Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: How can I get a rotated version of principal components in eigenvalue normalisation?


From   "Stas Kolenikov" <skolenik@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: How can I get a rotated version of principal components in eigenvalue normalisation?
Date   Fri, 22 Aug 2008 09:40:38 -0500

A brute force solution: go to return values (type -ereturn list-),
identify what the vector of eigenvalues is (it's -e(Ev)- in Stata 10),
and then manually scale the scores:

pca <your variables here>
local k = <number of scores you want to use>
predict score1 - score`k'
mat lambda = e(Ev)
forvalues i=1/`k' {
   replace score`i' = score`i'*sqrt(lambda[1,`i'])
}

I've no idea how that will go with rotation. I personally never had
much trust in rotations, anyway...

On Thu, Aug 21, 2008 at 3:32 PM, Thomas Lux <tholux2000@yahoo.de> wrote:
> Stata 9 displays principal components in unit normalisation, but one can diplay the principal components in eigenvalue normalisation by typing:  estat loadings, cnorm(eigen)
> But how can I get a rotated version of these principal components in eigenvalue normalisation?
>

-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index