Python scenarios.load方法代碼示例

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_particle_env(game_name, benchmark=False):
    import multiagent.scenarios as scenarios

    scenario = scenarios.load(game_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:
        env = ParticleEnv(
            world,
            scenario.reset_world,
            scenario.reward,
            scenario.observation,
            scenario.benchmark_data,
        )
    else:
        env = ParticleEnv(
            world, scenario.reset_world, scenario.reward, scenario.observation
        )
    return env


# environment for all agents in the multiagent world
# currently code assumes that no agents will be created/destroyed at runtime! 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_particle_env(game_name, benchmark=False):
    import multiagent.scenarios as scenarios
    scenario = scenarios.load(game_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:
        env = ParticleEnv(world, scenario.reset_world, scenario.reward, scenario.observation, scenario.benchmark_data)
    else:
        env = ParticleEnv(world, scenario.reset_world, scenario.reward, scenario.observation)
    return env






# environment for all agents in the multiagent world
# currently code assumes that no agents will be created/destroyed at runtime! 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def parse_args():
    parser = argparse.ArgumentParser("Reinforcement Learning experiments for multiagent environments")
    # Environment
    parser.add_argument("--scenario", type=str, default="simple", help="name of the scenario script")
    parser.add_argument("--max-episode-len", type=int, default=25, help="maximum episode length")
    parser.add_argument("--num-episodes", type=int, default=60000, help="number of episodes")
    parser.add_argument("--num-adversaries", type=int, default=0, help="number of adversaries")
    parser.add_argument("--good-policy", type=str, default="maddpg", help="policy for good agents")
    parser.add_argument("--bad-policy", type=str, default="maddpg", help="policy of adversaries")
    # Core training parameters
    parser.add_argument("--lr", type=float, default=1e-2, help="learning rate for Adam optimizer")
    parser.add_argument("--gamma", type=float, default=0.95, help="discount factor")
    parser.add_argument("--batch-size", type=int, default=1024, help="number of episodes to optimize at the same time")
    parser.add_argument("--num-units", type=int, default=64, help="number of units in the mlp")
    parser.add_argument("--adv-eps", type=float, default=1e-3, help="adversarial training rate")
    parser.add_argument("--adv-eps-s", type=float, default=1e-5, help="small adversarial training rate")
    # Checkpointing
    parser.add_argument("--exp-name", type=str, default=None, help="name of the experiment")
    parser.add_argument("--save-dir", type=str, default="/tmp/policy/", help="directory in which training state and model should be saved")
    parser.add_argument("--save-rate", type=int, default=1000, help="save model once every time this many episodes are completed")
    parser.add_argument("--load-name", type=str, default="", help="name of which training state and model are loaded, leave blank to load seperately")
    parser.add_argument("--load-good", type=str, default="", help="which good policy to load")
    parser.add_argument("--load-bad", type=str, default="", help="which bad policy to load")
    # Evaluation
    parser.add_argument("--test", action="store_true", default=False)
    parser.add_argument("--restore", action="store_true", default=False)
    parser.add_argument("--display", action="store_true", default=False)
    parser.add_argument("--benchmark", action="store_true", default=False)
    parser.add_argument("--benchmark-iters", type=int, default=100000, help="number of iterations run for benchmarking")
    parser.add_argument("--benchmark-dir", type=str, default="./benchmark_files/", help="directory where benchmark data is saved")
    parser.add_argument("--plots-dir", type=str, default="./learning_curves/", help="directory where plot data is saved")
    return parser.parse_args() 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_env(scenario_name, arglist, benchmark=False):
    from multiagent.environment import MultiAgentEnv
    import multiagent.scenarios as scenarios

    # load scenario from script
    scenario = scenarios.load(scenario_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation, scenario.benchmark_data)
    else:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation)
    return env 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_env(scenario_name, benchmark=False):
    '''
    Creates a MultiAgentEnv object as env. This can be used similar to a gym
    environment by calling env.reset() and env.step().
    Use env.render() to view the environment on the screen.

    Input:
        scenario_name   :   name of the scenario from ./scenarios/ to be Returns
                            (without the .py extension)
        benchmark       :   whether you want to produce benchmarking data
                            (usually only done during evaluation)

    Some useful env properties (see environment.py):
        .observation_space  :   Returns the observation space for each agent
        .action_space       :   Returns the action space for each agent
        .n                  :   Returns the number of Agents
    '''
    from multiagent.environment import MultiAgentEnv
    import multiagent.scenarios as scenarios

    # load scenario from script
    scenario = scenarios.load(scenario_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:        
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation, scenario.benchmark_data)
    else:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation)
    return env 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_env(scenario_name, benchmark=False):
    '''
    Creates a MultiAgentEnv object as env. This can be used similar to a gym
    environment by calling env.reset() and env.step().
    Use env.render() to view the environment on the screen.

    Input:
        scenario_name   :   name of the scenario from ./scenarios/ to be Returns
                            (without the .py extension)
        benchmark       :   whether you want to produce benchmarking data
                            (usually only done during evaluation)

    Some useful env properties (see environment.py):
        .observation_space  :   Returns the observation space for each agent
        .action_space       :   Returns the action space for each agent
        .n                  :   Returns the number of Agents
    '''
    from multiagent.environment import MultiAgentEnv
    import multiagent.scenarios as scenarios

    # load scenario from script
    scenario = scenarios.load(scenario_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation, scenario.benchmark_data)
    else:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation,
                            done_callback=scenario.done)
    return env 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def parse_args():
    parser = argparse.ArgumentParser("Reinforcement Learning experiments for multiagent environments")
    # Environment
    parser.add_argument("--scenario", type=str, default="simple", help="name of the scenario script")
    parser.add_argument("--max-episode-len", type=int, default=25, help="maximum episode length")
    parser.add_argument("--num-episodes", type=int, default=60000, help="number of episodes")
    parser.add_argument("--num-adversaries", type=int, default=0, help="number of adversaries")
    parser.add_argument("--good-policy", type=str, default="maddpg", help="policy for good agents")
    parser.add_argument("--adv-policy", type=str, default="maddpg", help="policy of adversaries")
    # Core training parameters
    parser.add_argument("--lr", type=float, default=1e-2, help="learning rate for Adam optimizer")
    parser.add_argument("--gamma", type=float, default=0.95, help="discount factor")
    parser.add_argument("--batch-size", type=int, default=1024, help="number of episodes to optimize at the same time")
    parser.add_argument("--num-units", type=int, default=64, help="number of units in the mlp")
    # Checkpointing
    parser.add_argument("--exp-name", type=str, default=None, help="name of the experiment")
    parser.add_argument("--save-dir", type=str, default="/tmp/policy/", help="directory in which training state and model should be saved")
    parser.add_argument("--save-rate", type=int, default=1000, help="save model once every time this many episodes are completed")
    parser.add_argument("--load-dir", type=str, default="", help="directory in which training state and model are loaded")
    # Evaluation
    parser.add_argument("--restore", action="store_true", default=False)
    parser.add_argument("--display", action="store_true", default=False)
    parser.add_argument("--benchmark", action="store_true", default=False)
    parser.add_argument("--benchmark-iters", type=int, default=100000, help="number of iterations run for benchmarking")
    parser.add_argument("--benchmark-dir", type=str, default="./benchmark_files/", help="directory where benchmark data is saved")
    parser.add_argument("--plots-dir", type=str, default="./learning_curves/", help="directory where plot data is saved")
    return parser.parse_args() 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_env(scenario_name, benchmark=False):
    # load scenario from script
    scenario = scenarios.load(scenario_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation, scenario.benchmark_data)
    else:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation)
    return env 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_multiagent_env(env_id, num_agents, dist_threshold, arena_size, identity_size):
    scenario = scenarios.load(env_id+".py").Scenario(num_agents=num_agents, dist_threshold=dist_threshold,
                                                     arena_size=arena_size, identity_size=identity_size)
    world = scenario.make_world()

    env = MultiAgentEnv(world=world, 
                        reset_callback=scenario.reset_world, 
                        reward_callback=scenario.reward, 
                        observation_callback=scenario.observation,
                        info_callback=scenario.info if hasattr(scenario, 'info') else None,
                        discrete_action=True,
                        done_callback=scenario.done,
                        cam_range=arena_size
                        )
    return env 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_env(scenario_name, benchmark=False):
    '''
    Creates a MultiAgentEnv object as env. This can be used similar to a gym
    environment by calling env.reset() and env.step().
    Use env.render() to view the environment on the screen.

    Input:
        scenario_name   :   name of the scenario from ./scenarios/ to be Returns
                            (without the .py extension)
        benchmark       :   whether you want to produce benchmarking data
                            (usually only done during evaluation)

    Some useful env properties (see environment.py):
        .observation_space  :   Returns the observation space for each agent
        .action_space       :   Returns the action space for each agent
        .n                  :   Returns the number of Agents
    '''
    from multiagent.environment import MultiAgentEnv
    import multiagent.scenarios as scenarios

    # load scenario from script
    scenario = scenarios.load(scenario_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:        
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation, scenario.benchmark_data, done_callback = scenario.done)
    else:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward, scenario.observation, done_callback = scenario.done)
    return env 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_env(scenario_name, benchmark=False, discrete_action=False):
    '''
    Creates a MultiAgentEnv object as env. This can be used similar to a gym
    environment by calling env.reset() and env.step().
    Use env.render() to view the environment on the screen.

    Input:
        scenario_name   :   name of the scenario from ./scenarios/ to be Returns
                            (without the .py extension)
        benchmark       :   whether you want to produce benchmarking data
                            (usually only done during evaluation)

    Some useful env properties (see environment.py):
        .observation_space  :   Returns the observation space for each agent
        .action_space       :   Returns the action space for each agent
        .n                  :   Returns the number of Agents
    '''
    from multiagent.environment import MultiAgentEnv
    import multiagent.scenarios as scenarios

    # load scenario from script
    scenario = scenarios.load(scenario_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if benchmark:        
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward,
                            scenario.observation, scenario.benchmark_data,
                            discrete_action=discrete_action)
    else:
        env = MultiAgentEnv(world, scenario.reset_world, scenario.reward,
                            scenario.observation,
                            discrete_action=discrete_action)
    return env 

# 需要導入模塊: from multiagent import scenarios [as 別名]
# 或者: from multiagent.scenarios import load [as 別名]
def make_env(scenario_name, benchmark=False, discrete_action=False):
    '''
    Creates a MultiAgentEnv object as env. This can be used similar to a gym
    environment by calling env.reset() and env.step().
    Use env.render() to view the environment on the screen.

    Input:
        scenario_name   :   name of the scenario from ./scenarios/ to be Returns
                            (without the .py extension)
        benchmark       :   whether you want to produce benchmarking data
                            (usually only done during evaluation)

    Some useful env properties (see environment.py):
        .observation_space  :   Returns the observation space for each agent
        .action_space       :   Returns the action space for each agent
        .n                  :   Returns the number of Agents
    '''
    from multiagent.environment import MultiAgentEnv
    import multiagent.scenarios as old_scenarios
    import envs.mpe_scenarios as new_scenarios

    # load scenario from script
    try:
        scenario = old_scenarios.load(scenario_name + ".py").Scenario()
    except:
        scenario = new_scenarios.load(scenario_name + ".py").Scenario()
    # create world
    world = scenario.make_world()
    # create multiagent environment
    if hasattr(scenario, 'post_step'):
        post_step = scenario.post_step
    else:
        post_step = None
    if benchmark:        
        env = MultiAgentEnv(world, reset_callback=scenario.reset_world,
                            reward_callback=scenario.reward,
                            observation_callback=scenario.observation,
                            post_step_callback=post_step,
                            info_callback=scenario.benchmark_data,
                            discrete_action=discrete_action)
    else:
        env = MultiAgentEnv(world, reset_callback=scenario.reset_world,
                            reward_callback=scenario.reward,
                            observation_callback=scenario.observation,
                            post_step_callback=post_step,
                            discrete_action=discrete_action)
    return env