No, changing memory or platform is probably not the issue, as constraint
programming really reserves some traps and surprises for newcomers. First
fo all, you should _never_ have to generate all solutions first so you can
sort them. This is pretty much the whole point of using CSP. You should be
using one of the two "classic" Oz approaches to optimization (see tutorial
and below).

Please check the following to see if you are on the right path:

1 - you should not be using anything like SearchAll (Search.all etc.), are
you?

2 - is your cost function "materialized" as a completely constrained FD
variable before ANY distribution step (i.e. calls to FD.distribute or
FS.distribute)?

3 - Approach 1 - use Search.one and, as the _very first_ distribution
command in your script, do something like {FD.distribute naive
[CostVariable]}. The "naive" option only applies if what you want is to
minimize CostVaribale;

4 - Approach 2 - use Search.best. In this case, you should make sure that,
before any distribution step, the solution variable of the script has been
completely intantantiated and constrained.

If you have trouble understanding the above points, it would help if you
posted your code.

Cheers,

Jorge.

mostafa eslahi escreveu:
> Thanks you...
>   Indeed, my csp have much more than one million solutions and I'm looking
> for best of them w.r.t a cost function (my fault, I mean one million
> choice nodes).
>   I tried all instructions of recomputation, but nothing changed.
>   Do you have any idea?
>   What do you think about changing my memory or platform.(I use windows xp
> as operating system).
>   Best
>
>
>
>
> ---------------------------------
>
>
> Yahoo!         Canada Toolbar : Search from anywhere on         the web
> and bookmark your favourite sites. Download it now!
> _________________________________________________________________________________
> mozart-users mailing list
> [email protected]
> http://www.mozart-oz.org/mailman/listinfo/mozart-users


Jorge M. Pelizzoni
ICMC - Universidade de São Paulo

_________________________________________________________________________________
mozart-users mailing list                               
[email protected]
http://www.mozart-oz.org/mailman/listinfo/mozart-users

Reply via email to