Date:         Tue, 19 Jun 2007 10:40:02 -0400
Reply-To:     Sigurd Hermansen <HERMANS1@WESTAT.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Sigurd Hermansen <HERMANS1@WESTAT.COM>
Subject:      Re: Cross-referencing to value of another case
Comments: To: Will <schimpanski@gmx.de>
In-Reply-To:  <5dq5stF35i8q6U1@mid.individual.net>
Content-Type: text/plain; charset="us-ascii"

Will:
proc sql;
   create table twins as
   select subject,twinpair,max(zygosity) as zygosity
   from dataset1
   group by twinpair
   order by subject
   ;
quit;

The query selects the maximum value of zygosity for a twin pair. SAS SQL
happens to put missing values at the bottom of the collating sequence
(convenient in this case, but dangerous in general). I am assuming that
subjects without twins appear only once in a table.
S

-----Original Message-----
From: owner-sas-l@listserv.uga.edu [mailto:owner-sas-l@listserv.uga.edu]
On Behalf Of Will
Sent: Tuesday, June 19, 2007 10:01 AM
To: sas-l@uga.edu
Subject: Re: Cross-referencing to value of another case


Arild.Skogmo@klp.no skrev:
 > what does your data look like?

Hallo,

to make my question more understandable:
How do I get from dataset1 to dataset2?

DATA  dataset1;
INPUT
subject  twinpair zygosity;
DATALINES;
0011     100      1
0012     100      .
0013     .        .        /* singleton, non-twin*/
0021     101      2
0022     101      .
RUN;

DATA  dataset2;
INPUT
subject  twinpair zygosity;
DATALINES;
0011     100      1
0012     100      1
0013     .        .        /* singleton, non-twin*/
0021     101      2
0022     101      2
RUN;

Thanks,

Will