datamutant 0.1.0

A Python library for advanced data analysis algorithms

DataMutant

A Python library for advanced data analysis algorithms with a modular architecture that supports multiple specialized tools.

🧬 What is DataMutant?

DataMutant is designed to be a comprehensive toolkit for data scientists and researchers, providing efficient implementations of advanced algorithms for data analysis, pattern recognition, and machine learning. Each algorithm is organized as a separate module with its own examples and documentation.

🚀 Features

• Modular Architecture: Each algorithm is self-contained with examples and detailed documentation
• High Performance: Optimized implementations using PyTorch
• Comprehensive Testing: Full test coverage for all algorithms
• Type Safety: Complete type hints and validation
• Extensible Design: Easy to add new algorithms

📦 Available Algorithms

1. Sublinear Monotonicity Score

• Location: datamutant.monotonicity_score
• Purpose: Measure how close sequences are to being monotonically increasing
• Complexity: O((1/ε) log n) - sublinear sampling algorithm
• Documentation: View detailed docs

📚 Installation

pip install datamutant

For development:

pip install datamutant[dev]

🔥 Quick Start

import torch
from datamutant.monotonicity_score import sublinear_monotonicity_score

# Analyze sequence monotonicity
sequence = torch.tensor([1.0, 2.0, 3.0, 2.5, 4.0])
score = sublinear_monotonicity_score(sequence)
print(f"Monotonicity score: {score}")  # ~0.8

🏗️ Project Structure

datamutant/
├── datamutant/                          # Main package
│   ├── __init__.py                      # Package exports
│   └── monotonicity_score/              # Monotonicity algorithm
│       ├── __init__.py                  # Module exports
│       ├── sublinear.py                 # Core algorithm implementation
│       ├── models.py                    # Neural network models
│       ├── README.md                    # Algorithm documentation
│       └── examples/                    # Usage examples
│           ├── basic_usage.py
│           └── model_training.py
├── tests/                               # Comprehensive tests
│   └── monotonicity_score/
│       ├── test_core.py
│       └── test_models.py
├── setup.py                             # Package setup
├── pyproject.toml                       # Modern packaging
└── README.md                            # This file

📖 Algorithm Documentation

Each algorithm has its own detailed documentation:

• Sublinear Monotonicity Score: ./datamutant/monotonicity_score/README.md

🎯 Examples

Each algorithm includes comprehensive examples in its own directory:

# Run monotonicity score examples
python -m datamutant.monotonicity_score.examples.basic_usage
python -m datamutant.monotonicity_score.examples.model_training

🧪 Testing

Run all tests:

pytest

Run tests for specific algorithms:

pytest tests/monotonicity_score/

🤝 Contributing

1. Each new algorithm should be in its own module under datamutant/
2. Include comprehensive documentation in the algorithm's README.md
3. Provide examples in the examples/ subdirectory
4. Add tests in the corresponding tests/ directory
5. Update this main README with a short description

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.