python - OpenAI Gym: env.reset() throwing OSError

Question

When I call env.reset() on 'SuperMatioBros2-vo', I get an OSError. Here is my code -

## Setting up the game

import gym
from nes_py.wrappers import JoypadSpace
import gym_super_mario_bros
from gym_super_mario_bros.actions import SIMPLE_MOVEMENT, COMPLEX_MOVEMENT
env = gym_super_mario_bros.make('SuperMarioBros2-v0')
env = JoypadSpace(env, COMPLEX_MOVEMENT)
done = True
for step in range(2500):
    if done:
        state = env.reset()
    state, reward, done, info = env.step(env.action_space.sample())
    env.render()
env.close()
env.reset()

Here is the error that I have been receiving -

OSError                                   Traceback (most recent call last)
<ipython-input-6-c5672cf456ee> in <module>
----> 1 env.reset()

~anaconda3libsite-packages
es_pywrappersjoypad_space.py in reset(self)
     76     def reset(self):
     77         """Reset the environment and return the initial observation."""
---> 78         return self.env.reset()
     79 
     80     def get_keys_to_action(self):

~anaconda3libsite-packagesgymwrappersime_limit.py in reset(self, **kwargs)
     23     def reset(self, **kwargs):
     24         self._elapsed_steps = 0
---> 25         return self.env.reset(**kwargs)

~anaconda3libsite-packages
es_py
es_env.py in reset(self)
    256         # reset the emulator
    257         if self._has_backup:
--> 258             self._restore()
    259         else:
    260             _LIB.Reset(self._env)

~anaconda3libsite-packages
es_py
es_env.py in _restore(self)
    218     def _restore(self):
    219         """Restore the backup state into the NES emulator."""
--> 220         _LIB.Restore(self._env)
    221 
    222     def _will_reset(self):

OSError: exception: access violation reading 0x000000000003C200

However, this issue doesn't seem to arise with the Cartpole environment -

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
    if done:
        state = env.reset()
    
env.close()
env.reset()

I am using env.reset() for a different purpose in my code. But this is what I came up with inorder to build a minimum reproducible example.

question from:https://stackoverflow.com/questions/65890165/openai-gym-env-reset-throwing-oserror

Answer 1

Waitting for answers