wPattern

A Python project for detecting W (double-bottom) stock price patterns using rule-based technical analysis, with optional machine learning experiments.

This repo is focused on research and experimentation, not live trading.

---

Overview

The goal of this project is to:

• Systematically define and detect W patterns in historical stock data
• Scan large universes (e.g. S&P 500, Nasdaq)
• Study pattern frequency and post-pattern outcomes
• Experiment with ML models on top of rule-based detection

---

Key Features

• Rule-based W pattern detection
• Batch scans across stock universes
• Multiple experiment phases (formation-only, outcome-labeled, stricter rules)
• CSV exports and chart visualizations
• Optional ML workflows (feature engineering, classification)

---

Repository Structure (Simplified)

wPattern/
├── rules1.py              # Core W-pattern rules
├── phase1_formation_only/
├── phase2_with_outcomes/
├── strict_w_patterns/
├── sp500_scans/
├── training_data/
├── yfML/
└── plots / csv exports

---

Data

• Historical OHLCV data pulled via yfinance
• Typical lookback window: ~60–120 trading days

---

Disclaimer

For educational and research purposes only. Not financial advice.

---

Author

Ivan Liu Economics & Data Science @ USC