Installing OpenAI Gym on Ubuntu

Installing OpenAI Gym on Ubuntu

OpenAI Gym is a toolkit for developing and comparing reinforcement learning algorithms. Here’s how you can install OpenAI Gym on Ubuntu.

Reference : 

• https://gym.openai.com/
• https://github.com/openai/gym#installing-everything
• https://gym.openai.com/docs
• http://www.modulabs.co.kr/RL_library/1705  #모두의 연구소

1) Install Required Packages

Before installing OpenAI Gym, you'll need to install some required dependencies. Open a terminal and run the following command:

$ sudo apt-get install -y python-numpy python-dev cmake zlib1g-dev libjpeg-dev xvfb libav-tools xorg-dev python-opengl libboost-all-dev libsdl2-dev swig

These packages include necessary libraries for handling things like numpy, CMake, image processing, Xvfb (for virtual displays), and OpenGL.

2) Install OpenAI Gym

1. Clone the OpenAI Gym repository from GitHub:

   bash
   git clone https://github.com/openai/gym.git
   

2. Navigate to the cloned directory:

   bash
   cd gym
   

3. Install Gym with additional dependencies. If you’re using Python 3.5, run the following command:

   bash
   pip install -e '.[all]'
   

   This will install Gym along with all optional dependencies (such as environments for Atari, robotics, and more).

4. If you are using Python 2.7, run the following:

   bash
   pip2.7 install -e '.[all]'
   

   This will install Gym and the necessary dependencies for Python 2.7.

3) RL Test Code: Simple Reinforcement Learning with OpenAI Gym

Now that you have OpenAI Gym installed, let's go through a simple Reinforcement Learning (RL) test example using the CartPole-v1 environment. In this example, we will implement a basic random agent that will take random actions in the environment.

Test Code for Reinforcement Learning (Random Agent)

import gym

env = gym.make('CartPole-v0')

env.reset()

for _ in range(1000):

    env.render()

    env.step(env.action_space.sample()) # take a random actio

4) 실행

$ python CartPole.py # 파이썬 2.7 $ python3.5 CartPole.py # 파이썬 3.5