Date:         Tue, 26 Jun 2007 12:40:51 -0400
Reply-To:     Crescent Martin <crescent_b_martin@FANNIEMAE.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Crescent Martin <crescent_b_martin@FANNIEMAE.COM>
Subject:      parameter truncated in ODS parameterestimates output (proc genmod)

I'm running a Proc Genmod and writing the parameter estimates to a dataset
using ODS.  In this output dataset, the variable Parameter appears to have
maximum length of 20, so it cuts off the variables that I put into my model
statement.

My model statement contains 3 crossed variables:
statusLB*randomlag*season
(statusLB and randomlag and class variables and are listed in the class
statement; season is continuous)

But in the parameter file, it's shown as:
season*status*random

Is there any way to keep Proc Genmod from truncating my variable names? I'd
like to tie the parameter estimates back to the original dataset and so
would like the variable names to be unchanged. I tried using Proc Template
to format it, (based on this post: http://www.listserv.uga.edu/cgi-bin/wa?
A2=ind0303A&L=sas-l&P=R4638)  but it only added extra spaces to the end.

 ods path(prepend)
  WORK.TEMPLATE(UPDATE) ;
 run;

 proc template;
  edit Stat.Genmod.ParameterEstimates;
    define Parameter;
       format $50.;
    end;
  end;
 run;

And of course, the easy answer is to use shorter variable names, but now
I'm curious if there's a way to fix it, not just work around it.