help-glpk
[Top][All Lists]
Advanced

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

Re: [Help-glpk] Modelling Advice Request - Project Tasks


From: jeff . kantor
Subject: Re: [Help-glpk] Modelling Advice Request - Project Tasks
Date: Thu, 21 Jan 2010 18:09:37 +0300

On Jan 21, 2010 9:30am, Andrew Makhorin <address@hidden> wrote:
> > On that subject, a member of this list has privately sent me a model
> 
> > which does this.
> 
> 
> 
> > Thanks to EVERYBODY who took the time to respond.
> 
> 
> 
> > Here #39;s the new model, which uses a parameter for data input,
> 
> > minimises the number of variables needed, and works very successfully:
> 
> 
> 
> Mixed gomory cuts (i.e. glpsol --gomory) allows solving your new model
> 
> with glpsol about 20 times faster than without using cuts.
> 
> 
> 
> (Note that there are known more efficient formulations of the linear
> 
> ordering problem.)
> 
Indeed.  Adding constraints to remove cycles from the solution space
reduces the time about another 30% at the cost of increased memory requirements.
 I #39;m experimenting a bit right now, should
have an example model ready shortly.

Jeff

> 
> 
> 
> 
> 
> _______________________________________________
> 
> Help-glpk mailing list
> 
> address@hidden
> 
> http://lists.gnu.org/mailman/listinfo/help-glpk
> 
On Jan 21, 2010 9:30am, Andrew Makhorin <address@hidden> wrote:
> > On that subject, a member of this list has privately sent me a model
>
> > which does this.
>
>
>
> > Thanks to EVERYBODY who took the time to respond.
>
>
>
> > Here's the new model, which uses a parameter for data input,
>
> > minimises the number of variables needed, and works very successfully:
>
>
>
> Mixed gomory cuts (i.e. glpsol --gomory) allows solving your new model
>
> with glpsol about 20 times faster than without using cuts.
>
>
>
> (Note that there are known more efficient formulations of the linear
>
> ordering problem.)
>
Indeed. Adding constraints to remove cycles from the solution space
reduces the time about another 30% at the cost of increased memory requirements.
I'm experimenting a bit right now, should
have an example model ready shortly.

Jeff

>
>
>
>
>
> _______________________________________________
>
> Help-glpk mailing list
>
> address@hidden
>
> http://lists.gnu.org/mailman/listinfo/help-glpk
>
reply via email to

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