help-glpk
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Help-glpk] IOS, Re: integer objective


From: Michael Hennebry
Subject: [Help-glpk] IOS, Re: integer objective
Date: Mon, 24 Nov 2003 12:41:14 -0600 (CST)

On Sat, 22 Nov 2003, Andrew Makhorin wrote:

> >If the cost function is known to be integer
> >because it is an integral combination of integers,
> >will GLPK (MIP and BCS) use this in pruning?
>
> Yes. The mip solver informs about that as follows:
>
> Integer optimization begins...
> Objective function is integral
> +    29: mip =     not found yet; lp =   2.520571739e+03 (1, 0)
> +   137: mip =   3.716000000e+03; lp =   2.520571739e+03 (31, 56)
>
> However, implied integer variables are not recognized.

Thank you for your reply.

Perhaps a third category of variable would be useful: implied integer.
The solver wouldn't branch on it, but its integrality could allow it
to be fixed on the basis of reduced cost.
It could also be used to determine whether the objective was integral.

> Starting from the version 4.2 glpk has been provided with a new
> component, IOS (Integer Optimization Suite), which replaces BCS, so the
> latter will be removed from the package. In IOS also there is a flag to
> indicate if the objective is integral.

Is there documentation on IOS?
I fgrepped the doc files from glpk-4.2
for both IOS and ios and didn't find it.

-- 
Mike   address@hidden
"History is written by the winners."  --  the losers





reply via email to

[Prev in Thread] Current Thread [Next in Thread]