Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/coelhosilva/nuriqabe
nuriQabe is a Python package for solving Nurikabe puzzles with Q-Learning.
https://github.com/coelhosilva/nuriqabe
board-game q-learning
Last synced: 27 days ago
JSON representation
nuriQabe is a Python package for solving Nurikabe puzzles with Q-Learning.
- Host: GitHub
- URL: https://github.com/coelhosilva/nuriqabe
- Owner: coelhosilva
- License: mit
- Created: 2023-12-04T21:54:14.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2023-12-05T22:27:49.000Z (about 1 year ago)
- Last Synced: 2024-11-10T11:37:58.510Z (3 months ago)
- Topics: board-game, q-learning
- Language: Python
- Homepage:
- Size: 354 KB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# nuriQabe
`nuriQabe` is a Python package for solving Nurikabe puzzles with Q-Learning.
## Installation
### From this repo
Create a project folder and navigate to it:
1. ```mkdir nurikabe-qlearning-user```
2. ```cd nurikabe-qlearning-user```
From a terminal within the `nurikabe-qlearning-user` project folder, create a new `Python (>=3.10.4)` virtual environment:
3. ```python -m venv venv```
Activate the virtual environment:
4. Bash/Zsh: `source venv/bin/activate` | Windows CMD: `venv\Scripts\activate`
Install the nuriQabe package:
5. ```pip install git+https://github.com/coelhosilva/nuriqabe.git```
### From source
Alternatively, one may clone this repo and perform a local installation according to the following steps:
1. ```git clone https://github.com/coelhosilva/nuriqabe.git``` (HTTP) or ```git clone [email protected]:coelhosilva/nuriqabe.git``` (SSH)
2. ```cd nuriqabe```
3. ```python -m venv venv```
4. Bash/Zsh: `source venv/bin/activate` | Windows CMD: `venv\Scripts\activate`
5. ```pip install .```
## Getting started
`nuriQabe` provides a Q-Learning-based solver for the board game Nurikabe. The package is divided into the following modules: `board_instances`, with problem formulations and solutions; `game`, which implements the `NurikabeBoard` object along the game rules and deterministic strategies for progressing the game; `geometry`, with auxiliary functions for dealing with the grid, such as measuring distances; `graphics`, with plotting tools; and `solvers`, with the Q-Learning solver as well as a naive brute force alternative.
The following code exemplifies how to instantiate a game and solve it using the Q-Learning solver. For seeing the board charts, opt for a ">Jupyter: Run Current File in Interactive Window" run configuration in vscode or a "Run File In Python Console" in PyCharm. For that, also make sure the venv has the `ipython` and `ipykernel` installed.
```python
from nuriqabe.board_instances import read_board
from nuriqabe.solvers.qlearning import QLearningSolver
grid, solution = read_board(9999)
board_solver = QLearningSolver(problem_grid=grid)
board_solver.board.plot() # inspect base board, filled with deterministic rules
board_solver.solve(max_n_episodes=100, max_n_steps=200)
board_solver.plot_solution()
```
For inspecting the available board instances, please refer to the following code:
```python
from nuriqabe.board_instances import boards_summary
# Pandas DataFrame with id, shape, size, whether it is a square board, and n_islands
df_boards = boards_summary()
print(df_boards)
```