Mujoco-py: TypeError: 'numpy.intc' object is not iterable

[pedagil]

Describe the bug
I'm able to use env that requires mujoco such as 'CartPole-v1', but can't render any of Fetch robotic env. This is likely something like TypeError:'numpy.int32' object is not iterable for make('FetchReach-v1') #324 . Fix fix advanced mujoco envs for windows path separation #1220 Does not work for me

To Reproduce
import gym # openAi gym
from gym import envs

env = gym.make('FetchSlide-v1')
env.reset()
for _ in range(1000):
env.render()
env.step(env.action_space.sample()) # take a random action
env.close()

Error Messages
Traceback (most recent call last):
File "C:/Users/petrosp/PycharmProjects/FetchReach/FetchReach.py", line 4, in

env = gym.make('FetchSlide-v1')

File "C:\Users\petrosp\Anaconda3\envs\MainEnv\lib\site-packages\gym\envs\registration.py", line 156, in make
return registry.make(id, **kwargs)

File "C:\Users\petrosp\Anaconda3\envs\MainEnv\lib\site-packages\gym\envs\registration.py", line 101, in make
env = spec.make(**kwargs)

File "C:\Users\petrosp\Anaconda3\envs\MainEnv\lib\site-packages\gym\envs\registration.py", line 73, in make
env = cls(**_kwargs)

File "C:\Users\petrosp\Anaconda3\envs\MainEnv\lib\site-packages\gym\envs\robotics\fetch\slide.py", line 24, in __init__
initial_qpos=initial_qpos, reward_type=reward_type)

File "C:\Users\petrosp\Anaconda3\envs\MainEnv\lib\site-packages\gym\envs\robotics\fetch_env.py", line 48, in __init__
initial_qpos=initial_qpos)

File "C:\Users\petrosp\Anaconda3\envs\MainEnv\lib\site-packages\gym\envs\robotics\robot_env.py", line 36, in __init__
self._env_setup(initial_qpos=initial_qpos)

File "C:\Users\petrosp\Anaconda3\envs\MainEnv\lib\site-packages\gym\envs\robotics\fetch_env.py", line 172, in _env_setup
self.sim.data.set_joint_qpos(name, value)

File "mujoco_py\generated/wrappers.pxi", line 2547, in mujoco_py.cymj.PyMjData.set_joint_qpos
TypeError: 'numpy.intc' object is not iterable

Desktop (please complete the following information):

• OS: Windows 10
• Python Version 3.7.5
• Mujoco Version [e.g. 1.50]
• mujoco-py version [e.g. 1.50.1.68]

[ysd007]

I also have this problem.

[TneitaP]

Same problem happened in:

• Windows 10
  -Python Version 3.7.6
• Mujoco200
• mujoco-py 2.0.2.9

[TneitaP]

@pedagil I cloned the current master of mujoco-py, changed all isinstance(addr, (int, np.int32, np.int64))|-> hasattr(addr, '__int__') in mujoco-py\mujoco_py\generated\wrappers.pxi as discussed at TypeError:'numpy.int32' object is not iterable.
Finally, I used pip install -e . in the cloned mujoco-py mater repo, and successfully created the hand related repo, e.g.env = gym.make('HandManipulatePenTouchSensorsDense-v0')

[leey127]

I have the same problem, on Windows 10. Seems that the solutions for TypeError:'numpy.int32' object is not iterable. can not work for this issue.

[Iskandor]

@pedagil I cloned the current master of mujoco-py, changed all isinstance(addr, (int, np.int32, np.int64))|-> hasattr(addr, '__int__') in mujoco-py\mujoco_py\generated\wrappers.pxi as discussed at TypeError:'numpy.int32' object is not iterable.
Finally, I used pip install -e . in the cloned mujoco-py mater repo, and successfully created the hand related repo, e.g.env = gym.make('HandManipulatePenTouchSensorsDense-v0')

Worked for me for FetchReach-v1 environment. Thanks a lot.

[KindleKin]

@pedagil I cloned the current master of mujoco-py, changed all isinstance(addr, (int, np.int32, np.int64))|-> hasattr(addr, '__int__') in mujoco-py\mujoco_py\generated\wrappers.pxi as discussed at TypeError:'numpy.int32' object is not iterable.
Finally, I used pip install -e . in the cloned mujoco-py mater repo, and successfully created the hand related repo, e.g.env = gym.make('HandManipulatePenTouchSensorsDense-v0')

Where is your setup.py of mujoco-py ? How to pip install -e .

[KindleKin]

@pedagil I cloned the current master of mujoco-py, changed all isinstance(addr, (int, np.int32, np.int64))|-> hasattr(addr, '__int__') in mujoco-py\mujoco_py\generated\wrappers.pxi as discussed at TypeError:'numpy.int32' object is not iterable.
Finally, I used pip install -e . in the cloned mujoco-py mater repo, and successfully created the hand related repo, e.g.env = gym.make('HandManipulatePenTouchSensorsDense-v0')

Where is your setup.py of mujoco-py ? How to pip install -e .

Oh，I've got it！

[paulaxa1]

@pedagil I cloned the current master of mujoco-py, changed all isinstance(addr, (int, np.int32, np.int64))|-> hasattr(addr, '__int__') in mujoco-py\mujoco_py\generated\wrappers.pxi as discussed at TypeError:'numpy.int32' object is not iterable.
Finally, I used pip install -e . in the cloned mujoco-py mater repo, and successfully created the hand related repo, e.g.env = gym.make('HandManipulatePenTouchSensorsDense-v0')

I have no words. Not all heroes wear capes. Thank you.