| Date: | Fri, 3 Sep 2004 18:53:55 GMT |
| Reply-To: | JP <crughy@BIGFOOT.COM> |
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
| From: | JP <crughy@BIGFOOT.COM> |
| Organization: | Pas très organisé |
| Subject: | CATMOD error message |
|---|
Hi all,
I am trying to model a polytomous logistic model using CATMOD.
I have 3 categories for the outcome variables (1, 2 and 3).
Those are ordonned values (Certain/Probable, possible, unrelated).
See for an example:
http://www.id.unizh.ch/software/unix/statmath/sas/sasdoc/stat/chap22/sect41
.htm
Here is a simplified model I tried:
PROC CATMOD DATA=fiche_event4;
RESPONSE 1 0.5 0;
DIRECT nbpres;
MODEL q2bis = nbpres / NOITER NOPROFILE CLPARM ;
WHERE qpat0;
RUN;
ERROR: The response functions are linearly dependent since the number of
functions per
population,1, is greater than or equal to the number of response
levels, 1, in
population2.
1611 QUIT;
It works when I do not specify the RESPONSE structure (so it create 2
fucntions). But I only want to have one.
It stops. I don't understand the message. I did a search in the archives.
It came up several times, without any answers.
Any idea or explanation?
Oups! I juste discovered that taking out nbpres (nb of prescriptions,
continuous) solve the pb. I am gonna work on that, but feel free to help
with any insight.
JP
|