Frozen Lake Solver

This project implements a reinforcement learning agent to solve the Frozen Lake environment from OpenAI Gym. It uses value iteration to find the optimal policy.

Installation

1. Clone the repository: git clone https://github.com/wuhungmao/frozen_lake.git
2. Navigate to the project directory: cd frozen-lake-solver
3. Create a virtual environment (recommended): python3 -m venv .venv
4. Activate the virtual environment: source .venv/bin/activate (Linux/macOS) or .venv\Scripts\activate (Windows)
5. Install the required dependencies: pip install -r requirements.txt

File Structure

• main.py: The main script that runs the value iteration algorithm and extracts the policy.
• frozen_lake.py: Contains the implementation of the Frozen Lake environment wrapper and the value iteration algorithm.
• test_frozen_lake.py: Contains the unit tests for the code.
• requirements.txt: Lists the project's dependencies (gym, numpy, etc.).

Usage

To run the main program:

python3 main.py --map_size 8 --is_slippery