Date:         Mon, 5 Dec 2005 14:45:57 +0100
Reply-To:     a.smulders@beke.nl
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Antoon Smulders <a.smulders@beke.nl>
Subject:      Two aggregate questions
In-Reply-To:  <FABE7BF4C8A9B647A083DCEB99A792CF14A785@beke01.beke.local>
Content-Type: text/plain; charset="us-ascii"

Hello list members,

Often I have the following (related) problems:
I have a file with multiple records for each subject, for instance
(literally) criminal records, with fields (variables) like:
Id, date, place, crime.
Now often I want to aggregate this file in such a way that all VALUES of a
certain variable are represented as new VARIABLES in the aggregated file.
For instance:
AGGREGATE OUTFILE * /BREAK Id
        /murder = PIN (crime, 1,1)
        /robbery = PIN (crime, 2, 2)
        ... etc
My first question is if it is possible to automate this. (Maybe creating
automatic variable names like agg001 agg002 etc.)

Next I often want to have the frequencies of the values in the aggregated
file. In the above example: Id John commits variable murder 1 time, robbery
4 times etc.
What I do until now is add an aggregate variable:
 /number = N
and then calculate the original frequencies with the "pin variable".
I also considered aggregating the file in two steps:
Is there a more elegant way?

Thanks in advance.

Antoon Smulders