Francisco Javier Santos Alamillos
2010-Jul-20 18:33 UTC
[R] How can I get the principal components after a varimax rotation using Varimax function?
Dear R users, I have some question about Varimax rotation of the loadings obtained from a PCA. Imagine X is a field where the rows are the observations and the columns are the variables. I obtain the loadings (L), the principal components (PC) and the percentage of the variance for each mode (PVar) from PCA. I would like to do a varimax rotation in order to find an easiest physical interpretation of the results. A way to do this task is to use the varimax function. I apply this function to the loadings as: L_varimax = L %*% varimax(L)$rotmat Now I would like to calculate the rotated principal components. It is possible to calculate it as pcs_varimax = scale(X) %*% L_varimax? If I do it, the standard desviation of the rotated principal components are not equal to 1. How can I solve it? How can I get the percentage of the variance for each rotated principal component? Thanks in advance [[alternative HTML version deleted]]