Date:         Fri, 27 Jul 2007 02:54:27 -0400
Reply-To:     Gerhard Hellriegel <gerhard.hellriegel@T-ONLINE.DE>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Gerhard Hellriegel <gerhard.hellriegel@T-ONLINE.DE>
Subject:      Re: A DATA MANUPULATION QUESTION

if you want it in groups, you could do the following:

1. sort the dataset BY A
2. initialize the NEWB variable at every FIRST.A (use BY in the DATA -
step)
like:
  ....
  by a;
  if first.a then newb=" ";
  ...
3. cumulate the values (like in the other examples)
4. test, if the end-of-group is detected. If yes, put it out.
like
  ...
  if last.a then output;

only to give you a explanation of the logic. All other things around
(retain newb, concatenation, ...) stay like before.

Gerhard



On Thu, 26 Jul 2007 14:23:13 -0400, John Edward
<entry_mid_level_sas@YAHOO.COM> wrote:

>Mr. Tobby,
>
>It worked great But I just provided a part of the problem it was my
>mistake. Please look at the dataset and problem more closely:
>
>I have the following dataset with variabel A and B:
>
>A              B
>---           -------
>
>0001        Choka Moka
>0002        Cola Coca
>0002        Monro Choca
>0002        Hola Boka Loka
>0003        Poka
>
>
>I will need the following dataset:
>A          B
>----    --------------------------------------
>0001    Choka Moka
>0002    Cola Coca, Monro Choca, Hola Boka Loka
>0003    Poka
>
>
>your previous code retaining value of A  to the following obsevation B
>and puting  "," even in the first observation. I know it is my mistake by
>not giving you more detail problem.
>
>
>Thank you Mr. Toby