This dataset contains 3000 small parity games (games-small) and 3000 large parity games (games-large), all of which were randomly generated using [PGSolver]. For the small and large parity game, the numbers of vertices were uniformly sampled from {10,…,200} and {200,…,3000}, respectively. Moreover, for each game with n vertices, the colors as well as the relative outdegrees of its vertices as well as its outdegree were uniformly sampled from {0,…,n} and {n/100,…n/2}, respectively. Structure of the Dataset This dataset contains two zip-archives, namely games-small.zip and games-large.zip, each containing the respective dataset. Each zip-file contains the directories games and solutions. Each directory games contains files game_0000.txt through game_2999.txt, where each file contains one parity game in the input format understood by PGSolver. Moreover, each directory solutions contains files solution_game_0000.txt through solution_game_2999.txt, where solutions/solution_game_XXXX.txt contains the solution for games/game_XXXX.txt, generated as follows: pgsolver -global recursive games/game_XXXX.txt > solutions/solution_game_XXXX.txt

{"references": ["O. Friedmann, M. Lange. Solving Parity Games in Practice. In Proc. of the 7th Int. Symp. on Automated Technology for Verification and Analysis, ATVA'09, Macao, China, 2009 volume 5799 of Lecture Notes in Computer Science, pages 182-196, Springer, 2009. DOI: 10.1007/978-3-642-04761-9_15", "Hecking, T. gnn_pr_solver [Computer software]. https://github.com/DLR-SC/gnn_pr_solver. DOI: 10.5281/zenodo.7198473"]}