[Scip] problems with solving LPs

Stefan Heinz heinz at zib.de
Mon Aug 31 10:47:46 MEST 2009


Hi Frank,

the problem we know of with the objective limit does not hold for CPLEX. 

Since we do not have Xpress I have no experience how SCIP works with Xpress 
together.

Can you send me a log file?

Best Stefan


Am Montag 31 August 2009 schrieb Frank Hennig:
> Hi Stefan,
>
> I use cplex 11.2.
> Xpress performed even worse with respect to the problems (or the
> coupling between Xpress and SCIP is not as comfortable as between cplex
> and SCIP).
>
> Is there a bugfix too?
>
> Frank
>
> -----Original Message-----
> From: Stefan Heinz [mailto:heinz at zib.de]
> Sent: 31. august 2009 10:38
> To: scip at zib.de
> Cc: Frank Hennig
> Subject: Re: [Scip] problems with solving LPs
>
> Hi Frank,
>
> my first question is which LP solver did you link to SCIP?
>
> On the other hand there is Bug in SCIP Version 1.1 with respect to the
> objective limits for the LP solver.
>
>
> Best Stefan
>
> Am Freitag 28 August 2009 schrieb Frank Hennig:
> > Hi,
> >
> >
> >
> > I run a Branch-and-Price algorithm that solves very many LPs in its
> > nodes.
> >
> >
> >
> > Unfortunatelly I very often encounter the error message:
> >
> >
> >
> > src/scip/solve.c:670] ERROR: Error <-6> in function call
> > [src/scip/solve.c:1812] ERROR: Error <-6> in function call
> > [src/scip/solve.c:2443] ERROR: Error <-6> in function call
> > [src/scip/solve.c:3018] ERROR: Error <-6> in function call
> > [src/scip/scip.c:5221] ERROR: Error <-6> in function call
> >
> >
> >
> >
> >
> > which leads to abortion of the SCIPsolve routine.
> >
> >
> >
> > One possible reason is: [src/scip/lp.c:10772] debug: unresolved error
> > while resolving LP in order to exceed the objlimit
> >
> >
> >
> >
> >
> > The errors like to occur in the SCIP heuristics (preferably diving)
>
> but
>
> > also during solving a node.
> >
> >
> >
> >
> >
> > Is their any chance to improve the performance?
> >
> >
> >
> >
> >
> > This costs time and also may lead to random solvability.
> >
> >
> >
> >
> >
> > Thank you very much.
> >
> >
> >
> > Regards,
> >
> >
> >
> > Frank
> >
> >
> >
> >
> >
> >
> >
> > ________________________________________________________
> >
> >
> >
> > Frank Hennig  -  Naval Architect, PhD Candidate
> >
> >
> >
> > Department of Industrial Economics and Technology Management
> >
> > Norwegian University of Science and Technology
> >
> > Alfred Getz veg 3
> >
> > NO-7491 Trondheim, Norway
> >
> >
> >
> > Phone:   +49 30 314 23 598
> >
> > Email:      frank.hennig at iot.ntnu.no




More information about the Scip mailing list