pettingzoo

Pistonball - PettingZoo Documentation

from pettingzoo.butterfly import pistonball_v6
env = pistonball_v6.env(render_mode="human")
env.reset(seed=42)
i = 0
for agent in env.agent_iter(12000):
    observation, reward, termination, truncation, info = env.last()

    if termination or truncation:
        action = None
    else:
        action = env.action_space(agent).sample() # this is where you would insert your policy
        
    env.step(action) 
    i = i+1
env.close()
print(i)

from pettingzoo.butterfly import pistonball_v6
parallel_env = pistonball_v6.parallel_env(render_mode="human",max_cycles=159)
observations = parallel_env.reset(seed=42)
i = 0
while parallel_env.agents:
    # this is where you would insert your policy
    actions = {agent: parallel_env.action_space(agent).sample() for agent in parallel_env.agents}  
    print(    actions)
    observations, rewards, terminations, truncations, infos = parallel_env.step(actions)
    i = i+1
    print(observations['piston_19'].shape)
parallel_env.close()
print(i)

发现跑一会就结束，原来是max_cycles默认值是125？

/home/r/Pictures/Screenshot from 2023-05-30 11-28-09.png