A Note on a New Divisive Classificatory Program for Mixed Data