Date:         Tue, 8 Feb 2000 14:03:59 -0600
Reply-To:     "Muzny, Robert V." <MUZNRV@INTEGRIS-HEALTH.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Muzny, Robert V." <MUZNRV@INTEGRIS-HEALTH.COM>
Subject:      Re: Record Selection
Comments: To: "anirdesh@HOTMAIL.COM" <anirdesh@HOTMAIL.COM>
Content-Type: text/plain; charset="iso-8859-1"

How about this (untested)

data a;
        retain grd1;
        set yourdata;
        by name;
        if first.name then grd1=grd;
        if last.name then
                if grd1=grd then output;
run;

Data set a should contain only records where the first grade in the sequence
is equal to the last grade in the sequence.

RVM

original text:
Date:    Tue, 8 Feb 2000 08:35:18 PST
From:    George Smith <anirdesh@HOTMAIL.COM>
Subject: Record Selection

I HAVE RECORDS AS FOLLOWS  :

NAME EDTE GRD

ABC    890123 04
ABC    900413 05
ABC    970624 06
ABC    991213 04
XYZ    910123  07
XYZ    940413  08
XYZ    980624  08
XYZ    991101  11
OPQ   870130  06
OPQ   980130  06
OPQ   000123  06

I WANT TO FIND OUT HOW MANY PERSONS ARE THERE IN A DATABASE, WHO HAVE SAME
GRADE
IN THE BEGINING AND AT THE LAST.
(  IN THIS CASE PERSON ABC AND OPQ HAVE SAME GRADE; 04 AND 06 RESPECTIVELY,
E-RESPECTIVE OF EDATE)