Kai and Peter and all, This is the most efficient approach among the replies that I've seen to this request. Once you read moves.dat, you can run a simple CROSSTABS or TABLES command. However, there is 1 error to correct.

compute dest = stops(#i) .

should be

compute dest = stops(#i+1) .

which I'm sure is what Kai intended. I also wanted to point out that you can save directly to an SPSS system file in a loop and/or DO IF set by using the XSAVE command, as in

do if ~sysmis(dest) . XSAVE OUTFILE = 'moves.dat' / KEEP from dest . end if .

If you plan to use the new file in SPSS, anyway, this can avoid a lot of worries regarding missing values, rounding, spacing, etc. You might also want to consider a variable name other than "from", which is a keyword in some SPSS commands - maybe "depart". David Matheson SPSS Technical Support

-----Original Message----- From: Kai Borgolte [mailto:Kai.Borgolte@HOME.IVM.DE] Sent: Friday, February 26, 1999 10:40 AM To: SPSSX-L@UGA.CC.UGA.EDU Subject: Re: Help with syntax

"Peter Newcombe" <newc@psy.uq.edu.au> schrieb in <7b3bdn$uu0$1@bunyip.cc.uq.edu.au>:

>We >would like to gain data relevant to the number of times people travelled >from point A to Point B (regardless of whether it may have been from Stop 1 >to 2 or Stop 9 to 10) as well as times (ie frequency of trips) travelled >from Point B to Point A.

The following syntax creates a data file with one case for each move. Hope that helps.

vector stops = stop1 to stop10 . loop #i = 1 to 9 . compute from = stops(#i) . compute dest = stops(#i) . do if ~sysmis(dest) . write outfile = 'moves.dat' / from ' ' dest . end if . end loop . execute .

data list file = 'moves.dat' list / from dest . -- Kai Borgolte, Bonn