Here you can find benchmark datasets developed in the project.
When you use them and publish some results, please cite us.
problem: Weak coloring numbers and uniform quasi-wideness
subtitle: Graphs, orderings, uniform quasi-wideness output
link: http://kernelization-experiments.mimuw.edu.pl/datasets/wcol-uqw-data.tgz
author: Wojciech Nadara, Marcin Pilipczuk, Felix Reidl, Roman Rabinovich, Sebastian Siebertz
Graphs, output orderings, and output uniform quasi-wideness results for the experiments on weak coloring numbers and uniform quasi-wideness. The graphs has been gathered from a number of sources, including Koblenz Network Collection, PACE 2016 Track B instances, random instances used by O’Brien and Sullivan, and random planar graphs. This package is licensed under the CC-BY-SA 2.0 license.
problem: Hamiltonian Cycle on graphs of bounded treewidth
subtitle: 7 random instances with a cycle
link: available at the repository https://github.com/stalowyjez/hc_tw_experiments
author: Michał Ziobro, Marcin Pilipczuk
7 instances of random graphs with a Hamiltonian Cycle sampled as follows: starting from a long cycle, a number of random edges with endpoints close to a cycle is added. The tests are intended to give advantage to the rank-based approach over the naive one.
problem: Hamiltonian Cycle on graphs of bounded treewidth
subtitle: Tree decompositions for the moderate-treewidth instances
link: available at the repository https://github.com/stalowyjez/hc_tw_experiments/tree/master/decompositions
author: Michał Ziobro, Marcin Pilipczuk
A number of moderate-width tree decompositions for 19 instances from the Flinders Hamiltonian Cycle Challenge, used in our experiments. They were computed using Strasser’s entry to PACE 2017 Track A.
problem: Feedback Vertex Set
subtitle: small toy instances
link: http://kernelization-experiments.mimuw.edu.pl/datasets/fvs-small-instances.tgz
author: Marcin Pilipczuk
A number of small instances, developed while working on the entry to PACE 2016 track B. Each instance (file with suffix .graph) is in the same format as at the challenge, and is accompanied with a .opt file with a sample optimum solution.