mancala 0.2.1

Mancala written in Python, playable in CLI (GUI coming soon)!

Mancala

Mancala board game written in python.

Features & Road maps

• Mancala playable on CLI
• Cmpatible with the gym API
• Can train RL agents
• Mancala playable on GUI

Installation

$ pip install mancala

Usage

Play a game with agents

$ mancala play --player0 human --player1 random

Compare each agents and plot their win rates

The values are player0's (first move) win rates in percentage

$ mancala arena
            p0_random  p0_exact  p0_max  p0_minimax
p1_random        40.0      12.0     2.0         0.0
p1_exact         76.0      38.0    29.0         2.0
p1_max           91.0      55.0    27.0         1.0
p1_minimax       99.0      87.0    81.0        28.0

Algorithms

Mancala is a game with perfect information. マンカラは完全情報ゲームです。

Value Iteration

Using Dynamic Programming (DP), calculate value for states and memorize them. Use the value to plan future actions.

Other implementations

• OpenSpiel value_iteration algorithm https://github.com/deepmind/open_spiel/blob/master/open_spiel/python/algorithms/value_iteration.py example https://github.com/deepmind/open_spiel/blob/master/open_spiel/python/examples/value_iteration.py

Policy Iteration

Using Dynamic Programming (DP), calculate value for states and memorize them. Use the value and policy for planning.

References

• https://github.com/mdavolio/mancala

Multi agent RL

• https://github.com/deepmind/open_spiel
• https://github.com/PettingZoo-Team/PettingZoo