Does anyone know how to generate a latin hyper cube sampling of a multivariate distribution with mean, and covariance specified in python. It would be the equivalent of the X = lhsnorm (mu, sigma, n) in matlab. The answer provided here ( Is there a python equivalent to Matlab function lhsnorm?) does not address the question. It simply returns multivariate normal samples just like same as using . R = mvnrnd(mu,sigma,n) as seen here https://www.mathworks.com/help/stats/mvnrnd.html in matlab.
Asked
Active
Viewed 221 times
0
-
Perhaps https://pythonhosted.org/pyDOE/randomized.html ? – Severin Pappadeux Jul 25 '19 at 21:45
-
i have also looked at the pyDOE/randomized, it generates LHS samples but does not give the option of introducing correlated multivariate samples via the mean and covariance matrices. – george Jul 29 '19 at 16:29