Date:         Fri, 7 Feb 1997 10:23:14 -0800
Reply-To:     gxx18300@ggr.co.uk, BWRogers@MSN.COM
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Bruce Rogers <gxx18300@GGR.CO.UK>
Organization: Medical IR, Glaxo Wellcome
Subject:      Re: adding records to a dataset
Comments: To: jupiterbowl <scot@wam.umd.edu>
Content-Type: text/plain; charset=us-ascii

jupiterbowl wrote:
>
 
> Is there a general strategy for adding records to reflect a missing value
> for SCORE in the missing months (Months 2,4-11 in the above case), besides
> editing the dataset the hard way?
>
 
Scot
 
There are several ways of doing this in a data step, one of which I've
included below (there may be better ways, but I don't have time to
examine all the possibilities right now) . I think it's pretty
self-explanatory, but if not email me and I'll fill in the details.
 
data new ;
set old;
by client ;
drop stormon storscor prevmon ;
 
   prevmon = lag(month) ;
   stormon = month ;
   storscor = score ;
 
   if first.client then prevmon = 0 ;
   do month = (prevmon+1) to (stormon-1) ;
      score = . ;
      output ;
   end;
 
   score = storscor ;
   month = stormon ;
   output ;
 
   if last.client & month ^= 12
      then do month = (stormon+1) to 12 ;
      score = . ;
      output ;
   end;
run;