In SPSS you can save a variable with the k-means cluster membership and then use this in the Scattergram to change the color/shape (can't remember which) of the points in your scagttergraph. I'm sure other packages do this too. "ben" <[EMAIL PROTECTED]> wrote in message news:ec03db5b.0211070901.493ffaac@;posting.google.com... > Hi everyone, > I need help. > I am looking for a statistical program that knows how to do k-means > cluster analysis and to show the different clusters on a scatter plot > graph. The clusters can be differentiated by their color or by a > circle around the center point of each one, or even just by a > cluster-number on the center point of each one. > I’ve been trying to do this with STATISTICA but without success, > I get the scatter plot graph with the observations clustered and > numbered but this gives me a big cloud of numbers and it covers the > observation points themselves. > I am in serious need of help with this. > Either doing it in STATISTICA or with some other program. > > Thanks > Ben
. . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
