Message-ID: <200505231801.j4NI1i7S005677@magnolia.isciences.com>
Date: 2005-05-23T18:01:38Z
From: Thomas M. Parris
Subject: Can't reproduce clusplot princomp results.
Dear R folk:
Perhaps I'm just dense today, but I am having trouble reproducing the
principal components plotted and summarized by clusplot. Here is a brief
example using the pluton dataset. clusplot reports that the first two
principal components explain 99.7% of the variability. But this is not what
princomp is reporting. I would greatly appreciate any advice.
With best regards,
-- Tom
> R.version
_
platform i386-pc-mingw32
arch i386
os mingw32
system i386, mingw32
status
major 2
minor 0.1
year 2004
month 11
day 15
language R
> require("cluster")
[1] TRUE
> pluton.agnes <- agnes(pluton)
> clusters <- cutree(as.hclust(pluton.agnes), h=4.00)
> clusplot(pluton, clusters, lines=0)
> pca <- princomp(pluton, cor=TRUE)
> loadings(pca)
Loadings:
Comp.1 Comp.2 Comp.3 Comp.4
Pu238 0.521 0.348 0.714 0.313
Pu239 -0.540 0.837
Pu240 0.418 -0.835 0.353
Pu241 0.512 0.418 -0.698 0.277
Comp.1 Comp.2 Comp.3 Comp.4
SS loadings 1.00 1.00 1.00 1.00
Proportion Var 0.25 0.25 0.25 0.25
Cumulative Var 0.25 0.50 0.75 1.00