群益 Ticker Backend

A lightweight ticker backend for listening to market data, enabling algo-trading, backtesting, and real-time data visualization. Currently implemented in pure Python with an intuitive GUI frontend.

---

🔍 Overview

• Purpose: Provide a reliable, low-latency engine for market data ingestion and processing.

• Use Cases:

  1. Live Trading: Connect to broker feeds for real-time algo execution.
  2. Backtesting: Replay tick data for strategy validation.
  3. Visualization: Plot order flow, footprints, and custom charts.

🚀 Features

• Modular Architecture: Separate core engine and GUI for flexibility.
• Real‑Time Ticking: Buffer and batch‐process ticks every 0.3s for efficiency.
• Footprint & OHLC Aggregation: Generate minute bars with volume at price, delta, and trade-side breakdown.
• Extensible Interface: Python API to dispatch custom handlers or integrate with external modules.

📐 Architecture

1. COM Callback Layer: OnNotifyTicksLONG receives raw ticks from broker.
2. Buffering: Ticks are appended to in‐memory queues keyed by symbol.
3. Batch Aggregator: Every N miliseconds, _agg_tick processes new ticks into Bar objects.
4. GUI Binding: PySide6 visualizes incoming bars and orderflow charts.
5. Redis: In-memory database system for better IPC latency

🔧 Roadmap

• Built‑in Algo Trading App: Develop a simple UI for strategy live monitoring. (optionally strategy configuration)
• WebGL\WebGPU Migration: Take advantage of browser GPU framework, cross-platform.(FastAPI, websocket)
• Rust Migration: Port hot‐path functions (OnNotifyTicksLONG, _agg_tick) to Rust for sub‐millisecond throughput while retaining Python bindings. simd, avx2
• Plugin System: Allow community‐driven extensions for custom indicators and data sources.

🤝 Contributing

Contributions, issues, and feature requests are welcome!