Date:         Sun, 11 Jan 1998 06:24:06 GMT
Reply-To:     Ashok Natarajan <anataraj@UCI.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Ashok Natarajan <anataraj@UCI.EDU>
Organization: University of California, Irvine
Subject:      Re: Help with Macro

Thanks to Anthony Ayiomamitis and Roland Rashleigh-Berry, I think I am
making progress in understanding how macros work.  However, I confused
myself while writing my previous message, and am still having
problems.
 
While I had indicated that the regressor variables were x1-x50, they
really are a little more descriptive, and have variable labels.  The
macro was supposed to retain the variable name and label value, which
is why I had an array defined for the 50 variables.
 
It is relatively easy to rename the variables and use Anthony's
solution, and this in fact works, but I can't differentiate between x1
"a super important variable" and x50 "a minor unimportant variable".
So let me define the problem again, hopefully a little more clearly
this time.
 
data a ;
        infile 'a.dat' ;
        input y abc1 def2 ghi3 ........ xyz50 ;
        label abc1 = "super important"
                 .......
                      xyz50 = "minor and unimportant" ;
        run ;
 
%macro myreg(j) ;
        proc reg data=a ;
                model y= x{&j} ;
                run ;
%mend ;
 
data _NULL_ ;
        set a ;
        array x{50} abc1 def2 ...... xyz50 ;
        do j = 1 to 50 ;
        %myreg(j)
        end ;
 
Will this work?  I am going to try it out.
 
Ashok