Empirical Analysis of Runtime Distributions of Neighbourhood Search for the Graph Coloring Problem — Code and Data

Abstract

Code and experimental data accompanying the paper "Empirical Analysis of Runtime Distributions of Neighbourhood Search for the Graph Coloring Problem." Contains Python and C implementations of four neighbourhood search algorithms (Random Walk, TABUCOL, PARTIALCOL, Ising heuristic), experiment runners, raw result data (300 trials × 4 algorithms × 27 planted-partition instances + 14 DIMACS-equivalent instances), and figure generation scripts.