Yet another solution using 'do until' trick which i learned from SAS-L....

data practice_dates; input id date mmddyy8. test passed $; format date mmddyy10.; datalines; 1234 01021969 2345 Y 1234 01301969 3456 N 3157 02031969 2345 N 3157 02201969 2897 N 3157 04151969 2345 Y 1011 02051969 2345 N 1011 02211969 2345 N 1011 05201969 2897 N 2468 03211969 2234 Y 2468 07151969 2255 Y ; run ;

proc sort;by id ;

data final (where = ( new in (1,2) ) drop=yes no ) ; do until (last.id) ; set practice_dates ; by id ; select(passed); when('Y') yes=1; when('N') no=1; otherwise; end ; end ;

if yes and no then new=1 ; else if yes then new=2 ; else if no then new=3 ;

do until (last.id) ; set practice_dates ; by id ;

output ; end ; run ;

proc print data=final(drop= new ); run ;

On Fri, 16 May 2008 10:53:45 -0700, geraden72011 <geraden72011@GMAIL.COM> wrote:

>I have the following example i created: > >data practice_dates; >input id date mmddyy8. test passed $; >format date mmddyy10.; >datalines; >1234 01021969 2345 Y >1234 01301969 3456 N >3157 02031969 2345 N >3157 02201969 2897 N >3157 04151969 2345 Y >1011 02051969 2345 N >1011 02211969 2345 N >1011 05201969 2897 N >2468 03211969 2234 Y >2468 07151969 2255 Y >; >proc sort data=practice_dates; >by id date; >run; > >What I want to do is to create a new datafile that has all of the >records for the ID's that have a passed = 'Y' for any of the tests. > >so for the above example my new datafile would have all of the >following > >1234 01021969 2345 Y >1234 01301969 3456 N >3157 02031969 2345 N >3157 02201969 2897 N >3157 04151969 2345 Y >2468 03211969 2234 Y >2468 07151969 2255 Y > >any help is greatly appreciated.