以前に、WSLでOpenAI Gymを動かす記事を書いた。
その際は、Windows側にX Window Serverの「VcXsrv」をインストールして行ったが、Windows 11では、WSLgがX Window Serverの役割を果たすので、Windows側のセットアップなしにLinuxのGUIを扱えるようになった。
はじめ、WSLにDockerをインストールして、Dockerコンテナ上で動かそうとしたが、GUIアプリの起動はできるがウィンドウが表示されないため、あきらめて、WSLのUbuntu上に直接、環境構築した。
「wsl --export」でインストール直後のUbuntuの環境を保存しておくと、WSLに同じディストリビューションで複数環境を構築できるため、Dockerがなくてもそれほど困らないだろう。
OpenAI Gymの環境構築手順
Anacondaインストール
$ wget https://repo.anaconda.com/archive/Anaconda3-2021.11-Linux-x86_64.sh $ bash Anaconda3-2021.11-Linux-x86_64.sh
で、Anacondaをインストールする。
OpenAI Gymインストール
$ pip install gym
実行できるかipythonで、以下のコードを入力して確認する。
import gym
env = gym.make('CartPole-v0')
env.reset()
env.render()以下のエラーが表示される。
ModuleNotFoundError Traceback (most recent call last)
~/anaconda3/lib/python3.9/site-packages/gym/envs/classic_control/rendering.py in <module>
14 try:
---> 15 import pyglet
16 except ImportError as e:
ModuleNotFoundError: No module named 'pyglet'
During handling of the above exception, another exception occurred:
ImportError Traceback (most recent call last)
<ipython-input-4-0fb217d1bc14> in <module>
1 for _ in range(1000):
----> 2 env.render()
3 env.step(env.action_space.sample()) # take a random action
4 env.close()
~/anaconda3/lib/python3.9/site-packages/gym/core.py in render(self, mode, **kwargs)
293
294 def render(self, mode="human", **kwargs):
--> 295 return self.env.render(mode, **kwargs)
296
297 def close(self):
~/anaconda3/lib/python3.9/site-packages/gym/envs/classic_control/cartpole.py in render(self, mode)
177
178 if self.viewer is None:
--> 179 from gym.envs.classic_control import rendering
180
181 self.viewer = rendering.Viewer(screen_width, screen_height)
~/anaconda3/lib/python3.9/site-packages/gym/envs/classic_control/rendering.py in <module>
15 import pyglet
16 except ImportError as e:
---> 17 raise ImportError(
18 """
19 Cannot import pyglet.
ImportError:
Cannot import pyglet.
HINT: you can install pyglet directly via 'pip install pyglet'.
But if you really just want to install all Gym dependencies and not have to think about it,
'pip install -e .[all]' or 'pip install gym[all]' will do it.
pip install gym[all]をインストールしろということなので、実行すると、MuJoCoをインストールしろと言われる。
Running setup.py clean for mujoco-py
Failed to build box2d-py mujoco-py
Installing collected packages: lockfile, importlib-resources, glfw, pyglet, opencv-python, mujoco-py, lz4, box2d-py, ale-py
Running setup.py install for mujoco-py ... error
ERROR: Command errored out with exit status 1:
command: /home/xxxx/anaconda3/bin/python -u -c 'import io, os, sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/setup.py'"'"'; __file__='"'"'/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)(__file__) if os.path.exists(__file__) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' install --record /tmp/pip-record-s0ldhh8m/install-record.txt --single-version-externally-managed --compile --install-headers /home/xxxx/anaconda3/include/python3.9/mujoco-py
cwd: /tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/
Complete output (56 lines):
running install
running build
You appear to be missing MuJoCo. We expected to find the file here: /home/xxxx/.mujoco/mjpro150
This package only provides python bindings, the library must be installed separately.
Please follow the instructions on the README to install MuJoCo
https://github.com/openai/mujoco-py#install-mujoco
Which can be downloaded from the website
https://www.roboti.us/index.html
Traceback (most recent call last):
File "<string>", line 1, in <module>
File "/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/setup.py", line 32, in <module>
setup(
File "/home/xxxx/anaconda3/lib/python3.9/site-packages/setuptools/__init__.py", line 153, in setup
return distutils.core.setup(**attrs)
File "/home/xxxx/anaconda3/lib/python3.9/distutils/core.py", line 148, in setup
dist.run_commands()
File "/home/xxxx/anaconda3/lib/python3.9/distutils/dist.py", line 966, in run_commands
self.run_command(cmd)
File "/home/xxxx/anaconda3/lib/python3.9/distutils/dist.py", line 985, in run_command
cmd_obj.run()
File "/home/xxxx/anaconda3/lib/python3.9/site-packages/setuptools/command/install.py", line 61, in run
return orig.install.run(self)
File "/home/xxxx/anaconda3/lib/python3.9/distutils/command/install.py", line 546, in run
self.run_command('build')
File "/home/xxxx/anaconda3/lib/python3.9/distutils/cmd.py", line 313, in run_command
self.distribution.run_command(command)
File "/home/xxxx/anaconda3/lib/python3.9/distutils/dist.py", line 985, in run_command
cmd_obj.run()
File "/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/setup.py", line 28, in run
import mujoco_py # noqa: force build
File "/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/mujoco_py/__init__.py", line 3, in <module>
from mujoco_py.builder import cymj, ignore_mujoco_warnings, functions, MujocoException
File "/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/mujoco_py/builder.py", line 502, in <module>
mjpro_path, key_path = discover_mujoco()
File "/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/mujoco_py/utils.py", line 93, in discover_mujoco
raise Exception(message)
Exception:
You appear to be missing MuJoCo. We expected to find the file here: /home/xxxx/.mujoco/mjpro150
This package only provides python bindings, the library must be installed separately.
Please follow the instructions on the README to install MuJoCo
https://github.com/openai/mujoco-py#install-mujoco
Which can be downloaded from the website
https://www.roboti.us/index.html
----------------------------------------
ERROR: Command errored out with exit status 1: /home/xxxx/anaconda3/bin/python -u -c 'import io, os, sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/setup.py'"'"'; __file__='"'"'/tmp/pip-install-p63nh23w/mujoco-py_a2ff8ae5e2664451813a1a75ca3ccbfd/setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)(__file__) if os.path.exists(__file__) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' install --record /tmp/pip-record-s0ldhh8m/install-record.txt --single-version-externally-managed --compile --install-headers /home/xxxx/anaconda3/include/python3.9/mujoco-py Check the logs for full command output.MuJoCoは、最近オープンソースになったが、バージョン1.5はライセンスが必要なため、gym[all]ではなくgym[atari]をインストールすることにした。
$ pip install gym[atari]
再び、動作を確認したところ、今度は、python-openglをインストールしろと言われた。
ImportError: Library "GLU" not found.
During handling of the above exception, another exception occurred:
ImportError Traceback (most recent call last)
<ipython-input-4-d9761596d5d9> in <module>
----> 1 env.render()
~/anaconda3/lib/python3.9/site-packages/gym/core.py in render(self, mode, **kwargs)
293
294 def render(self, mode="human", **kwargs):
--> 295 return self.env.render(mode, **kwargs)
296
297 def close(self):
~/anaconda3/lib/python3.9/site-packages/gym/envs/classic_control/cartpole.py in render(self, mode)
177
178 if self.viewer is None:
--> 179 from gym.envs.classic_control import rendering
180
181 self.viewer = rendering.Viewer(screen_width, screen_height)
~/anaconda3/lib/python3.9/site-packages/gym/envs/classic_control/rendering.py in <module>
27 from pyglet.gl import *
28 except ImportError as e:
---> 29 raise ImportError(
30 """
31 Error occurred while running `from pyglet.gl import *`
ImportError:
Error occurred while running `from pyglet.gl import *`
HINT: make sure you have OpenGL installed. On Ubuntu, you can run 'apt-get install python-opengl'.
If you're running on a server, you may need a virtual frame buffer; something like this should work:
'xvfb-run -s "-screen 0 1400x900x24" python <your_script.py>'
$ pip install python-opengl
再度、動作を確認したことろ、無事CartPoleの画面が表示できた。
初回は数十秒くらい時間がかかった。
