ZVT is a modular quant framework designed for quantitative trading research and backtesting. It provides a unified API for data recording, querying, and strategy development, catering to researchers and traders who need to analyze market data and build predictive models. The framework aims to simplify the process of data acquisition, processing, and strategy execution.

How It Works

ZVT employs a data-centric approach, defining tradable entities (like stocks, ETFs) and their associated events (like price changes, financial reports). It uses a schema-based system for data management, allowing for unified record_data and query_data methods across different data providers. Strategies can be implemented in a "solo" (free-style) or "formal" (structured factor-based) manner, with results visualized through a Dash/Plotly UI or accessed via a REST API.

Quick Start & Requirements

• Install: python3 -m pip install -U zvt
• UI: Run zvt open http://127.0.0.1:8050/
• Server: Install uvicorn, run zvt_server
• Frontend: Deploy from https://github.com/zvtvz/zvt_ui
• Dependencies: Python 3.8+ recommended. Data providers like 'em' (Eastmoney) and 'sina' are used.

Highlighted Details

• Unified API for data recording and querying across multiple providers (e.g., 'em', 'sina').
• Supports various data types: stocks (China, US, HK), indices, ETFs, financial factors, and events.
• Offers both a standalone UI for backtesting and research, and a REST API for real-time interaction.
• Enables strategy development through "solo" (flexible) and "formal" (factor-based) modes.
• Includes machine learning integration for predictive modeling.

Maintenance & Community

• The project states it does not guarantee backward compatibility and may deprecate features as the author's thoughts evolve.
• Contact: WeChat (foolcage), Zhihu.

Licensing & Compatibility

• The README does not explicitly state a license.

Limitations & Caveats

• The project explicitly states no backward compatibility is guaranteed, implying potential breaking changes.
• Real-time quotes for the new UI rely on QMT data source, requiring contact with the author for access.
• The author's evolving thoughts may lead to changes in maintenance focus.