Clustering comparisons with similarity matrix

[Stafford Phillip-APS095 <phillip.stafford@motorola.com>]

Dear S-group;

I am trying to compare several commercial and shareware clustering programs 
(k-means, SOM, HMM) to the built-in S+ clustering algorithms using a standard 
normalized symmetrical similarity matrix that I created from a BLAST algorithm. 
 Question: has anyone written code to output a simple list of cluster ownership 
and row name?  All the S+ clustering algorithms give me a different type of 
output for group membership.  Also, has anyone out there written an HMM in S 
code?


Thanks for your help and advice;

Phillip Stafford
Motorola