Error when using env.render() along with stable-baselines3

[shivakanthsujit]

Hello,

In the gym.Env class, the first argument is defined as the mode, which is either "human" or "rgb_array". In Crafter, the render function takes only one argument, size. This causes errors when using stable-baselines3 to evaluate policies and create videos since their helper functions assume envs follow gym.Env and have the first argument as the mode.

A fix would be to just add a dummy first argument mode to the render function

crafter/crafter/env.py

Lines 120 to 122 in e955b11

	def render(self, size=None):
	size = size or self._size
	unit = size // self._view

[romnatall]

its didnt work for me, still this error

[romnatall]

to solve this

recorder.py

def reset(self):
    obs = self._env.reset()
    self._frames = [self._env.render( size=self._size)]
    return obs

  def step(self, action):
    obs, reward, done, info = self._env.step(action)
    self._frames.append(self._env.render( size=self._size))
    if done:
      self._save()
    return obs, reward, done, info

env.py

#in __init__
self.metadata = {
        "render_modes": [ "rgb_array"],
        "render_fps": 4,
    }

def render( self ,mode="rgb_array", size=None):
    size = size or self._size
    unit = size // self._view
    canvas = np.zeros(tuple(size) + (3,), np.uint8)
    local_view = self._local_view(self._player, unit)
    item_view = self._item_view(self._player.inventory, unit)
    view = np.concatenate([local_view, item_view], 1)
    border = (size - (size // self._view) * self._view) // 2
    (x, y), (w, h) = border, view.shape[:2]
    canvas[x: x + w, y: y + h] = view

    return canvas.transpose((1, 0, 2))

  def _obs(self):
    return self.render(None)

also delete pycache folder in crafter and gym

[danijar]

Thanks! I'll leave this open in case it's helpful for other people. I don't know if updating the env is worth it at this point since OpenAI seems to have already changed the Gym interface in the meantime.

[yhs0602]

The changed code provided here works well with latest sb3 and latest gym. What about merging the PR #25 ?