Date:         Wed, 27 Sep 2006 15:05:24 -0400
Reply-To:     Mehdi Soleymani <mehdi_soleymani@SOFTHOME.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Mehdi Soleymani <mehdi_soleymani@SOFTHOME.NET>
Subject:      Re: Generating from Singular Normal
Comments: To: "David L. Cassell" <davidlcassell@MSN.COM>

David,
it is no difference in IML or Data Step, the problem is having a variance-
covariance matrix which is Singular, that is a k*k matrix with rank r < k,
How I can generate multivariate normal observation.
HOWEVER, could you, David, explain your approch more?
Thanks.