gym-match3's Introduction

gym-match3

gym-match3 is an environment for reinforcement learning purposes. It replicates Match-3 board and allows to create your own levels.

Getting started

Installing

git clone https://github.com/kamildar/gym-match3.git
cd gym-match3
pip install -e .

Running the tests

For testing the environment after installation run following command in tests directory of a project:

python -m unittest test_game

Usage

Here’s a bare minimum example of getting something running.

from gym_match3.envs import Match3Env

env = Match3Env()
obs, reward, done, info = env.step(0)

For more information on gym interface visit gym documentation

Levels

Current number of levels is 30, but environment supports custom levels. Example:

from gym_match3.envs import Match3Env
from gym_match3.envs.levels import LEVELS #  default levels
from gym_match3.envs.levels import Match3Levels, Level


custom_level = Level(h=9, w=9, n_shape=6, board=[
    [-1, -1, -1, -1,  0, -1, -1, -1, -1],
    [-1, -1, -1,  0,  0,  0, -1, -1, -1],
    [-1, -1,  0,  0,  0,  0,  0, -1, -1],
    [-1,  0,  0,  0,  0,  0,  0,  0, -1],
    [ 0,  0,  0,  0,  0,  0,  0,  0,  0],
    [-1,  0,  0,  0,  0,  0,  0,  0, -1],
    [-1, -1,  0,  0,  0,  0,  0, -1, -1],
    [-1, -1, -1,  0,  0,  0, -1, -1, -1],
    [-1, -1, -1, -1,  0, -1, -1, -1, -1],
])

# create an instance with extended levels
custom_m3_levels = Match3Levels(levels=LEVELS + [custom_level]) 
env = Match3Env(levels=custom_m3_levels)

Recommend Projects