Date:         Fri, 12 Jun 2009 10:58:01 -0700
Reply-To:     "Richard A. DeVenezia" <rdevenezia@WILDBLUE.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Richard A. DeVenezia" <rdevenezia@WILDBLUE.NET>
Organization: http://groups.google.com
Subject:      Re: Table Transform
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

On Jun 12, 1:06 pm, Emma G <emma.gottesman.h...@googlemail.com> wrote:
> I have a dataset with 2 fields, an ID and a part-number, both
> character variables. Here's a sample of made-up but representative
> data:
> ID     Part-Num
> 1256 789-9867
> 2345 657-8988
> 2345 879-7866
> 3444 666-1234
> 3444 666-1234
> 5678 789-9867
>
> I need to transform this to a table that has one row per id and then a
> variable containing all the part-numbers for that ID. It should look
> something like this:
> ID     Part-Num List
> 1256 789-9867
> 2345 657-8988, 879-786
> 3444 666-1234, 666-1234
> 5678 789-9867
>
> There is no maximum number of part-numbers for an ID, it's one or
> more.

If the number of parts per id is not excessive, the parts list can be
stored in a maximum length character string and the list can be
accumulated in a do over group loop.

data IdPartsList / view=IdPartsList;
  do until (last.id);
    set IdParts;
    by ID;
    length partslist $32767;
    partslist = catx(', ',partslist,part);
  end;
run;

--
Richard A. DeVenezia
http://www.devenezia.com