LiuAlgoTrader is a comprehensive, multi-process framework designed for algorithmic trading, targeting developers and traders seeking to build, test, and deploy automated investment strategies. It simplifies the entire lifecycle from development and hyper-parameter optimization to predictive model training and portfolio management, supporting US Equities and Crypto.

How It Works

The framework employs a scalable, multi-process architecture to handle complex trading operations efficiently. It integrates with various data and trading APIs (Alpaca, Gemini, Polygon.io, Tradier) for real-time data and execution. A key advantage is its automated analysis of trading sessions, which can be leveraged for hyper-parameter tuning and training predictive ML models, offering a robust solution for both laptop and server deployments.

Quick Start & Requirements

• Install: pip install liualgotrader
• Configure: Run liu quickstart to configure environment variables and set up a Dockerized PostgreSQL database with test data.
• Prerequisites: Alpaca Markets account (paper or funded), optional Polygon.io subscription, Docker Engine, and Docker Compose.
• Resources: Official documentation and installation FAQ are available.

Highlighted Details

• Supports trading and back-testing for US Equities and Crypto.
• Provides a browser-based UI for back-testing and analysis.
• Includes a wide range of analytical tools like tear-sheets, gain/loss analysis, and anchored-VWAPs.
• Machine Learning features (LSTM, Attention/Transformer) are in development.

Maintenance & Community

The project acknowledges several contributors and provides an email for suggestions and feedback. Further details on evolution and design concepts are available in the design folder.

Licensing & Compatibility

The README does not explicitly state the license type. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

Machine Learning features are marked as "work in progress." The Tradier API integration is in Beta. The README does not detail specific performance benchmarks or system requirements beyond the need for Docker.