Message-ID: <20090417122315.rage2wjzlwkgoo84@www.ifremer.fr>
Date: 2009-04-17T02:23:15Z
From: Katell.Hamon at ifremer.fr
Subject: SAS-like cluster statistics for ward's minimum variance method
Hi all,
I am looking for an easy way to compute statistics out of my cluster
analysis using ward's method. Ideally it would be a table like that one
http://www.tau.ac.il/cc/pages/docs/sas8/stat/chap23/sect2.htm#clug2
with various validation statistics for the last few number of clusters.
The diagnostic I'm the more interested in is the R-squared which
represents the variance explained.
Is there a function doing this directly?
If not, Is there a way to convert the height into R^2 and an easy way
to extract the height for each number of cluster?
Thanks in advance,
Katell
--
Katell Hamon
Ph.D. Student
TAFI ? University of Tasmania / CSIRO / IFREMER Dpt of Marine Economics
Marine Research Laboratories, Nubeena Crescent, Taroona, TAS 7053
Email: Katell.Hamon at ifremer.fr Tel. (TAFI) (+61) 03 6227 7225
Tel. (CSIRO) (+61) 03 6232 5486
Mob.(+61) 04 3735 3067