def main(): '''Simple function to bootstrap a game. Use this as an example to set up your training env. ''' # Print all possible environments in the Pommerman registry print(pommerman.REGISTRY) # Create a set of agents (exactly four) agent_list = [ agents.BerlinAgent(), agents.DortmundAgent(), agents.BerlinAgent(), agents.DortmundAgent(), # agents.DockerAgent("pommerman/simple-agent", port=12345), ] # Make the "Free-For-All" environment using the agent list env = pommerman.make('PommeTeamCompetitionFast-v0', agent_list) wins = 0 ties = 0 survived_agents = [] nof_plays = 100 # Run the episodes just like OpenAI Gym for i_episode in range(nof_plays): print("Game " + str(i_episode)) state = env.reset(i_episode) done = False while not done: #env.render() actions = env.act(state) state, reward, done, info = env.step(actions) print(info) if info['result'] == pommerman.constants.Result.Tie: ties += 1 elif info['winners'] == [1, 3]: wins += 1 else: print(info['result']) print(info['winners']) print("Lost with seed: " + str(i_episode)) print('Episode {} finished'.format(i_episode)) survived_agents.extend(state[0]['alive']) env.close() survived_team_0 = survived_agents.count(10) + survived_agents.count(12) survived_team_1 = survived_agents.count(11) + survived_agents.count(13) kills = nof_plays * 2 - survived_team_0 death = nof_plays * 2 - survived_team_1 print("kills / death / ratio: ", kills, " / ", death, " / ", kills/max(0.1, death)) winRatio = str(wins / max(1, (nof_plays - ties))) print("wins: " + str(wins) + "/" + str(nof_plays - ties) + " = " + winRatio) file = open("/tmp/hypertune_result_winrate.txt", "w") file.write(winRatio) file.close() file = open("/tmp/hypertune_result_killdeath_diff.txt", "w") file.write(str(kills - death)) file.close()
def main(): '''Simple function to bootstrap a game. Use this as an example to set up your training env. ''' # Print all possible environments in the Pommerman registry print(pommerman.REGISTRY) # Create a set of agents (exactly four) agent_list = [ agents.BerlinAgent(), agents.DortmundAgent(), agents.BerlinAgent(), agents.DortmundAgent(), # agents.DockerAgent("pommerman/simple-agent", port=12345), ] # Make the "Free-For-All" environment using the agent list env = pommerman.make('PommeTeamCompetition-v0', agent_list) # Run the episodes just like OpenAI Gym for i_episode in range(1): state = env.reset(18) done = False while not done: env.render() #record_json_dir='/tmp/' actions = env.act(state) state, reward, done, info = env.step(actions) print(info) print('Episode {} finished'.format(i_episode)) env.close()