Crypto Pattern Recognition Engine

A Python library for detecting technical-analysis patterns in cryptocurrency OHLCV price data. Experimental / educational project.

What it does

• Technical indicators: RSI, MACD, Bollinger Bands, Stochastic, VWAP, moving-average cross (SMA/EMA), ATR, OBV, ADX, Parabolic SAR.
• Chart patterns: head & shoulders, triangles, double top/bottom, flags, pennants, wedges, cup & handle, rectangle, diamond.
• Candlestick patterns: doji, hammer/hanging man, engulfing, morning/evening star, three soldiers/crows, shooting star.
• Combination strategies (consensus / weighted / confirmation voting).
• A paper-trading simulator with position sizing and basic risk metrics (Sharpe, Sortino, VaR) and a simple portfolio rebalancer.

Quick start

git clone https://github.com/overkillkulture/crypto-pattern-recognition-engine.git
cd crypto-pattern-recognition-engine
python -m venv venv
source venv/bin/activate      # Windows: venv\Scripts\activate
pip install -r requirements.txt

python examples/demo_patterns_offline.py

from src.patterns.optimized import OptimizedRSIPattern

rsi = OptimizedRSIPattern(use_cache=True)
patterns = rsi.detect(your_ohlcv_data)
for p in patterns:
    print(p.pattern_name, p.signal, round(p.confidence, 2))

See examples/ for the trading simulator, multi-strategy backtest, and portfolio demos.

Status

Early-stage and not actively maintained. Indicators, chart/candlestick patterns, the paper-trading simulator, and the benchmark suite are implemented. Live data feeds, exchange connectors, and ML classification are not.

Disclaimer

For educational and research purposes only. This is not financial advice. Cryptocurrency trading carries significant risk; test thoroughly and do your own research before risking capital.

License

MIT License - see the LICENSE file.