Local Max-Cut algorithm test environment with Python package NetworkX

Author: Dhamma Kimpara

Python 3.5.2

Packages: NetworkX, random, copy, matplotlib, tkinter, numpy

AMS subject classifications: 68Q25, 90C10, 90C27

The complexity of local max-cut with FLIP-neighborhood is known for finite simple graphs with maximum degree 3 and maximum degree 5 and above. Hence this repo works towards establishing a polynomial time algorithm for graphs with maximum degree 4. (Elsaesser and Tobias)

files: poljakTest.py, GraphGen.py

Testing feasibility of Poljak's (SICOMP 1995) method on finite simple graphs with maximum degree 4. outputs objective function vector c and matrix m as text file for input into mathematica LinearProgramming(c,m,b,lu,dom) solver.

files: algTest.py, Alg.py, Verifier.py

Provided test environment for problem with algorithm input, solution verifier, error identification, and graph display.

Poljak, S. 1995: http://dl.acm.org/citation.cfm?id=210959

Elsaesser and Tobias: https://arxiv.org/abs/1004.5329 for an overview

Work done at Operations Research at the Technical University of Munich