I have a sample of associated variables xi=(xi1,xi2,xi3,...). I want to generate random vectors from the distribution implied by this sample. The xijs are correlated and not necessarily normal.
I've looked up some information on both Cholesky and Spectral Decomposition but as far as I can tell these methods only apply to normal distribution. How would i generate a vector using correlations and bootstrapped distributions from the sample?
I just realised I posted in the Pre-U section. I will repost in the appropriate forum. Can a moderator delete this post.