I want to generate some samples from a set of correlated exponentially distributed random variables. I have the correlation matrix between these random variables.
For multivariate normal distributions, one can easily use Cholesky/Eigenvalue decomposition to generate uncorrelated samples and then bring them in the correlated space.