| Date: | Tue, 6 May 2008 22:03:50 -0400 |
|---|
| Reply-To: | Cordelia Flynn <prime431@YAHOO.COM> |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
|---|
| From: | Cordelia Flynn <prime431@YAHOO.COM> |
|---|
| Subject: | Re: Simulating variance components + Cholesky |
|---|
"Pollack, Simcha" <SPollack@Winthrop.org> wrote:
I would prefer replying to the whole list. Can you show me how?
>>> To post a reply to the list you should joint the list.
>>> thanks - CF
The answer to your question, based on discussions with Phil Gibbs of
SAS, is that the Cholesky decomposition allows one to generate
multivariate normal simulated data with arbitrary intercorrelations,
means and sd.
You can use Call Vnormal to do it without having to program the IML.
Simcha
| 
