Dataset for Minimum Spanning Tree Problem with Conflict Constraints
Description
This dataset provides all benchmark instances and source code used in the article “Minimum Spanning Tree Problem with Conflict Constraints (MSTC)”. It enables full reproducibility of the computational experiments reported in the paper and serves as a reference collection for future studies on MSTC and related combinatorial optimization problems. It contains the Restricted Test Set (RTS) of 48 small–medium instances, and the Extended Random Set (ERS) of 335 large-scale instances. The dataset also includes the source code of the instance generator for reproducibility of the MSTC instances, and the source code of the Stable Set Based Branch-and-bound Algorithm for the MSTC problem. Both source codes are written in C++ 17. These instances were generated to benchmark exact and heuristic algorithms for the MSTC problem. Researchers may reuse or extend the dataset to test new formulations, relaxations, or heuristics. STRONG and WEAK formulations are included to be used for benchmarking in both multi-thread and single-thread settings. They are solved with commercial solver Gurobi 10.0.3, which is called from a Python program. All data are provided in plain-text format for easy parsing. All experiments were executed on Windows 10 using Visual Studio 2022 (MSVC v17, C++17), as detailed in the article’s Online Supplement.
Files
Steps to reproduce
1. Download RTS.zip and/or ERS.zip. 2. Compile MSTC_instance_generator.cpp and MSTC_stablesetBB.cpp with Visual Studio 2022 (MSVC v17, C++17 standard). 3. Use the generator to recreate instances if desired. 4. Run the Branch-and-Bound solver (MSTC_stablesetBB.exe) on any instance file. 5. For benchmarking, compile the codes within MSTC_formulations.zip using a Python compiler (e.g. Anaconda Spyder 3). 6. Full computational settings, environment, and validation procedures are detailed in the Online Supplement (EJOR-D-25-00433). Related article (submitted): İzer, M.U., Öncan, T., Altınel, K., Finding conflict-free minimum weight spanning trees using maximal stable sets of the conflict graph, Submitted to European Journal of Operations Research. (2025).
Institutions
- Bogazici Universitesi Endustri Muhendisligi Bolumu
- Galatasaray Universitesi Muhendislik ve Teknoloji Fakultesi