series) – list of datetime objects associated with each instance id. Each one of these will be the last time in the new datetime series for each instance id

  • window_size (pd.Timedelta, optional) – amount of time between each datetime in each new cutoff series

  • num_windows (int, optional) – number of windows in each new cutoff series

  • start (list, optional) – list of start times for each instance id