Date:         Tue, 14 Dec 2004 05:00:42 -0800
Reply-To:     Dennis Diskin <diskin@SNET.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Dennis Diskin <diskin@SNET.NET>
Subject:      Re: how to retain last year's information if missing??
Comments: To: xzhou80@HOTMAIL.COM
Content-Type: text/plain; charset=us-ascii

xzhou,

I didn't read what you wanted carefully enough and I think I got it backwards. You can get the result you asked for from this code by first sorting on firm_id and DESCENDING year. Then run the datastep without the year in the BY statement.

Regards,
Dennis Diskin


Dennis Diskin <diskin@snet.net> wrote:xzhou,

Try this:

data new(drop=_year1 _year2);
set old (rename=year=_year2);
by firm_id year;
retain _year1;
if first.firm_id then _year1 = _year2 - 1;
do year = _year1 +1 to _year2;
  output;
end;
_year1 = _year2;
run;

HTH,
Dennis Diskin

xzhou80@HOTMAIL.COM wrote:
I was wondering whether you have some 'smart' codes doing the
following: :)



Basically, what I want is to eliminate missing observations for some
years of the firms because I know they exists, by coping the previous
year's info to this year's.



original data set:

firm_Id year x

1004 1990 13

1004 1991 12

1004 1992 9

1004 1993 9

1004 1994 10

1004 1995 10

1004 1996 10 **

1004 2000 9



1011 1990 7

1011 1991 7

1011 1992 8 **

1011 1994 7

1011 1995 7



The data set that I want:

1004 1990 13

1004 1991 12

1004 1992 9

1004 1993 9

1004 1994 10

1004 1995 10

1004 1996 10

1004 1997 10 **

1004 1998 10 **

1004 1999 10 **

1004 2000 9



1011 1990 7

1011 1991 7

1011 1992 8

1011 1993 8 **

1011 1994 7

1011 1995 7



Thanks very much.