I am currently working on a clustering project and would like to obtain statistics for the number of clusters to include. In SAS you get a pseudo-F statistic and a cubic clustering criterion. Has anyone developed a function to get these values? Sean