Skip to main content

Gym Environment for 3D Tic Tac Toe

Project description

gym-tic-tac-toe3D


OpenAI Gym Environment for a Two-Player 3D Tic-Tac-Toe. Github Link: https://github.com/OUStudent/gym_tic_tac_toe3D

Requirements


  • gym
  • Numpy
  • Matplotlib

Install


pip install gym-tic-tac-toe3D

How it Works

Tic Tac Toe is usually played on a 3x3 grid where the objective is for one player to line up their tokens in a straight line of three. This is an extremely easy and trivial game; however, one can extend the difficulty by stacking 3x3 layers to create a 3x3x3 cube. Now the objective is to line up three tokens in any of the directions.

Here are three example games where the action was randomly generated:

Game 1

Game 2

Game 3

How to Use

The environment is a Two-Player Game, where Blue denotes Player 1 and Red denotes Player 2. The state and action contains all 27 possible positions, 9 for the first layer, 9 for the second, and 9 for the third.

The input has three possible integer values for eac position, -1 for opponent, 0 for empty, and 1 for current player. Note that no matter whom the Player is, these values holds true. The reward is a two value list where the first index represents the reward for Player 1 and the second for Player 2. The reward value should only be used after the game is completed. Players are rewarded for wins and receive extra points for how fast they win; while players are penalized for losing and how fast they lost. For example, after a game has completed the final reward could be [20, -10], which rewards Player 1 20 points while penalizing Player 2 -10 points.

Here is an example on how to create the environment:

import gym
import gym_tic_tac_toe3D
env = gym.make("tic_tac_toe3D-v0")

games = 3  # best of three
player1_reward = 0
player2_reward = 0
for i in range(0, games):
    state = env.reset()
    done = False
    player = 1
    while not done:
        env.render(player=player)
        plt.pause(0.5)
        while True:
            action = env.action_space.sample()
            # Need to check if action is available in state space
            if state[action] == 0:  
                break
    
        state, reward, done, info = env.step(action, player=player)
        # switch players
        if player == 1:
            player = 2
        else:
            player = 1
    # final render after completion of game to see final move
    env.render(player=player)
    plt.pause(1)
    player1_reward += reward[0]
    player2_reward += reward[1]
    
    

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

gym_tic_tac_toe3D-0.0.2.tar.gz (2.5 kB view details)

Uploaded Source

File details

Details for the file gym_tic_tac_toe3D-0.0.2.tar.gz.

File metadata

  • Download URL: gym_tic_tac_toe3D-0.0.2.tar.gz
  • Upload date:
  • Size: 2.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.6.1 pkginfo/1.7.1 requests/2.25.0 requests-toolbelt/0.9.1 tqdm/4.54.1 CPython/3.9.1

File hashes

Hashes for gym_tic_tac_toe3D-0.0.2.tar.gz
Algorithm Hash digest
SHA256 12c70f8bb9f362fa7d55ac5bd55ad9d44438670ee6702ea2df0efd638f2ba0b3
MD5 340bf60e6172c9db506509c67ce2ea56
BLAKE2b-256 0b8f4d7bc0646f8a646076d0162a521e821403aafa44936e2443adef93ffa370

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page