Date:         Fri, 12 Apr 1996 10:04:58 -0400
Reply-To:     Ronald Cody EdD <cody@UMDNJ.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Ronald Cody EdD <cody@UMDNJ.EDU>
Subject:      Re: HELP: "Merging" datasets
Comments: To: teiwa@CENTRUM.DK
In-Reply-To:  <no.id> from "Anders Brabaek" at Apr 12, 96 01:59:03 pm

Anders Brabaek writes:
>
> I have a small problem "merging" 2 datasets.
>
> One dataset, dataset1,consists of one observation and 2 variables: y1, y2.
> The other dataset, dataset2, consists of 7600 observations, each observation has 3 variables: x1, x2, x3.
> Now I want to put these two datasets together,
> so that I will get 7600 observations,.each consisting of 2+3=5 variables.
>
> Like this:
> Dataset1
> Obs     y1      y2
> 1       1       0
>
> Dataset2:
> Obs     x1      x2      x3
> 1       m       24      33
> 2       f       45      98
> etc
>
> Final
> Obs     x1      x2      x3      y1      y2
> 1       m       24      33      1       0
> 2       f       45      98      1       0
> ...
> 7600    m       18      70      1       0
>
>
> When I run the following code
>
> data Final;
>   merge dataset1 dataset2;
> run;
>
> I only get the observation from dataset1
> added to the first observation in dataset2,
> but I want to add it to all of the observations in dataset2.
>
> I suppose this should be a fairly easy operation,
> but being a novice with SAS I need help.
>
> Please, I need help.
> Thanks in advance.
> Anders Brabaek
>
Actually, it is not so easy or obvious.  You need to employ a "trick" if
you use a data step approach.  If you do a conditional SET statement on
the data set with one observation, the values (Y1 and Y2) will be
RETAINED (variables coming in with a SET are automatically retained) and
will be added to all obs in the other data set.  The code (not tested)
would look like this:
 
DATA FINAL;
   SET DATASET2;
   IF _N_ = 1 THEN SET DATASET1;
RUN;
 
 
Hope this helps.  You can see an example of this (and lots of other
examples) in SAS Programming by Example (Cody and Pass) available from
the SAS Institute.  Adding a single observation to each obs in another
data set is covered on page 190.  Ron
   --
 -----------------------------------------------------
|  Ronald Cody, Ed.D.                                 |
|  Robert Wood Johnson Medical School                 |
|  Department of Environmental and Community Medicine |
|  675 Hoes Lane                                      |
|  Piscataway, NJ 08854                               |
|  -------------------------------------------------- |
|  Voice Phone (908)235-4490                          |
|  Fax         (908)235-4569                          |
|  E-Mail      CODY@UMDNJ.EDU                         |
 -----------------------------------------------------