[R] centroid representation and MANOVA

[julien colomb]

hi all.

I have a matrix of data with 5 different groups and 20 individual 
response per group, and about 12 variables collected for each. I want to 
represent the result in a 2D plot. PCA is not so good because the 
difference between the groups is not obvious. I have seen, in a recent 
paper, people doing a MANOVA and representing it in a centroid plot 
(they used Matlab to do it).
I would like to do the same with R, but don't know what to do after the 
MANOVA... and nothing can be found in the Archives.

Any suggestions?

-- 

Dr. Julien Colomb



FB Biologie, Chemie, Pharmazie
Institut für Biologie
Neurobiologie -
Königin-Luise-Str. 28/30
14195 Berlin
Germany


http://www.neurobiologie.fu-berlin.de
Tel: 0049.30.83851860
Fax: 0049.30.83855455

skype: thrawny1
FedEX number:3000-6496-5