Identifies the column by which the dataset is grouped for purposes of applying the transform.