Minesweeper with AI and Pygame

Introduction

This project is a Python implementation of the classic Minesweeper game, enhanced with an AI-powered solver. The AI uses logical reasoning to deduce the positions of mines and safe cells. The game is designed with a graphical interface created using Pygame.

---

Features

1. Classic Minesweeper Gameplay:

   • Reveal cells to find mines.
   • Right-click to flag cells as mines.
   • Avoid detonating a mine to win the game.

2. AI Solver:

   • The AI uses a knowledge base and logical reasoning to identify safe cells or mines.
   • Incorporates:
     • Safe move deduction: Based on known information, marks certain cells as safe.
     • Random move fallback: If no logical move is available, the AI makes a random move.
     • Inference system: The AI learns from every revealed cell and updates its knowledge base.

3. Graphical User Interface:

   • Interactive grid for gameplay.
   • Visual elements for flags, mines, and revealed cells.
   • Dynamic buttons for:
     • Autoplaying with AI.
     • Showing AI inferences (safe and mine predictions).
     • Resetting the game.

4. Autoplay:

   • Watch the AI automatically solve the Minesweeper board step-by-step.

---

Getting Started

Prerequisites

• Python 3.x
• Install the following Python libraries:

  pip install pygame

Files in the Project

1. minesweeper.py: Contains the game logic and AI logic.
2. main.py: Runs the graphical interface using Pygame.
3. Assets:
   • Fonts: Located in assets/fonts/ (e.g., OpenSans-Regular.ttf).
   • Images: Located in assets/images/ (e.g., flag.png, mine.png).

---

How to Run

1. Clone or Download the Repository:

   • Clone the repository or download it as a zip file.

2. Navigate to the Project Directory:

   cd <project-directory>

3. Run the Game:

   python runner.py

---

Game Instructions

Objective:

Mark all the mines on the board without triggering any of them.

Controls:

• Left-Click: Reveal a cell.
• Right-Click: Flag a cell as a mine.
• Buttons:
  • Autoplay: Lets the AI take over and solve the board.
  • AI Move: Allows the AI to make a single move.
  • Reset: Resets the game to start over.
  • Show Inference: Highlights AI's safe and mine predictions.

Win Condition:

• Flag all mines correctly without triggering any.

Lose Condition:

• Clicking on a mine detonates it, ending the game.

---

AI Algorithm

Core Concepts:

1. Knowledge Base:

   • The AI maintains a set of logical sentences about the game board.
   • Each sentence represents a relationship between a group of cells and the number of mines in them.

2. Inference:

   • The AI deduces:
     • Safe cells: Cells that can be revealed without risk.
     • Mines: Cells that must contain mines.
   • It uses techniques like subset reasoning to infer new information.

3. Fallback:

   • When no safe move can be logically deduced, the AI chooses a random unexplored cell.

---

Code Overview

minesweeper.py

1. Minesweeper Class:

   • Handles the game board and mine placement.
   • Tracks the number of nearby mines for any cell.
   • Provides utility methods for checking mines and determining win/loss states.

2. Sentence Class:

   • Represents logical statements about cells and their mine counts.
   • Provides methods for marking cells as safe or mines, enabling the AI to refine its knowledge base.

3. MinesweeperAI Class:

   • Implements the AI logic to deduce safe cells and mines.
   • Maintains knowledge of all moves made, safe cells, and mines.
   • Dynamically updates its knowledge base with each new revelation.

---

runner.py

1. Game Setup:

   • Initializes the Minesweeper board and AI.
   • Loads assets (e.g., images and fonts) for rendering.

2. Graphical User Interface:

   • Displays the Minesweeper board, cells, and interactive buttons.
   • Highlights flagged cells, revealed cells, and inferred safe/mine cells.

3. Gameplay Loop:

   • Handles user input for cell clicks and button presses.
   • Integrates AI to make safe or random moves during autoplay.

---

Project Structure

📂 Project Folder
├── 📂 assets
│   ├── 📂 fonts
│   │   └── OpenSans-Regular.ttf
│   ├── 📂 images
│       ├── flag.png
│       ├── mine.png
│       ├── mine-red.png
├── .gitignore
├── minesweeper.py
├── README.py
└── runner.py

---

Future Improvements

• Add support for customizable grid sizes and mine counts through a user interface.
• Improve AI inference for edge cases with low mine densities.
• Include a leaderboard to track game completions and performance metrics.

---

Acknowledgments

• Pygame: For creating an intuitive library to design 2D games.
• Inspiration: The classic Minesweeper game and AI algorithms based on logical reasoning.