Date:         Sat, 13 Jun 2009 02:18:25 -0700
Reply-To:     paulinem11 <pauline.macouillard@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         paulinem11 <pauline.macouillard@GMAIL.COM>
Organization: http://groups.google.com
Subject:      Re: Error message in proc catmod
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

On 12 juin, 20:46, rh...@UNMC.EDU (Robin R High) wrote:
> Are any of the variables "continuous" -- if so, they need to be specified
> on a DIRECT statement, i.e., as specified in the documentation:
>
> DIRECT < variables > ;
> The DIRECT statement lists numeric independent variables to be treated in
> a quantitative, rather than qualitative, way. etc..
> I only know enough about CATMOD to "rejoice greatly" that PROCs LOGISTIC,
> GENMOD, or GLIMMIX exist (unless you really need adjacent logits) --
> clogics or glogits might also be an option.
>
> Robin high
> UNMC
>
> paulinem11 <pauline.macouill...@GMAIL.COM>
> Sent by: "SAS(r) Discussion" <SA...@LISTSERV.UGA.EDU>
> 06/12/2009 09:47 AM
> Please respond to
> paulinem11 <pauline.macouill...@GMAIL.COM>
>
> To
> SA...@LISTSERV.UGA.EDU
> cc
>
> Subject
> Error message in proc catmod
>
> I use the proc catmod for an adjacent-category logistic model, and
> when my model contains a lot of variable there is this message in the
> journal
> "The response functions are linearly dependent since the number of
> functions per population, 2, is greater than or equal to the number of
> response levels, 1, in population 1"
>
> What does it mean?
>
> Thanks
> Pauline
>
> I give you my code SAS
>
> proc catmod data= table order=data;
> model gr_score= actphy0 age_cl rcsp_r gly_gr hta_gr niv_etud sexe
> sitfamille tabac0 tg_gr imc_c2 tail_gr
> gly_gr*niv_etud gly_gr*sexe hta_gr*actphy0 tg_gr*age_cl imc_c2*actphy0
> imc_c2*age_cl imc_c2*tabac0
> response alogit;
> format gr_score score. actphy0 actphy. age_cl age_cl. rcsp_r rcsp_r.
> gly_gr gly_gr. hta_gr hta_gr. sexe sexe. sitfamille sitfamille.
> tabac0 tabac. tg_gr bio_gr. imc_c2 imc_2c. tail_gr tail_gr.;
> where age0<64;
> run;


Thanks for your answer.
But there is no "continous" variable in my model.
When I use a model with less variables, the error message doesn't
appear.
So I don't understand, where the problem can be.

Thanks
Pauline