ocplib-simplex

A library implementing a simplex algorithm, in a functional style, for solving systems of linear inequalities and optimizing linear objective functions

Overview

ocplib-simplex is a (fully) functional OCaml implementation of the simplex algorithm for solving systems of linear inequalities. The implementation is incremental and backtrackable. It is able to extract unsat-cores for unsatisfiable problems. Versions > 0.1 also support linear optimization

Dependencies

ocplib-simplex requires 4.01.0 or higher and ocamlfind. You can use make opam-deps to install dependencies in the current switch.

Build and Install Instructions

The easiest way to install ocplib-simplex is to use OPAM:

$ opam install ocplib-simplex

If you want to install ocplib-simplex from sources, use the following instructions:

$ make opam-deps (if you are using OPAM and some deps are missing)
$ make
$ make install

to compile and install ocplib-simplex on your system. You can uninstall the library with make uninstall.

Minimal Examples

Solving a system of linear inequalities: see the file tests/standalone_minimal.ml

Linear optimization: see the file tests/standalone_minimal_maximization.ml

Contributing

Don't hesitate to report encountered bugs on this Git repo's issues tracker.

TODO

• the code is not (well) documented,

• some parts of the code need factorization/simplification,

• some invariants (check unsat-core, linear optimization) are missing.

Licensing

ocplib-simplex is Copyright (C) --- OCamlPro. it is distributed under the terms of the GNU Lesser General Public License (LGPL) version 2.1 (see LICENSE file for more details).