It works for Ubuntu 16.04. But I have not tested it with Windows or Mac.
Issue or contact me if you find any problem: gityiheng@gmail.com
apt-get install -y python3-pip python3-numpy python3-dev cmake zlib1g-dev libjpeg-dev xvfb libav-tools xorg-dev python-opengl libboost-all-dev libsdl2-dev swig
git clone https://github.com/openai/gym.git
cd gym
pip3 install -e .
cd ..
git clone https://github.com/GitYiheng/gym-multipendulum.git
cd ./gym-multipendulum
pip3 install -e .
cd ..
import gym
import gym_multipendulum
env = gym.make('multipendulum-v0')
env.reset()
done = False
reward = 0
while not done:
action = env.sample_action()
observation, reward, done, _ = env.step(action)
env.render()