3D simulation of Gaussian.

Igor's built-in gauss() function can compute points in a 3D Gaussian, but lacks the correlation parameters. If you need the correlation parameters, you will need to compute it yourself with a user-defined function. The Wikipedia article about multivariate Gaussian is here: https://en.wikipedia.org/wiki/Multivariate_normal_distribution 

John's suggestion is well-taken, and proposes use of the standard transformation of easily-generated uncorrelated GRVs into correlated ones, based on the covariance matrix. However, another simple method using Cholesky decomposition has been given by Prof. Bechhoefer in a previous snippet:

https://www.wavemetrics.com/code-snippet/correlated-gaussian-random-var…

Thanks for the help.