[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Help-glpk] MIP stopping
From: |
Brady Hunsaker |
Subject: |
Re: [Help-glpk] MIP stopping |
Date: |
Wed, 18 May 2005 13:30:26 -0400 |
User-agent: |
Debian Thunderbird 1.0.2 (X11/20050331) |
Klas,
I think it makes sense to add the feature you mention; I don't think
GLPK has it already, but I haven't checked the code for it.
Realize, however, that your situation is a special case in which all
variables that appear in the objective are integer variables and all of
their objective coefficients are integers as well. The rule you
describe does not apply for general integer programs (even pure integer
programs).
Brady
Klas Markström wrote:
> Hi!
> I have been using glpsol to solve some
> minimisation integer programming problems. Right
> now the solver often finds the optimal integer
> solution with value opt fairly fast and then uses
> a lot of time to improve the lower bound from opt
> -1 to opt. Since the solver "knows" that this is
> an integer programming problem it should be able
> to stop as soon as it has improved the lower
> bound to anything strictly above the current
> optimum minus one.
>
> Is this going to be changed in some upcoming
> version of glpsol or is there some simple way to
> change the code so it will stop as soon as it
> knows that the current best is the optimum?
>
> /Klas
>
> --
>
> ==========================================================================
>
> Klas Markström email: address@hidden
> Department of Mathematics fax: (+46)90 786 52 22
> Umeå University phone: (+46)90 786 97 21
> S-901 87 Umea, Sweden
>
> URL: http://abel.math.umu.se/~klasm/
>
> ==========================================================================
>
>
--
Brady Hunsaker
Assistant Professor
Industrial Engineering
University of Pittsburgh
http://www.engr.pitt.edu/hunsaker/