[Scip] Unbounded LP Relaxation

Ambros Gleixner gleixner at zib.de
Thu Feb 24 15:38:40 MET 2011


Hi Steve,

this is a SoPlex issue: because SoPlex cannot provide the primal ray,
SCIP refuses to continue the solving process and so your pricer is not
even called anymore to catch and handle the unbounded LP state.

In our developer version, we have now implemented a method to obtain the
primal ray for unbounded LPs.  I will send you a patched version of
SoPlex.  The next official release will also have this fixed.

Best regards,
ambros




Am 18.02.2011 06:25, schrieb Stephen J Maher:
> Hi All,
> 
> I am using SCIP to implement Benders decomposition. I am having some
> difficulty with some of the subproblems in that the LP relaxation
> becomes unbounded in some iterations of the program. When I run my
> program in debug mode I get the error
> 
> [src/scip/lpi_spx.cpp:3100] ERROR: SCIPlpiGetPrimalRay() not supported
> by SoPlex
> 
> What I was wondering is whether it is possible to identify at some stage
> in the SCIP solving process whether the LP is unbounded, and to stop the
> solving?
> 
> I am solving the subproblems with column generation so I have tried
> using the functions
> 
> SCIPlpiIsPrimalUnbounded and SCIPlpiIsDualUnbounded
> 
> however they both return false and then the error occurs.
> 
> Thanks a lot for your help,
> 
> Steve
> 
> 
> 
> _______________________________________________
> Scip mailing list
> Scip at zib.de
> http://listserv.zib.de/mailman/listinfo/scip

-- 
____________________________________________________________
Ambros M. Gleixner
Zuse Institute Berlin - Matheon - Berlin Mathematical School
http://www.zib.de/gleixner


More information about the Scip mailing list