Date:         Thu, 19 Jun 2008 16:33:30 -0400
Reply-To:     Gene Maguin <emaguin@buffalo.edu>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Gene Maguin <emaguin@buffalo.edu>
Subject:      Re: Can I Mimic the Solver in Excel Within SPSS?
In-Reply-To:  <47009A583B6592479DBC0C6DCD40DF9503C6AC0A@mission.harris.harrisinteractive.com>
Content-Type: text/plain; charset="us-ascii"

Zachary,

My big question is understanding your dataset.
Your description seems to be saying two different things. You begin ...

>>I have two continuous variables that are standardized.  When I cross
them I wind up with a 2 X 2 matrix.

This suggests a data set looking like

Id   x   y
01  -3  12
02  18  -8
03  -2  -4
Etc.

>>I find all of this stuff using CROSSTABS or a variation of CTABLES.  Now
I will do this crosstab for about 40 different breaks.  Let us assume
each break is a country within the world.

Ok so you have another column for the country variable. So you could have
recoded two variables, (x and y in my example) and crosstabbed them against
country as a third variable. So far nothing special. But then you say ...


>>So visualize, if you will, 40
rows followed by four columns of information.  Each row will have the
ROWPCT for the respective row so that they add up to 100% for each row.
Each of the columns designates which matrix the person falls into.

This seems to describe a new dataset which looks like this, I think

Country cell11 cell12 cell21 cell22
France    12     45     21     22
Taiwan    81     02     11     06
Etc.

So this reminds me of crosstab tables written to a file using one of the
crosstab options or else, output from an OMS operation. Is it?  Anyway, you
say

>>Now comes the real tricky part.  I do not know if 0.0 is the optimal
place to break my distribution.  I have an ideal 40 X 4 matrix that I
want my results to match so my goal is to use something like the SOLVER
to minimize the gap between my solution and the ideal that has been
submitted to me.

At this point, and to begin with, which dataset are you referring to? If the
second one, your request doesn't make sense. If the first one, what does
this 'ideal' solution look like? This brings up the second big question:
what are you really trying to do?

I'll need better explanation to understand what are you starting with and
what are you are trying to do.

Gene Maguin

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD