For users:

pip install famlafl

Or with poetry:

poetry add famlafl

famlafl/
├── backtest_statistics/    # Backtesting tools and statistics
├── bet_sizing/            # Position sizing and bet sizing tools
├── clustering/            # Clustering algorithms for financial data
├── codependence/          # Codependence and correlation metrics
├── cross_validation/      # Cross-validation for financial data
├── data_structures/       # Financial data structures
├── datasets/              # Sample datasets and loaders
├── ensemble/              # Ensemble methods
├── feature_importance/    # Feature importance analysis
├── features/             # Feature engineering tools
├── filters/              # Financial data filters
├── labeling/             # Financial data labeling tools
├── microstructural_features/  # Market microstructure features
├── multi_product/        # Multi-product analysis
├── online_portfolio_selection/  # Online portfolio selection
├── portfolio_optimization/  # Portfolio optimization tools
├── sample_weights/       # Sample weight generation
├── sampling/             # Financial data sampling
├── structural_breaks/    # Structural break detection
└── tests/               # Unit tests

# Run all tests
poetry run pytest

# Run tests with coverage
poetry run pytest --cov=famlafl --cov-report=html --cov-report=term

# Run specific test file
poetry run pytest famlafl/tests/test_specific_file.py

We welcome contributions from the community! Please see our Contributing Guidelines for more details on how to get involved.

This is a fork of mlfinlab (ArbitrageLab), developed by Hudson & Thames Quantitative Research.

Important

• All mlfinlab-derived code here remains under Hudson & Thames’s “all rights reserved” license.
• Any new or original code that I (Vadim Surin) wrote from scratch (and does not derive from mlfinlab code) is released under the BSD-3-Clause License. However, usage in combination with mlfinlab code is still governed by Hudson & Thames’s restrictions.

1. Hudson & Thames License (All Rights Reserved)
   The original mlfinlab portion of this repository is subject to the Hudson & Thames license (or see the license text included in this repo’s LICENSE file).

   Their license overrides any open-source terms with respect to the mlfinlab files.

2. BSD-3-Clause (for My Independent Code)
   Purely original files that do not include or derive from mlfinlab logic can be used under BSD-3-Clause terms.

   Note: If these files are used in conjunction with mlfinlab code, the combined work is effectively subject to Hudson & Thames’s license to the extent of mlfinlab’s portion.

Feel free to experiment with my additions, but remember mlfinlab’s license requires you to comply with Hudson & Thames’s terms for the original (and derived) code.