What to do if you get a DISPLAY error in gym.render ()

trouble

When I try to run the Open AI Gym sample code in a docker container, I get the following error message:

Terminal

Traceback (most recent call last):
  File "hello_gym.py", line 8, in <module>
    env.render()
  File "/usr/local/lib/python3.6/dist-packages/gym/core.py", line 240, in render
    return self.env.render(mode, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/gym/envs/classic_control/cartpole.py", line 174, in render
    from gym.envs.classic_control import rendering
  File "/usr/local/lib/python3.6/dist-packages/gym/envs/classic_control/rendering.py", line 25, in <module>
    from pyglet.gl import *
  File "/usr/local/lib/python3.6/dist-packages/pyglet/gl/__init__.py", line 244, in <module>
    import pyglet.window
  File "/usr/local/lib/python3.6/dist-packages/pyglet/window/__init__.py", line 1880, in <module>
    gl._create_shadow_window()
  File "/usr/local/lib/python3.6/dist-packages/pyglet/gl/__init__.py", line 220, in _create_shadow_window
    _shadow_window = Window(width=1, height=1, visible=False)
  File "/usr/local/lib/python3.6/dist-packages/pyglet/window/xlib/__init__.py", line 165, in __init__
    super(XlibWindow, self).__init__(*args, **kwargs)
  File "/usr/local/lib/python3.6/dist-packages/pyglet/window/__init__.py", line 570, in __init__
    display = pyglet.canvas.get_display()
  File "/usr/local/lib/python3.6/dist-packages/pyglet/canvas/__init__.py", line 94, in get_display
    return Display()
  File "/usr/local/lib/python3.6/dist-packages/pyglet/canvas/xlib.py", line 123, in __init__
    raise NoSuchDisplayException('Cannot connect to "%s"' % name)
pyglet.canvas.xlib.NoSuchDisplayException: Cannot connect to "None"

Execution environment

The execution environment is as follows.

I'm trying to set vnc in a Ubintu 18.04 based docker container and see the render () output in a browser.

Thoughts

--VNC is not set --X display is not set --gym sample code error (around render) --The version of pyglet does not match (pyglet was up to date)

Here is the sample code for the gym:

sample.py

import gym
env = gym.make('CartPole-v0')
env.reset()
for _ in range(1000):
    env.render()
    env.step(env.action_space.sample()) # take a random action
env.close()

(Actually, this code causes an error for another reason, but that's another article ...)

Cause

The DISPLAY was not set.

Check the environment variables in the terminal.

Terminal

$ printenv

The list is displayed in a row

Terminal

Abbreviation
LANG=C.UTF-8
OLDPWD=/home/user
VISIBLE=now
USER=user
PWD=/home/user/workspace
HOME=/home/user
Abbreviation

There is no indication of DISPLAY =. I should have set it ...

Setting the DISPLAY environment variable (example of failure)

I tried the following three methods, but they didn't work. If you want to know the solution quickly, Jump

Setting method 1

Originally it was set in the Dockerfile (I thought it was done)

Dockerfile

ENV DISPLAY=:0

But I couldn't set the environment variables in the container.

Setting method 2 Set with the option of docker run

The next method I tried was this method.

Terminal

docker run -e DISPLAY=$DISPLAY Below abbreviation

By the way, even if I changed -e to --env, it didn't work lol

Setting method 3 Set with startup.sh

startup.sh is, roughly speaking, a shell script that can be executed when a container is started. (I just don't understand w)

You can write the CMD command at the end of the Dockerfile, but if you want to execute multiple commands, Execute them by writing them in startup.sh.

Dockerfile

CMD ["/startup.sh"]

The reason I used startup.sh is that I wantedssh and vnc to be started by supervisord, so /usr/bin/supervisord -c /etc/supervisor/supervisord.conf is at least I had to run it in bash. Therefore, I wrote the command on two lines.

startup.sh

#!/bin/bash
/usr/bin/supervisord -c /etc/supervisor/supervisord.conf
export DISPLAY=:0

After all, this method did not resolve the error.

solution

Describe as follows in the Dockerfile.

RUN echo "export DISPLAY=:0"  >> /etc/profile

You have now set the environment variable inside the container to DISPLAY.

Summary

If you get the error pyglet.canvas.xlib.NoSuchDisplayException: Cannot connect to" None " when you run gym inside the docker container, The DISPLAY environment variable may not be set.

I was able to set it by executing RUN echo" export DISPLAY =: 0 ">>/etc/profile in the Dockerfile.