Date:         Wed, 17 Oct 2007 16:45:06 -0500
Reply-To:     Mary <mlhoward@avalon.net>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Mary <mlhoward@AVALON.NET>
Subject:      Re: Use all numeric variables as predictors in proc reg
Comments: To: ginanicolosi@HOTMAIL.COM
Content-Type: text/plain; charset="iso-8859-1"

Hi, Gina, 

You could try something like this (not exact):

-Mary

proc sql;

describe table dictionary.columns;

quit;

run;

proc sql;

create table num_vars as

select name, informat

from dictionary.columns

where memname='RESULTS5' and

libname='WORK';

quit;

{then get a random number of observations from 

the data set num_vars; let's say we put it in

a data set called num_vars_subset }

{Then put the names in num_vars_subset into

a macro variable }

proc sql;

select name

into :random_vars

from num_vars_subset;

quit;

Then use your macro variable to run against the model:

proc reg;

model result= &random_vars;

run;



  ----- Original Message ----- 
  From: Gina Nicolosi 
  To: SAS-L@LISTSERV.UGA.EDU 
  Sent: Tuesday, October 16, 2007 3:44 PM
  Subject: Use all numeric variables as predictors in proc reg


  I am writing a program which randomly selects predictor variables to
  be included in a regression.  Since the predictors during any one
  iteration can change, I can't write them out in the MODEL statement.
  Therefore, I was wondering if there was any way to specify that all
  numeric variables in the data set act as regressors? Kind of like
  putting _ALL_ or _NUMERIC_ after the equals sign when running PROC REG
  (but this approach doesn't work).  Any guidance would be greatly
  appreciated.