Date:         Tue, 7 Dec 2004 20:13:41 -0600
Reply-To:     "Dunn, Toby" <Toby.Dunn@TEA.STATE.TX.US>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Dunn, Toby" <Toby.Dunn@TEA.STATE.TX.US>
Subject:      Re: HELP:: by group  processing?
Comments: To: Steve Jones <st_jones77@YAHOO.COM>
Content-Type: text/plain; charset="iso-8859-1"

Steve,
 
There are so many questions i want to ask, chiefly why? But that wont solve your problem.  As you don't say and I am not a mind reader I will provide the frame work for you solve your problem and let you fill in the business logic for all your different situation.
 
 
data one (drop = __:);;
 set temp;
 by id;
 
retain  __w . __x . __yesno '    '  __lowhigh '    ';
 
if first.id then cnt = 0;
 
cnt + 1;
 
if cnt = 2 then do;
   __w = w;
  __x = x;
end;
 
if yesno   = 'yes' then   __yesno   = yesno ;
if highlow = 'mid' then   __lowhigh = lowhigh ;
 
if last.id then do:
 id = __id;
 w = __w;
 x = __x;
 yesno   = __yesno;
 highlow = __highlow;
 output;
end;
run;
 
You will need to determine what you will do in the cases like:
there is no value of 'mid' or 'yes', what to do when you have more than one of these values, and what observations you want w and x for.
 
HTH
Toby Dunn

________________________________

From: SAS(r) Discussion on behalf of Steve Jones
Sent: Tue 12/7/2004 5:48 PM
To: SAS-L@LISTSERV.UGA.EDU
Subject: HELP:: by group processing?



Hello all,

I have a dataset (an example below) that has multiple observations
per person/ID.

data temp;
  input id w x yesno $ lowhigh $;
  datalines;
  1 1 2 yes low
  1 2 3 yes mid
  1 2 4 no  high
  2 6 7 no  mid
  2 8 7 yes low
  2 4 8 no   .
  3 3 7 yes  high
  3 9 7 yes  mid
  3 8 8  .   mid
  ;
run;

My goal is to summarize the information by reducing
the data to one observation per ID.
To accomplish this, I want to keep specific values of some variables
in the dataset. For example, for the variable yesno, I want to keep
'yes' even though 'no' appeared in some instances. Similarly, for
variable lowhigh, I want to keep 'mid'.

In essence I want an output dataset that looks like:

  id w x  yesno lowhigh
  1  2 3  yes    mid
  2  8 7  yes    mid
  3  9 7  yes    mid


I appreciate any pointers.

Thanks,

SJ.




--
Posted via Mailgate.ORG Server - http://www.Mailgate.ORG