Super small 4 x 4 gridworld, re-worked for compatibility in OpenAI gym environment.
Original grid world implementation by sudarshanseshadri (https://github.com/sudarshanseshadri/gridworld)
To install:
git clone https://github.com/kristychoi/gym_gridworld.git </b>
cd gym_gridworld
pip3 install -e .
To use (x,y) state space:
import gym
import gym_gridworld
env = gym.make('gym_gridworld-v0')
To use one-hot encoding of states (same grid world):
import gym
import gym_gridworld
env = gym.make('gym_onehotgrid-v0')
TODOs:
- implement rendering
- better initialization of environment
- code cleanup