Date:         Thu, 12 Jul 2007 05:53:45 -0700
Reply-To:     Ucal Gyam <ucalgyam@yahoo.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Ucal Gyam <ucalgyam@yahoo.com>
Subject:      Re: Combining Two Datafiles
Comments: To: Richard Ristow <wrristow@mindspring.com>
Content-Type: text/plain; charset=ascii

Thanks for your response Richard .
You are right , if records match (on those three variables) in both files, I  want to keep both (all) records in my output file, not merge them into one.

Which part of syntax wont give me desired results if there both files have the same variables?




----- Original Message ----
From: Richard Ristow <wrristow@mindspring.com>
To: SPSSX-L@LISTSERV.UGA.EDU
Sent: Wednesday, July 11, 2007 11:55:02 PM
Subject: Re: Combining Two Datafiles


At 01:59 PM 7/11/2007, Ucal Gyam wrote:

>I have 2 datafiles FileA and FileB with same variables, I want to
>Create a third datafile FileC  containing only records from both FileA
>and FileB Where the values of the variables x1 , x2 and  x3  (ie same
>matching records on x1,x2 and x3) in both datafiles are equal .

If I read you correctly: if records match (on those three variables) in
both files, you want to keep both records in your output file, not
merge them into one. (The latter won't give the desired results anyway,
if there both files have the same variables.)

The following is not tested. It assumes that FileA, FileB, FileC are
valid file references, perhaps file handles. It uses datasets (SPSS 14
and beyond). If you have an earlier SPSS, you can make srtFileB a
scratch .SAV file, changing the dataset-handling logic to file-handling
logic.

This logic leaves variables FrmFileA, FrmFileB, InFileA and InFileB in
the output file. I think you should keep the first two of these. If you
don't want the others, add
    /DROP InFileA InFileB
to your SAVE statement.

GET FILE=FileB.
SORT CASES BY x1 x2 x3.
DATASET NAME    srtFileB.

GET FILE=FileA.
SORT CASES BY x1 x2 x3.

ADD FILES
    /FILE=*        /IN=FrmFileA
    /FILE=srtFileB /IN=FrmFileB
    /BY x1 x2 x3.

AGGREGATE OUTFILE=* MODE=ADDVARIABLES
    /BREAK = x1 x2 x3
    /InFileA 'x1-x2-x3 key found in FileA' = MAX(FrmFileA)
    /InFileB 'x1-x2-x3 key found in FileB' = MAX(FrmFileB).

SELECT IF   FrmFileA EQ 1
         AND FrmFileB EQ 1.

SAVE OUTFILE=FileC.



____________________________________________________________________________________
Pinpoint customers who are looking for what you sell.
http://searchmarketing.yahoo.com/