|
On Wed, 28 Jun 2006 11:16:21 -0400, Jim Groeneveld <jim2stat@YAHOO.CO.UK> wrote:
>Hi,
>
>See for a similar problem and a solution:
>http://listserv.uga.edu/cgi-bin/wa?A2=ind0606B&L=sas-l&P=R21295
>
>Regards - Jim.
>--
>Jim Groeneveld, Netherlands
>Statistician, SAS consultant
>home.hccnet.nl/jim.groeneveld
>
>
>On Wed, 28 Jun 2006 07:38:51 -0700, Apsteinberg <apsteinberg@HOTMAIL.COM>
>wrote:
>
>>Here is my issue...I have a table with 60K records. There are multiple
>>individual IDs for one site id. The data looks something like this:
>>Site_ID Individual_ID
>>1 1
>>1 2
>>1 55
>>3 8416
>>5 4568
>>9 46
>>13 88
>>13 959
These appear to be character variables, as in:
data have;
input Site_ID $ Individual_ID $;
cards;
1 1
1 2
1 55
3 8416
5 4568
9 46
13 88
13 959
;
>>
>>I need the data to be listed like this:
>>
>>Site_ID Individual_ID
>>1 1, 2, 55
>>3 8416
>>5 4568
>>9 46
>>13 88, 959
Try
data want1(keep = Site_ID Individual_IDs);
do until (last.site_id);
set have;
by site_id notsorted;
length Individual_IDs $ 40;
Individual_IDs = catx(', ',Individual_IDs,Individual_ID);
end;
length Individual_IDs $ 40;
run;
Then use PROC PRINT;
>>
>>Or like this:
>>Site_ID Individual_ID Individual_ID Individual_ID
>>1 1 2 55
>>3 8416
>>5 4568
>>9 46
>>13 88 959
Try
proc transpose data=have
out=want2(drop=_name_) prefix=Individual_ID;
by site_id notsorted;
var Individual_ID;
run;
and, again, PROC PRINT.
>>
>>I need to get it so that only one site id is listed per column
>>corresponding to every individual id for the site id. Right now I have
>>it so that if I have more than one individual id it is a new column.
>>Please help. Thanks
PROC REPORT might be a one-step possibility.
| 
