QuantifiLib

QuantifiLib is a research-oriented Python library for quantitative finance, developed by Quantifi Sogang.
It provides a modular and extensible framework for systematic trading research, including data loading, event-based labeling, signal generation, backtesting, portfolio optimization, time series modeling, causal inference, and synthetic data generation.

QuantifiLib is designed to serve as a full-stack infrastructure for academic finance research, integrating both traditional econometric models and modern machine learning techniques.

Whether you're building event-driven strategies, training machine learning models with purged cross-validation, or simulating asset prices with GANs, QuantifiLib offers a unified environment to streamline your research process.

📁 Project Structure

quantifilib/
├── data/
│   ├── data_loader/
│   │   ├── yfinance_loader.py
│   │   ├── fred_loader.py
│   │   └── naver_loader.py
│   └── stock_universe/
│       └── wikipedia.py
├── features/
│   ├── bar_sampling/
│   │   ├── bar_feature.py
│   │   ├── base_bars.py
│   │   ├── core.py
│   │   ├── imbalance_data_structures.py
│   │   ├── microstructure.py
│   │   ├── run_data_structures.py
│   │   ├── standard_data_structures.py
│   │   └── time_data_structures.py
├── metrics/
│   ├── liquidity/
│   │   ├── corwin_schultz.py
│   │   ├── lambda.py
│   │   ├── pin.py
│   │   └── roll_models.py
│   └── risk/
│   │   ├── market.py
│   │   └── strategy.py
├── strategy/   
│   ├── fundamental_based/
│   ├── ml_based/
│   │   └── suppport_vector_machine.py
│   ├── price_based/
│   │   ├── technical.py
│   │   └── triple_barrier.py
│   ├── statistical_based/
│   │   └── trend_search.py
│   └── base_label.py
└── utils/
    ├── multiprocess/
    │   ├── parts.py
    │   └── process_job.py
    └── fast_ewma.py