I am using R mainly for multiple linear regression and principal
components analyses and I am quite happy with it. I think it is a
worderful work you have done.
But I am having a problem when using Varimax for rotating loadings
obtaines from a princomp (I use the cor = TRUE option): the variance
across the components in the variables does not mantain. Let me explain
it with an example in which the variances, which should be the same, are
listed in both the unrotated and rotated models:
Unrotated model:
PC1 PC2 variance
x1 0,627
-0,514
0,66 *
x2 -0,759
0,068
0,58
x3 0,730
-0,337
0,65
x4 -0,494
-0,798
0,88
x5 -0,425
-0,832
0,87
x6 -0,767
0,168
0,62
total variance
4,25
Model rotated with R's varimax:
PC1 PC2 variance
x1 0,767
-0,158
0,61 **
x2 -0,735
-0,298
0,63
x3 0,801
0,048
0,64
x4 -0,186
-0,937
0,91
x5 -0,110
-0,934
0,88
x6 -0,777
-0,214
0,65
total variance
4,33
* calculated as 0,627*0,627 + 0,514*0,514
**calculated as 0,735*0,735 + 0,298*0,298
Is there something special to take care about when performing Varimax?
Thanks in advance for your help,
Enrique Gonz?lez
egonzalez at ctv.es
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._