Date:         Tue, 12 Jul 2005 16:20:51 -0400
Reply-To:     Richard Ristow <wrristow@mindspring.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Richard Ristow <wrristow@mindspring.com>
Subject:      Re: start with blank data file
Comments: To: Kim Jinnett <kjinnett@ibiweb.org>
In-Reply-To:  <MPECLJPOAEIJAOCEBHBJCEEGCBAA.kjinnett@ibiweb.org>
Content-Type: text/plain; charset="us-ascii"; format=flowed

At 03:16 PM 7/12/2005, Kim Jinnett wrote:

>If possible, I'd like to begin creating a data file from a blank data
>screen.  I have no data records at present and simply want to create a
>list of 500 random unique IDs to start.  Others have been so kind as
>to tell me how to create random unique IDs:
>
>compute rand=uniform(1).
>sort cases by rand.
>compute id=$casenum.

Glad I read this! I was about to send exactly this solution, to your
earlier posting.

>What is the file-related syntax preceding these statements so that I
>can invoke a blank data file and start from scratch.  Is that possible?

You do this sort of thing with an INPUT PROGRAM. Here's (untested)
code:

NEW FILE.
INPUT PROGRAM.
NUMERIC ID(N4).
LOOP #CASE = 1 TO 500.
.  COMPUTE ID = 0.
.  END CASE.
END LOOP.
END FILE.
END INPUT PROGRAM.
*  And, as before,  .
compute rand=uniform(1).
sort cases by rand.
compute id=$casenum.

HOWEVER: what does this buy you? That is, how does 500 empty cases in
'random' order, but numbered 1 to 500 in that order, differ from 500
empty cases numbered 1 to 500?.

Perhaps you want the number 1 to 500 in random order in in the data
editor, like this (still untested)?

NEW FILE.
INPUT PROGRAM.
NUMERIC ID(N4).
LOOP #CASE = 1 TO 500.
.  COMPUTE ID = #CASE.
.  END CASE.
END LOOP.
END FILE.
END INPUT PROGRAM.
*  And, as before,  .
compute rand=uniform(1).
sort cases by rand.

You'll have variable 'rand' in your file, that you'll probably want to
drop.