---

Mutant Reduction Framework

This repository provides an automated pipeline for mutant reduction in mutation testing, leveraging FSM-based mutation generation, feature extraction, clustering, and equivalent mutant detection. The framework helps optimize mutation testing by identifying and pruning redundant mutants, ultimately reducing computational costs.

---

Table of Contents

• 📌 Installation
• 📂 Project Structure
• 🚀 Pipeline Overview
• 🔧 Running the Pipeline
  • 1️⃣ Mutation Testing
  • 2️⃣ Feature Extraction
  • 3️⃣ Clustering
  • 4️⃣ Equivalent Mutant Detection
  • 5️⃣ Equivalent Mutant Pruning
  • 6️⃣ Evaluation
• 📊 Results
• ❓ Troubleshooting
• 👥 Contributors

---

📌 Installation

Ensure your system is updated and install the necessary dependencies:

sudo apt update
sudo apt install python3.8 python3.8-venv python3.8-dev default-jre

Setting Up the Virtual Environment

python3.8 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install -r requirements.txt

Install Additional Dependencies if Needed

pip install pytest-mutagen hdbscan matplotlib numpy pandas sklearn pyqt5

---

📂 Project Structure

/mutantReduction
|-- data/
|   |-- output/
|       |-- clustering/                  # Clustering results (DBSCAN/HDBSCAN/KMeans)
|       |-- equivalence_testing/         # Test execution data for equivalent mutant analysis
|       |-- evaluation_mutants/          # Selected mutants for evaluation
|       |-- evaluation_results.json      # Mutation testing evaluation results
|       |-- equivalent_mutants.json      # Identified equivalent mutants
|       |-- features.json                # Extracted features from mutants
|       |-- fsm_transitions.json         # FSM transition states for mutants
|       |-- mutants/                     # Generated mutants
|       |-- pruned_equivalent_mutants.json # Pruned mutants after equivalence analysis
|       |-- selected_mutants.json        # Final selected mutants after pruning
|
|-- src/
|   |-- benchmarks/                      # Benchmarking tools
|   |-- clustering/                       # Clustering algorithms (HDBSCAN, KMeans)
|   |-- equivalent_mutants/               # Equivalent mutant detection & pruning
|   |-- feature_extraction/               # Structural metrics extraction
|   |-- fsm_modeling/                     # FSM-based mutation testing
|   |-- mutation_testing/                 # Mutation testing execution & evaluation
|
|-- requirements.txt                      # Project dependencies
|-- README.md                             # Project documentation

---

🚀 Pipeline Overview

The framework follows these key steps:

1️⃣ Mutation Testing → Generate FSM-based mutants
2️⃣ Feature Extraction → Extract structural & behavioral features
3️⃣ Clustering → Identify redundant mutants via HDBSCAN/KMeans
4️⃣ Equivalent Mutant Detection → Detect semantically equivalent mutants
5️⃣ Equivalent Mutant Pruning → Keep only the most representative mutants
6️⃣ Evaluation → Run mutation testing on the pruned set

---

🔧 Running the Pipeline

1️⃣ Mutation Testing

Generates FSM-based mutants and stores them in data/output/mutants/.

python -m src.mutation_testing.mutpy_integration

📌 Output:

• Mutants stored in data/output/mutants/
• FSM transition data in data/output/fsm_transitions.json

---

2️⃣ Feature Extraction

Extracts structural metrics (e.g., function calls, arithmetic operations, conditionals, etc.).

python src/feature_extraction/structural_metrics.py

📌 Output:

• Extracted features stored in data/output/features.json

---

3️⃣ Clustering

Clusters mutants using HDBSCAN or KMeans.

Run HDBSCAN Clustering

python src/clustering/hdbscan_clustering.py

Run KMeans Clustering

python src/clustering/kmeans_clustering.py

📌 Output:

• data/output/clustering/hdbscan_clustering.png
• data/output/clustering/kmeans_clustering.png
• Cluster assignments in data/output/kmeans_cluster_assignments.json

---

4️⃣ Equivalent Mutant Detection

Detects equivalent mutants using:

• Structural analysis (code similarity)
• Behavioral analysis (test execution comparison)

python src/equivalent_mutants/equivalent_mutant_detector.py

📌 Output:

• Equivalent mutants stored in data/output/equivalent_mutants.json

---

5️⃣ Equivalent Mutant Pruning

Removes redundant mutants while keeping the most representative ones.

python src/equivalent_mutants/equivalent_mutant_pruner.py

📌 Output:

• Pruned mutants stored in data/output/pruned_equivalent_mutants.json
• Selected mutants stored in data/output/selected_mutants.json
• Evaluation mutants moved to data/output/evaluation_mutants/

---

6️⃣ Evaluation

Runs mutation testing only on the pruned set of mutants.

python src/mutation_testing/mutpy_evaluation.py

📌 Output:

• Evaluation results in data/output/evaluation_results.json

---

📊 Results

After evaluation, compare:

• Mutation Score Before vs. After Reduction
• Impact of Clustering (HDBSCAN vs. KMeans)
• Effectiveness of Equivalent Mutant Pruning
• Computational Savings from Pruning Redundant Mutants

---

❓ Troubleshooting

Dependency Issues

If dependencies are missing, install them:

pip install -r requirements.txt

Matplotlib GUI Issues

If Matplotlib fails to render plots, install PyQt5:

pip install pyqt5

Missing Java for AST Extraction

Ensure Java is installed:

sudo apt install default-jre

---

👥 Contributors

• Felipe Orlando & João Paulo Nogueira - Research & Implementation

Project Presentation

Applying Machine Learning to Optimize Mutation Testing in Python Projects