Date:         Wed, 7 Feb 1996 10:35:01 -0500
Reply-To:     Christopher Zogby <chris@PHOR.COM>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Christopher Zogby <chris@PHOR.COM>
Organization: Pharmaceutical Outcomes Research, Inc.
Subject:      Re: Counting the # of variables in a SAS dataset
In-Reply-To:  <199602062049.PAA04466@ace.phor.com>

Sorry for the re-repost, but my last updated post could yield inaccurate
results if the last record of 'work.out' is not the greatest value of varnum.
A sort will fix that.
 
 
Corrected code below.
 
 
%macro numvars(dsn);
proc contents data=&dsn noprint out=out(keep=varnum);
   run;
 
proc sort;
by varnum;
 
 
data _null_;
   set out end=eof;
 
if eof then call symput('varnum',trim(varnum));
run;
 
%put &varnum=;
 
 
%mend numvars;
 
 
 
 
regards,
 
Chris
 
 
+*******************************************+
| Christopher Zogby                         |
| SAS Programmer/Analyst                    |
| Pharmaceutical Outcomes Research, Inc.    |
| 435 Lawrence Bell Drive                   |
| Williamsville, NY 14221                   |
|   Phone: (716) 633-3463                   |
|     Fax: (716) 633-7404                   |
|       e-mail: chris@phor.com              |
+*******************************************+