|Date: ||Mon, 8 Oct 2001 14:25:28 +0200|
|Reply-To: ||HALIMI RaphaŽl <Raphael.HALIMI@GEP.FR>|
|Sender: ||"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>|
|From: ||HALIMI RaphaŽl <Raphael.HALIMI@GEP.FR>|
|Subject: ||Scoring a new datasets with proc MIXED or GLM|
|Content-Type: ||text/plain; charset="iso-8859-1"|
I fitted a linear model thanks to a MIXED procedure (or GLM because I'm
working with fixed effects). Now, I'm looking for an easy way to score a new
I looked to several directions, in vain :
1- in proc LOGISTIC, you can output the parameters with an OUTEST statement,
then run the LOGISTIC proc again with the INEST= and MAXITER=0 statements.
2- in proc REG, you can output the parameters then use the SCORE proc.
With the MIXED (or GLM) proc, I didn't find the right way to use my first
estimation onto a new datasets.
Did anyone meet the same problem ?