IMP-MARL dataset

IMP-MARL dataset Authors: Pascal Leroy ([e-mail]), Pablo G. Morato ([e-mail]), Jonathan Pisane, Athanasios Kolios, Damien Ernst GitHub repository: https://github.com/moratodpg/imp_marl IMP-MARL environments: struct_uc: k-out-of-n system 2-out-of-3, 4-out-of-5, 9-out-of-10, 48-out-of-50, 95-out-of-100 (with and without campaign cost) struct_c: correlated k-out-of-n system 2-out-of-3, 4-out-of-5, 9-out-of-10, 48-out-of-50, 95-out-of-100 (with and without campaign cost) owf: offshore wind farm 1 wind turbine, 2 wind turbines, 5 wind turbines, 25 wind turbines, 50 wind turbines (with and without campaign cost) MARL methods investigated: QMIX, QVMIX, QPLEX, COMA, FACMAC, IQL, DQN. This dataset contains: MARL logs.zip => The configuration, execution, and result files of all investigated MARL methods. Each zip file contains the logs from an IMP-MARL environment. There, test and train files are stored for each method: cout.txt contains stored console ouputs. config.json contains the list of all parameter values for the experiments. info.json is the log file, listing reward values, for example. run.json contains simply logged information about the host, typically hardware. best_agent_networks.zip =>The controller networks’ weights of the best MARL-based policies. Controller files are stored as a pytorch file for each method/environment investigated (agent.th). heur_logs.zip => The configuration, execution, and result files of the computed expert-based heuristic policies. /heur_search => logs resulting from all evaluated heuristic policies, stored as numpy files. /heur_search => results from the best performing heuristics over 10,000 policy realisations, stored as pickle files.