A faster way to build quantitative features, targets, and alternative bars in Python. Powered by Numba.
Quantreo is a Python toolbox that helps you build quantitative features, targets, and alternative bars, faster and more efficiently.
It’s designed for quantitative researchers, data scientists, and traders who want to transform raw market data into ML-ready datasets in just a few lines of code. From volatility estimators to event-based bars and regime indicators, Quantreo provides a high-performance foundation for systematic trading research.
- High-performance: Built with Numba and vectorization for ultra-fast computations
- Research-oriented: Designed for feature and target engineering in quant finance
- Modular & Pythonic: Clean, intuitive API compatible with pandas & scikit-learn
- Alternative bars included: Time, tick, volume & run bars
- Robust by design: Functions implemented to avoid data leakage and look-ahead bias.
quantreo.features_engineering: Volatility, trend, entropy, and statistical indicatorsquantreo.target_engineering: triple barrier, event-based targets and regime classificationquantreo.alternative_bars: Time, tick, volume, and run barsquantreo.datasets: Synthetic or real OHLCV generators for fast testing
pip install quantreoIf the installation completes successfully, you're all set to start using Quantreo. If you encounter any issues, please refer to our documentation for troubleshooting tips.
Quantreo provides a collection of ready-to-use tools for feature engineering, target construction, and alternative bar generation, all optimized for performance and ease of use.
Jump to the Quick start section below to see how it works in practice.
Quantreo makes it easy to test, explore, and create new quantitative features and targets in just a few lines.
Use the built-in dataset generator to quickly get a realistic sample of market data.
Perfect for experimenting or creating your own custom functions.
from quantreo.datasets import load_generated_ohlcv
df = load_generated_ohlcv()
df.head()Compute a ready-to-use volatility feature in just two lines.
import quantreo.features_engineering as fe
df["parkinson_vol"] = fe.volatility.parkinson_volatility(df=df, high_col="high", low_col="low", window_size=30)Then, you can easily visualize your computed feature to check its behavior over time:
Want to go further? Here are a few examples of what you can build with Quantreo:
- Meta-Labeling Tutorial: Use Quantreo to generate dynamic targets and improve model precision.
- Multi-Asset Feature Engineering: Combine multiple assets to uncover cross-market patterns.
- PCA in Trading: Apply principal component analysis to understand market structure.
See all tutorials here → Quantreo for Beginners
- Explore our documentation to learn how Quantreo works.
- Ask questions or report issues directly on our GitHub repository.
- Discover practical examples and in-depth use cases in our tutorials section.
- Learn how to create features and targets like a pro in the ML4Trading course.
If you use Quantreo in your research or project, please consider citing it as:
Inglese, L. (2025). Quantreo: A Python library for quantitative feature and target engineering.
Available at: https://github.com/Quantreo/quantreo
MIT © Lucas Inglese, 2025
Feel free to reach out at lucas@quantreo.com.