Minimal Function Cache is a repository for a proposed cut generation method in the thesis of Acadia Larsen which uses an explicit method select optimal cuts over a space of valid intersection cuts for a given MIP relaxation basis.
The repository is in an alpha state.
-
Experimental Cut and Branch Solver using explicitly optimized cuts over the space of minimal functions with a finite number of breakpoints.
-
Code for generating representative elements of minimal functions for a semialgebraic parametric description of minimal functions in the Gomory Johnson Group Problem using high performance computing cluster.
-
A cache of computed representative elements.
-
Reproducible computational experiments.
- Illustrate concept of explicit optimal cut selection.
- Reproducible on local and HPC machines.
- Mathematical correctness of cut generation.
- Interface with current optimization software
pyscipopt. - Demonstrate use of
passsagemathin application.
- Performance and code optimization; the cut generation technique presented is a proof of concept and is intended to function (for practical problem) with an excess of compute.
- Documentation light, code follows thesis.
The code is released under the GNU General Public License, version 2, or any later version as published by the Free Software Foundation.