Date:         Mon, 28 Jul 2008 16:50:35 -0400
Reply-To:     "Howard Schreier <hs AT dc-sug DOT org>"
              <schreier.junk.mail@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Howard Schreier <hs AT dc-sug DOT org>"
              <schreier.junk.mail@GMAIL.COM>
Subject:      Re: How to put variable name ?

On Mon, 28 Jul 2008 16:36:00 -0400, Fehd, Ronald J. (CDC/CCHIS/NCPHI)
<rjf2@CDC.GOV> wrote:

>> From: Adriano Rodrigues
>> Checking some data with:
>> data errors;
>> set all;
>> array vars(*) p17a p17b    p17c  p17d  p17e  p17f  p17g  p17h
>>  p17i  p17j
>> p17k  p17l  p17m    p17n  p17o  p17p  p17q  p17r  p17s  p17t
>> p17u  p17v
>> p17x  p17w  p17y  p17z;
>>
>> do i=1 to dim(vars);
>> if vars(i) notin (1,2) then put 'quest=' quest 'error='vars(i)  ;
>> end ;
>> run;
>
>> I appreciate better way for this case to check errors without
>> array too :-)
>
>This solution is non-trivial
>as it requires building formats for each variable you want to check.

In general, yes. But here it seems the same validation rule (=1 or =2)
pertains to all of the questions.

The need for loops and arrays is a consequence of the wide data structure.
If each response were a separate observation, the array would not be needed.

>
>INVALID: A Data Review Macro
>
>http://www.nesug.info/proceedings/nesug01/at/at1008.pdf
>
>http://www.lexjansen.com/pharmasug/2004/datamanagement/dm06.pdf
>
>Ron Fehd     he-wishes-he-were-retired-from-data-review
>          or the macro maven  CDC Atlanta GA USA RJF2 at cdc dot gov