Date:         Tue, 26 May 1998 21:49:35 -0400
Reply-To:     "Luke T. O'Brien" <Luke_T@WEBTV.NET>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         "Luke T. O'Brien" <Luke_T@WEBTV.NET>
Organization: WebTV Subscriber
Subject:      Re: Data cleaning problem
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII

This is just a follow up suggestion and it might be not be necessary for
your situation, but in trying to clean up data I find it useful to
output the bad observations to a separate data set instead of just
deleting. This way you can check the bad data for what went wrong and
posibly correct it. Using this approach would look something like this:
 
Proc sort data_d ;
  By Rln Visit ;
   Run ;
 
Data Bad
         Good ;
Set data_d :
by Rln ;
retain date ;
If visit = 1 then date=month ;
If visit > 1 and  date > month then output Bad ;
else output Good;
Run ;
 
   One potential problem with this approach is that it assumes all the
entries for the first visit are correct. Also if second visits were not
supposed to occur in the same month as the first visit, that condition
would have to be added.
 
                                    Is mise le meas.