Date:         Thu, 28 Feb 2008 15:49:33 -0600
Reply-To:     Mary <mlhoward@avalon.net>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Mary <mlhoward@AVALON.NET>
Subject:      Re: Comparison of String variables
Comments: To: Vijayan Sundaram <vijchid@GMAIL.COM>
Content-Type: text/plain; charset="iso-8859-1"

Your idea sounds like a good one.  Perhaps something like code like this (not tested):

data set1;
   informat combined $2.;
   set ...;
   if id1 < id2 then
     combined=trim(id1) || trim(id2);
   else 
      combined=trim(id2) || trim(id1);
run;

proc sql;
    select distinct combined
    from set1
   order by combined;
quit;
run;
  ----- Original Message ----- 
  From: Vijayan Sundaram 
  To: SAS-L@LISTSERV.UGA.EDU 
  Sent: Thursday, February 28, 2008 3:08 PM
  Subject: Comparison of String variables


  I have data set which looks like this.

  OBS_NUM  ID1 ID2
  -----------------
     1     A    B
     2     B    A
     3     A    C
     4     A    D
     5     D    A

  I want to define the pairs (1,2) and (4,5) as duplicates and end up with
  the following dataset
    ID1 ID2
  ------------
      A    B
      A    C
      A    D

  My intial idea was to create a new variable called ID which will have a
  value of AB for both observations 1 and 2 and then do a nodup on the new
  variable. To create the new variable, I need to compare the A against B
  and find out which is first in the sort order.

  Any ideas?

  Thanks for the help