[S] Clustering: KMEANS question/concept

Falk Huettmann (k9wk@unb.ca)
Thu, 28 May 1998 16:22:41 +0000


I am trying to do clustering, KMEANS, with SPLUS, but after 2
hours reading the documentation and helpfiles I still cannot figure
it out (I just learn clustering ...). I checked-out SPSS and SYSSTAT
and it took me 10 minutes to get the concept and apply it to my data.
However, how could I do the same thing with my data now in SPLUS?

raw data: 18 columns (14 cols are continous data, 4 cols are
factorial), with 1034 cases/rows. Apparently, I do not understand how
to create the similarity matrix, centres and how to bring it into
KMEANS. When I use HCLUST, DIASY or KMEANS I get an error message
about object size etc.
Any help would be appreciated.



Falk Huettmann
University of New Brunswick (UNB)
ACWERN / Fac.Forestry
PO Box 44555
Tel. (+1) (506) 452 6033
Fax (+1) (506) 453 3538
Email k9wk@unb.ca
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news