Re: [R] SVD on a matix

the variance is the eigen values of the correlation matrix of yoru matrix X.cor <- cor(X) X.e <- eigen(X.cor) X.e$values# Eigenvalues of cor(X) = variances you're asking about kayj wrote: > > Hi All, > > I performed an svd on a matrix X and saved the first three column of the > left singular

[R] SVD on a matix

Hi All, I performed an svd on a matrix X and saved the first three column of the left singular matrix U. ( I assume that they correspond to the projection of the matrix on the first three eigen vectors that corresponds to the first three largest eigenvalues). I would like to know how much varian