Date:         Sun, 11 Jul 2004 23:58:05 -0400
Reply-To:     Thomas <tythong@YAHOO.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Thomas <tythong@YAHOO.COM>
Subject:      Filter Abnormal Values

Hi all,

I have a dataset as belows. May I know how to filter out the adnormal
values of X1 for each coname, which values for certain years are far below
50% (arbitrary??) or more from other years? So, for coname A, the values
of X1 for 1997 and 1996 are consistent, but the values of 1995 and 1994
are abnormal, so how to remove the obs of 1995 and 1994 from coname A.
Same for the other conames.

Thank you very much!
Thomas

*input;
coname    year     x1
A         1997     10300
A         1996     10000
A         1995     1000
A         1994     950
B         2002     203000
B         2001     200000
B         2000     189000
B         1999     5000
B         1998     3500

*expected output;
coname    year     x1
A         1997     10300
A         1996     10000
B         2002     203000
B         2001     200000
B         2000     189000