Date:         Mon, 31 Mar 2003 13:01:42 -0500
Reply-To:     diskin.dennis@KENDLE.COM
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         diskin.dennis@KENDLE.COM
Subject:      Re: Counting occurances of data
Comments: To: Brett Plummer <plummer@HSMGROUP.COM>
Content-type: text/plain; charset=us-ascii

Brett,

The simple way is to sort the dataset on VarA and then whatever other
variables the dataset is currently ordered on:

proc sort data=in;
by VarA otherIDs;
run;

data in;
set in;
by VarA;
if first.VarA then n_VarA = 0;
n_VarA+1;
run;

then possibly put back in the original order:
Proc sort data=in;
by otherIDs n_VarA;
run;

Other approaches are possible using an array to keep track of the counts
for each distinct value of VarA, but I doubt they would be practical since
you would need an array of at least Number of records divided by 4.

HTH,
Dennis Diskin





From:  Brett Plummer <plummer@HSMGROUP.COM>@LISTSERV.UGA.EDU> on 03/31/2003
       12:32 PM

Please respond to Brett Plummer <plummer@HSMGROUP.COM>

Sent by:    "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>


To:    SAS-L@LISTSERV.UGA.EDU
cc:

Subject:    Counting occurances of data


I have a large dataset where a specific instance of Variable A occurs up to
4 times.  The dataset is sorted by date.  I want to assign a new variable
that describes whether a specific instance of Variable A is the first,
second, third, or fourth time is has occured in the data.

Example of what I have:
RECORD VarA
1 1
2 2
3 3
4 1
5 1
6 2
7 2
8 3  etc.

Example of what I want to create:
RECORD VarA NewVar
1 1 1
2 2 1
3 3 1
4 1 2 (in record 4, VarA=1 for the second time)
5 1 3 (in record 5, VarA=1 for the third time)
6 2 2 (in record 6, VarA=2 for the second time)
7 2 3 (in record 7, VarA=2 for the third time)
8 3 2 etc.

Thank you,
Brett Plummer