EC2025W2 - MAXSAT Evolutionary Algorithm Solver 🧠

This project implements an Improved Binary Genetic Algorithm (IBGA) to solve Weighted MAXSAT problems using heuristic initialization and optional Lagrangian penalty ranking.

---

📂 Project Structure

EC2025W2_LCC/
├── Code/
│   ├── eda_105.ipynb            # Jupyter notebook for EDA or experiments
│   ├── IBGA.py                  # Improved Binary Genetic Algorithm implementation
│   ├── main.py                  # Main execution script
│   ├── maxsat.py                # MAXSAT problem definition and evaluation
├── Data/
│   ├── sbox_4.wcnf              # Example Weighted CNF instance (WCNF format)
├── Docs/
│   ├── images/                  # Images for report or visualization
│   ├── maxsat_report.pdf        # Final report on MAXSAT results
│   ├── optuna_eda104_lagrangian_results.csv  # Results of hyperparameter tuning or EDA
├── requirements.txt             # Python dependencies for pip installation
├── LICENSE.TXT                  # Project license (MIT or similar)
└── README.md                    # Project documentation

---

⚙️ Requirements

• Python 3.12+
• Recommended: Run inside a Conda environment

✅ Install dependencies:

pip install -r requirements.txt

---

🚀 How to Run the Project (Locally)

✅ Step 1: Activate your environment

conda activate env_evo

✅ Step 2: Navigate to the Code folder

cd Code

✅ Step 3: Run the evolutionary algorithm with your WCNF instance

python main.py --path ../Data/sbox_4.wcnf

✅ You can also provide an absolute path if preferred:

python main.py --path C:\Users\loren\Documents\ec2025w2\Data\sbox_4.wcnf

---

📈 Expected Output

• Percentage of satisfied clauses
• Best individual (binary string)
• Fitness score (number of satisfied clauses)
• Detailed DataFrame showing clause satisfaction

Example:

Percentage of Satisfied Clauses: 98.75%
Best Individual: [1, 0, 1, 1, 0, 1, ...]
Number of Satisfied Clauses: 395

---

📝 How to Export Environment Dependencies

✅ Export requirements.txt

conda activate env_evo
pip freeze > requirements.txt

---

📜 Author

Lorenzo - MSc Artificial Intelligence and Machine Learning
University of Birmingham | 2025

---

🗒️ License

MIT License