Search All of the Math Forum:

Views expressed in these public forums are not endorsed by NCTM or The Math Forum.

Notice: We are no longer accepting new posts, but the forums will continue to be readable.

Topic: Eigenvalues or Explained Variance of Rotated Principal Components
Replies: 1   Last Post: Feb 7, 2013 7:15 AM

 Search Thread: Advanced Search

 Greg Heath Posts: 6,387 Registered: 12/7/04
Re: Eigenvalues or Explained Variance of Rotated Principal Components
Posted: Feb 7, 2013 7:15 AM
 Plain Text Reply

"Mathew Guilfoyle" wrote in message <kethvb\$stp\$1@newscl01ah.mathworks.com>...
> Hi, looking for a way of determining the explained variance of principal components following rotation (i.e. after appying rotatefactors function).
>
> I have a dataset where the first 3 PCs explain >98% of the variance (approx 70, 20, and 8% per PC). To make them easier to interpret it is necessary to apply a varimax rotation (to first 3 PCs only). Clearly the total amount of explained variance across the 3 PCs will remain the same after rotation but it will be distributed very differently amongst them. In other programs (e.g. SPSS) the eigenvalues and explained variance of the rotated components is provided but in Matlab they are not.
>
> Is there an easy way of extracting these values? Can the rotation matrix output from rotatefactors somehow be applied to the vector of unrotated eigenvalues to get the rotated eigenvalues?
>
> Thanks for any help

Post to sci.stat.math

Greg

© The Math Forum at NCTM 1994-2018. All Rights Reserved.