Dear Ales,
I have found a matlab routine which does latin hypercube sampling
with random distributions. It seem to be quite straightforward to
transform it to R. If you or anybody else is interested - and it
works - I am happy to send the R version to him. The method is from
Iman and Conover and allows inducing rank correlation among input
variables, like you can do with @Risk, e.g. quite easily.
Many thanks and best regards,
Stefan
I do not belive (although I am not an expert) that such a pacake
exist in R.
You might try sampling first one variable and then the other
conditional on
the previous one?
I hope it helps;
Ales Ziberna
----- Original Message -----
From: <stefan.albrecht@allianz.com>
To: <R-help@stat.math.ethz.ch>
Sent: Monday, November 15, 2004 4:49 PM
Subject: [R] Multivariate Sampling
Dear all,
I am looking for routines which allow multi-variate sampling from
non-normal distributions (loglogistic) given correlations among the
variables.
Unfortunately, I could not find a suitable package for R. Does
anybody know one?
Many thanks and best regards,
Stefan Albrecht
[[alternative HTML version deleted]]
______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html
[[alternative HTML version deleted]]