A Python implementation of the Feasibility Pump heuristic
Feasibility Pump - find feasible solution for a Binary Programming problem.
The Feasibility Pump(FP) is a heuristic useful to find a feasible solution to Mixed Integer Programming problem. This implementation is specifically designed to work with Binary Programming problems. FP is not defined to find optimal solution nor a good one, it’s only purpose is to find a feasible solution in a small amount of time, even for those problem where a greedy algorithm cannot produce feasible result.
For a more detailed explanation please read the original paper
This project can be use as a stand alone application or as a library.
In order to find a feasible solution to a given 01MIP instance simply run
fp.py -i instance
the extension of the instance file will drive the back-end library to use the correct loader.
Available loaders can be obtained using:
fp.py -l
For more information about the usage you can:
fp.py -h
Read this docs to get more information about the currently know file formats:
this feature is not implemented at the current stage