Date: Fri, 30 Sep 2005 14:30:17 -0400
Reply-To: Sigurd Hermansen <HERMANS1@WESTAT.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Sigurd Hermansen <HERMANS1@WESTAT.COM>
Subject: complete separation in logistic regression
Content-Type: text/plain; charset="us-ascii"
Anyone have any experience with logistic regression under conditions of
complete separation. Heinze and Schemper have publications and SAS Macro
I'd have a special interest in hearing from anyone who has used the
algorithm and would know something about its performance
characteristics. Would also like to hear whether SAS statistical PROC's
handle complete or almost complete separation, or if someone has adapted
NLMIXED or other procedures to the problem.