Dear Guido, 

thanks for the answer!

* Guido Tack <[EMAIL PROTECTED]> [2007-04-14 21:38]:
> There is support for optimization in Gecode, using branch&bound  
> search. You have to define a function "constrain" in your Space. The  
> function gets one argument, which is the so-far best solution found,  
> and posts a constraint that any subsequent solution must be better,  
> pretty much like the constraints you mention above. Typically, you  
> use this to minimize or maximize a single variable, and post the real  
> objective function (like a weighted sum or an n-ary max) as a  
> constraint.
> 
> For simple examples, have a look at the photo or golomb source code.

It sounds like this is exactly what I expected. I have to take a closer look
to the mentioned examples and adjust them to my problem. Thanks a lot for the
hint!.


Cheers
stanio

_______________________________________________
Gecode users mailing list
[EMAIL PROTECTED]
https://www.gecode.org/mailman/listinfo/gecode-users

Reply via email to