Date:         Sun, 23 Nov 2003 22:32:57 -0500
Reply-To:     "Richard A. DeVenezia" <radevenz@IX.NETCOM.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Richard A. DeVenezia" <radevenz@IX.NETCOM.COM>
Subject:      Re: making an event history data

"praxis" <praxis6887@yahoo.com> wrote in message
news:bprjv7$4oi$1@geraldo.cc.utexas.edu...
> Hello,
>
> I want to build a event history type of data using this data set.
>
> id    sid    wave    dv    iv1
> 1      1    1          0      1
> 2      1    2          1      1
> 3      1    3          0      1
> 4      2    1          0      2
> 5      2    2          0      1
> 6      2    3          1      1
> 7      3    1          1      1
> 8      3    2          1      2
> 9      3    3          0      1
>
> Even though each of three individuals has data from 3 waves, I need to
take
> such cases as "id" 3, 8, and 9 out of this data set because these cases
were
> observed after the "event" was occurred (i.e., dv= 1). It would be nice if
> there is a way to control or use case number into the code, but I don't
> know. Please help.
>
> Thanks in advance.
> Jinseok
>


There are several techniques, some requiring retain, some not.
Presuming the observations of a sid group are always in adjacent rows.

Here is one requiring retain.

data filtered / view=filtered;
  set wave_events;
  by sid;
  retain event_occurred;
  drop event_occurred;
  if first.sid then event_occurred = 0;
  if not event_occurred; * subsetting if;
  if dv ^= 0 then event_occurred = 1;
  * implicit output;
run;

--
Richard A. DeVenezia
http://www.devenezia.com