Candlestick Pattern Recognition
Purpose
Identifies 15 classic candlestick patterns and generates trading signals:
Single-Candle Patterns (5)
| Pattern |
Signal |
Description |
| Hammer |
Bullish |
Long lower shadow with a small body at the top |
| Inverted Hammer |
Bullish |
Long upper shadow with a small body at the bottom |
| Shooting Star |
Bearish |
Long upper shadow with a small body at the bottom (appears after an uptrend) |
| Doji |
Neutral |
Open and close are nearly equal |
| Spinning Top |
Neutral |
Small body with roughly equal upper and lower shadows |
Double-Candle Patterns (5)
| Pattern |
Signal |
| Bullish Engulfing |
Bullish |
| Bearish Engulfing |
Bearish |
| Bullish Harami |
Bullish |
| Bearish Harami |
Bearish |
| Piercing Line |
Bullish |
| Dark Cloud Cover |
Bearish |
Triple-Candle Patterns (4)
| Pattern |
Signal |
| Morning Star |
Bullish |
| Evening Star |
Bearish |
| Three White Soldiers |
Bullish |
| Three Black Crows |
Bearish |
Signal Logic
Bullish patterns score +1, bearish patterns score -1. Go long when the total score is > 0, go short when it is < 0, and stand aside when it equals 0.
Parameters
| Parameter |
Default |
Description |
| body_pct |
0.1 |
Threshold for body-to-range ratio in a doji |
| shadow_ratio |
2.0 |
Ratio of shadow length to body length |
Dependencies
pip install pandas numpy requests
Signal Convention
1 = long, -1 = short, 0 = stand aside