[SCIP] Exact MIP

[Karl Däubel]

Dear SCIP-Team,

I recently came across the SCIP functionality to solve Mixed Integer 
Programs exactly, without any floating-point errors. Particularly, I 
found a link to a paper on the SCIP-website ([1]). In this paper the 
hybrid approach is advertised that tries to balance between safe 
floating point LP-bounds and, if needed, the exact rational LP-solution. 
The paper also states that all these approaches are implemented within 
the SCIP framework.

I tried to figure out how I could use this approach for a given MIP 
(lets say given in LP-format). Unfortunately I failed miserably. I could 
not find any parameters in the documentation that might be helpful nor 
was there anything related in the FAQ. While the paper was interesting 
to read, I could not find any guide how this can be enabled/used in SCIP.

Is there somewhere a step-by-step guide of something similar that I 
could follow to get started? Any kind of help would be highly 
appreciated. I am also sorry if I missed some obvious reference/link.

Also, can I compile SCIP to use multiple LP-solvers? As far as I 
understood it, an exact and an inexact solver were used for the hybrid 
approach.

Thank you in advance, Karl


[1]	A Hybrid Branch-and-Bound Approach for Exact Rational Mixed-Integer 
Programming
	William Cook, Thorsten Koch, Daniel E. Steffy, Kati Wolter
	Mathematical Programming Computation, 5(3), pp. 305-344, 2013
	https://link.springer.com/article/10.1007/s12532-013-0055-6

-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature
Type: application/pgp-signature
Size: 840 bytes
Desc: OpenPGP digital signature
URL: <http://listserv.zib.de/pipermail/scip/attachments/20210114/5cd0a022/attachment.asc>