A* Pac-Bot

Overview

Pac-Bot AI is a game that explores AI navigation through a grid-based maze environment. Inspired by the classic Pac-Man, the game features a player-controlled Pac-Man searching for food while being chased by four ghost agents. The project incorporates a fully AI-controlled multi-agent system where entities use the A* Search algorithm for pathfinding, leveraging heuristics and path costs for efficiency. Path costs are adapted as a proximity system in A* for Pac-bot to avoid ghosts. Additionally, the project implements BFS and DFS algorithms to provide comparison metrics for evaluating performance and efficiency.

Features

• Search Algorithms: Implements A*, BFS, and DFS for efficient pathfinding.
• Multi-Agent System: 1 Pac-Bot AI vs 4 Ghost AIs.
• Multiple Levels: Test different algorithm combinations.
• Food System: Randomly positioned respawnable food pellets in valid spaces.
• Cost Proximity: Path costs placed adjacent to ghosts, being costly for A* Pac-bot pathfind near it.
• Performance Metrics: Evaluates steps taken, time, food consumed and ram usage.
• Game Simulator: Simulate 50 games per AI matchup for metrics collection.

Interaction: How to Play

1. Menu Navigation:

   • Use the Up/Down Arrow Keys to select a difficulty level.
   • Use the Left/Right Arrow Keys to select the Pac-Bot AI algorithm (A*, BFS, or DFS).
   • Press Enter to start the game.

2. Observations:

   • Pac-Bot (yellow circle) will automatically navigate the maze to collect food (green circles) while avoiding ghosts (red circles).
   • The ghosts use different algorithms (DFS, BFS, or A*) to chase Pac-Bot, depending on the selected difficulty level.

3. Game Over:

   • The game ends if Pac-Bot collides with a ghost or the timer runs out.
   • Press ESC to quit the game after it ends.

4. Metrics:

   • During gameplay, the following metrics are displayed at the bottom of the screen:
     • Steps Taken: Number of steps Pac-Bot has moved.
     • Time Left: Remaining time in seconds.
     • Food Eaten: Number of food pellets collected.
     • Ram Used: Current memory usage.

Code Layout

• pacbot.py: Main game file that initializes and runs the game. Generates the maze layout, creates 1 pac-bot and 4 ghosts. Handles the algorithmic implementation of A*, BFS, and DFS. Also performs some metrics for steps taken, time, ram usage, and food count.
• memory_tracker.py: Measure ram usage metrics.
• simulations.py: Contains a simulator to simulate the game 50 times each per matchup, totaling 450 games.

Run Instructions

1. Install Python 3:
   • Ensure Python 3 is installed.
2. Install Required Libraries: Pygame-ce:
   • Run the following command:

     pip3 install pygame-ce

3. Running the Game:
   • Execute the following command in source directory:

     python3 pacbot.py

4. Running the Simulator:

• Use the following command:

python3 simulations.py

Requirements

• Python 3.x
• Libraries: (Only pygame-ce needs to be installed. The rest are standard libraries that should be imported.)
  • pygame-ce
  • heapq
  • random
  • deque
  • tracemalloc
  • csv

Data

No additional data files are necessary to run Pac-Bot AI. Though the simulations.py file does generate a results.csv file that provides some useful insights on Pac-Bot AI's metrics.