标签: openai-gym

我想知道CartPole-v0OpenAI Gym 中观察的规范（https://gym.openai.com/）。

例如，在以下代码中输出observation. 一种观察就像[-0.061586 -0.75893141 0.05793238 1.15547541]我想知道数字的含义。我想以任何方式知道其他的规范，Environments例如MountainCar-v0，MsPacman-v0等等。

我试图阅读https://github.com/openai/gym，但我不知道。你能告诉我知道规格的方法吗？

import gym
env = gym.make('CartPole-v0')
for i_episode in range(20):
    observation = env.reset()
    for t in range(100):
        env.render()
        print(observation)
        action = env.action_space.sample()
        observation, reward, done, info = env.step(action)
        if done:
            print("Episode finished after {} timesteps".format(t+1))
            break

（来自https://gym.openai.com/docs）

输出如下

[-0.061586   -0.75893141  0.05793238  1.15547541]
[-0.07676463 -0.95475889  0.08104189  1.46574644]
[-0.0958598  -1.15077434  0.11035682  1.78260485]
[-0.11887529 -0.95705275  0.14600892  1.5261692 …

当我尝试通过python pip在我的Windows机器上安装OpenAi Universe时,我得到以下stacktrace:

Traceback (most recent call last):
      File "<string>", line 1, in <module>
      File "C:\Users\Me\AppData\Local\Temp\pip-build-yjf_mrwx\fastzbarlight\setup.py", line 49, in <module>
        proc = subprocess.Popen(['ld', '-liconv'], stderr=subprocess.PIPE)
      File "E:\Python3.5.2\lib\subprocess.py", line 947, in __init__
        restore_signals, start_new_session)
      File "E:\Python3.5.2\lib\subprocess.py", line 1224, in _execute_child
        startupinfo)
    FileNotFoundError: [WinError 2] The system cannot find the file specified

而这个错误代码:

Command "python setup.py egg_info" failed with error code 1 in C:\Users\Me\AppData\Local\Temp\pip-build-yjf_mrwx\fastzbarlight\

我尝试了这里提到的一切.我还阅读了文档,发现了这个:

"While we don’t officially support Windows, we expect our code to be very …

我是强化学习的完全新手,并一直在寻找一个框架/模块,以轻松导航这个危险的地形.在我的搜索中,我遇到了两个模块keras-rl和OpenAI GYM.

我可以让他们两个在他们的WIKI上共享的示例上工作,但是它们带有预定义的环境,并且几乎没有关于如何设置我自己的自定义环境的信息.

如果有人能指出我的教程,或者只是向我解释如何设置非游戏环境,我真的很感激？

我想在笔记本中玩OpenAI体育馆，并且将体育馆内联渲染。

这是一个基本示例：

import matplotlib.pyplot as plt
import gym
from IPython import display
%matplotlib inline

env = gym.make('CartPole-v0')
env.reset()

for i in range(25):
   plt.imshow(env.render(mode='rgb_array'))
   display.display(plt.gcf())    
   display.clear_output(wait=True)
   env.step(env.action_space.sample()) # take a random action

env.close()

这行得通，我在笔记本中看到了健身房：

但！它还会打开一个交互式窗口，显示完全相同的内容。我不希望打开此窗口：

我根据 OpenAI Gym 框架创建了一个自定义环境；含step，reset，action，和reward功能。我的目标是在这个自定义环境上运行 OpenAI 基线。但在此之前，环境必须在 OpenAI 健身房注册。我想知道如何在 OpenAI 健身房注册自定义环境？另外，我是否应该修改 OpenAI 基线代码以包含此内容？

当我尝试从 docker google 计算引擎上的 docker 容器运行命令时，我遇到了此错误。

这是堆栈跟踪：

Traceback (most recent call last):
  File "train.py", line 16, in <module>
    from stable_baselines.ppo1 import PPO1
  File "/home/selfplay/.local/lib/python3.6/site-packages/stable_baselines/__init__.py", line 3, in <module>
    from stable_baselines.a2c import A2C
  File "/home/selfplay/.local/lib/python3.6/site-packages/stable_baselines/a2c/__init__.py", line 1, in <module>
    from stable_baselines.a2c.a2c import A2C
  File "/home/selfplay/.local/lib/python3.6/site-packages/stable_baselines/a2c/a2c.py", line 3, in <module>
    import gym
  File "/home/selfplay/.local/lib/python3.6/site-packages/gym/__init__.py", line 13, in <module>
    from gym.envs import make, spec, register
  File "/home/selfplay/.local/lib/python3.6/site-packages/gym/envs/__init__.py", line 10, in <module>
    _load_env_plugins()
  File "/home/selfplay/.local/lib/python3.6/site-packages/gym/envs/registration.py", line 269, in load_env_plugins
    context = contextlib.nullcontext()
AttributeError: module …

我正在与 RL 代理合作，并试图复制本文的发现，其中他们基于 Gym 开放 AI 制作了一个自定义跑酷环境，但是当我尝试渲染这个环境时遇到了问题。

import numpy as np
import time
import gym
import TeachMyAgent.environments

env = gym.make('parametric-continuous-parkour-v0', agent_body_type='fish', movable_creepers=True)
env.set_environment(input_vector=np.zeros(3), water_level = 0.1)
env.reset()

while True:
    _, _, d, _ = env.step(env.action_space.sample())
    env.render(mode='human')
    time.sleep(0.1)

c:\users\manu dwivedi\teachmyagent\TeachMyAgent\environments\envs\parametric_continuous_parkour.py in render(self, mode, draw_lidars)
    462 
    463     def render(self, mode='human', draw_lidars=True):
--> 464         from gym.envs.classic_control import rendering
    465         if self.viewer is None:
    466             self.viewer = rendering.Viewer(RENDERING_VIEWER_W, RENDERING_VIEWER_H)

ImportError: cannot import name 'rendering' from 'gym.envs.classic_control' (C:\ProgramData\Anaconda3\envs\teachagent\lib\site-packages\gym\envs\classic_control\__init__.py)

  [1]: https://github.com/flowersteam/TeachMyAgent

我认为这可能是这个自定义环境以及作者决定渲染它的方式的问题，但是，当我尝试

from gym.envs.classic_control …

我正在尝试在 Stable Baselines3 中使用自定义环境实现 SAC，但我不断收到标题中的错误。任何非策略算法都会发生该错误，而不仅仅是 SAC。

追溯：

File "<MY PROJECT PATH>\src\main.py", line 70, in <module>
  main()
File "<MY PROJECT PATH>\src\main.py", line 66, in main
  model.learn(total_timesteps=timesteps, reset_num_timesteps=False, tb_log_name=f"sac_{num_cars}_cars")
File "<MY PROJECT PATH>\venv\lib\site-packages\stable_baselines3\sac\sac.py", line 309, in learn
  return super().learn(
File "<MY PROJECT PATH>\venv\lib\site-packages\stable_baselines3\common\off_policy_algorithm.py", line 375, in learn
  self.train(batch_size=self.batch_size, gradient_steps=gradient_steps)
File "<MY PROJECT PATH>\venv\lib\site-packages\stable_baselines3\sac\sac.py", line 256, in train
  current_q_values = self.critic(replay_data.observations, replay_data.actions)
File "<MY PROJECT PATH>\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
  return forward_call(*input, **kwargs)
File "<MY PROJECT PATH>\venv\lib\site-packages\stable_baselines3\common\policies.py", line 885, in forward
  return tuple(q_net(qvalue_input) …

是否有可能将openai的健身房环境用于多智能体游戏？具体来说,我想模拟一个有四个玩家(代理商)的纸牌游戏.得分转弯的球员在下一回合开始.我如何模拟玩家之间的必要协调(例如,接下来是谁)？最终,我想对四个相互对抗的特工使用强化学习.

我想让下面的代码工作.

import gym

env = gym.make("CartPole-v0")

env.reset()

env.render()

运行前3行没有问题,但是当我运行第4行时,我得到错误:

Traceback (most recent call last):

  File "<ipython-input-3-a692a1a1ffe7>", line 1, in <module>
    env.render()

  File "/home/mikedoho/gym/gym/core.py", line 150, in render
    return self._render(mode=mode, close=close)

  File "/home/mikedoho/gym/gym/core.py", line 286, in _render
    return self.env.render(mode, close)

  File "/home/mikedoho/gym/gym/core.py", line 150, in render
    return self._render(mode=mode, close=close)

  File "/home/mikedoho/gym/gym/envs/classic_control/cartpole.py", line 116, in _render
    self.viewer = rendering.Viewer(screen_width, screen_height)

  File "/home/mikedoho/gym/gym/envs/classic_control/rendering.py", line 51, in __init__
    self.window = pyglet.window.Window(width=width, height=height, display=display)

  File "/home/mikedoho/anaconda3/lib/python3.6/site-packages/pyglet/window/__init__.py", line 504, in __init__
    screen = display.get_default_screen()

  File "/home/mikedoho/anaconda3/lib/python3.6/site-packages/pyglet/canvas/base.py", …