Stock Pattern Screenerv0.3.3

Use the 7 built-in pattern detectors to scan A-share stocks daily. The system normalizes scores across patterns like Cup-with-Handle and VCP, then aggregates confidence to rank stocks. Ideal for individual investors or small quant teams automating trade setup identification.

Backtest pattern-based strategies across A-share, Hong Kong, and crypto markets by switching data providers (Eastmoney, JoinQuant, etc.). The pipeline from data collection to visualization supports historical performance analysis and parameter optimization.

Extend the factor computation pipeline with custom factors (e.g., MACD crossover, volume breakout) and combine with pattern signals. The modular architecture allows adding new transformers and accumulators, enabling tailored stock selection logic.

Leverage the semantic locks to enforce sell-before-buy order execution and next-bar trading, reducing look-ahead bias. Generate clean trading signals with required fields, suitable for integrating with broker APIs or paper trading systems.

Analyze historical stock data to identify chart patterns and understand their predictive power. The normalized scoring and cross-detector calibration provide a consistent framework for studying pattern reliability, useful for finance courses or personal learning.

Offer the pattern screener as a cloud-based API or web app. Charge monthly fees for access to real-time scanning, backtesting, and premium data sources like QMT.

Deploy the skill package on client infrastructure, customize pattern parameters, and integrate with existing trading systems. Provide training and maintenance contracts.

Aggregate pattern signals from the screener and sell them as a data feed to algorithmic traders or robo-advisors. Tiered pricing based on frequency and coverage (e.g., daily A-share vs. multi-market real-time).