[R] strange loadings matrix after varimax rotation: PCA with prcomp in R

David Villegas Ríos chirleu at gmail.com
Mon Oct 20 15:36:33 CEST 2014


Hello.
I ran a PCA analysis on a dataset with 5 variables and retained two
components. I rotated them and now I want to predict the scores in a new
data set for which the original variables are available.

I normally use the predict.prcomp() function to predict using a prcomp
object. For example..

PC1=predict(data.pca, new.data)[,1]

...will predict the first component in the new dataset.

But how to do it with the rotated components/loadings, since after
varimax() we don't have a prcomp object anymore?

Thanks in advance!

David

	[[alternative HTML version deleted]]



More information about the R-help mailing list