Message-ID: <A03919E6-7425-11D7-A090-000393DB5846@msu.edu>
Date: 2003-04-21T18:18:18Z
From: Hector L. Ayala-del-Rio
Subject: multiv library
Hi
I am not sure if this question if appropriate for this forum but here
it goes.
I was trying to do PCA analysis using the multiv library and I realized
that this particular library contains 8 methods of PCA which use
different correlation covariance structures and normalization of the
data. When I analyze one particular data set with the 8 different
methods I get different results in terms of the graphical display and
percentage of the variance explained. My question which method is more
appropriate under what circumstances. If somebody knows a reference
that talks about the 8 different methods mentioned in documentation of
this library please let me know. I am familiar with the differences
between a correlation and a covariance matrix, but not with the other
ones. If you think that I should post this message in a different
mailing list, please let me know.
Thanks and apologies if there is any information in the archives that I
did not find :-).
Hector
H?ctor L. Ayala-del-R?o, Ph.D.
Center for Microbial Ecology &
Center for Genomic and Evolutionary Studies
on Microbial Life at Low Temperatures
Michigan State University
545 Plant & Soil Sciences Building
East Lansing, MI 48824-1325
Phone: 517-353-9021
Fax: 517-353-2917