Hi,I am puzzled with different results in R, compared to SPSS and
Senstools.Net. After running a PCA on the three software (for R, I use the PCA
function programmed on the FactoMineR package), I extracted the loadings. In
the three cases, the loadings are exactly identical.Then, I applied the varimax
rotation on these loadings, and unfortunately, the results differ. A quick look
at the resulting rotation matrix (rotmat) shows me, that the results are
"sorted" differently. Here are my results for the rotation matrix:On R: >
complet.varimax <- varimax(complet.loadings,normalize=T)$rotmat>
complet.varimax [,1] [,2] [,3] [,4] [,5] [,6] [,7]
[,8][1,] 0.529 -0.034 -0.411 -0.428 0.013 -0.297 -0.399 0.344[2,] -0.119
0.715 -0.194 -0.058 0.517 -0.219 -0.068 -0.337[3,] 0.081 0.087 0.722
-0.017 0.250 -0.472 0.092 0.413[4,] 0.110 0.039 -0.172 0.860 0.139
-0.013 -0.343 0.284[5,] -0.159 -0.353 -0.419 0.010 0.502 -0.114 0.578
0.275[6,!
] -0.122 -0.065 0.184 -0.240 0.502 0.670 -0.357 0.248[7,] 0.612
0.409 0.026 0.070 -0.061 0.420 0.497 0.159[8,] -0.522 0.427 -0.181
-0.107 -0.379 0.043 0.042 0.595On SPSS and Senstools.Net:Rotation dim 1
dim 2 dim 3 dim 4 dim 5 dim 6 dim 7 dim 8dim 1 -0,53
-0,41 -0,43 -0,40 0,03 -0,34 -0,30 -0,01dim 2
0,12 -0,19 -0,06 -0,07 -0,71 0,34 -0,22 -0,52dim 3
-0,08 0,72 -0,02 0,09 -0,09 -0,41 -0,47 -0,25dim 4
-0,11 -0,17 0,86 -0,34 -0,04 -0,28 -0,01 -0,14dim 5
0,16 -0,42 0,01 0,58 0,35 -0,27 -0,11 -0,50dim 6
0,12 0,18 -0,24 -0,36 0,07 -0,25 0,67 -0,50dim 7
0,61 -0,03 -0,07 -0,50 0,41 0,16 -0,42 -0,06dim 8
-0,52 0,18 0,11 -0,04 0,43 0,60 -0,04 -0,38 Here,
we can clearly see, that the results are!
similar, but sorted differently (in R, the second column correspond t
o the 5th column in SPSS or Senstools.Net).Concerning the loadings after
rotation, the results are different to (permutation are seen too).Please, could
someone explain me what's happening?Personally, I have the feeling that the
results differ...is it true? Which results should I trust?Should you need any
further information for answering me, do not hesitate to contact me ([EMAIL
PROTECTED]).I look forward to your response.Best regards, WORCH Thierry
_________________________________________________________________
[[replacing trailing spam]]
[[alternative HTML version deleted]]
______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
