Fingerprint Pattern Recognition
- 实用工具
- 版本: 1.0
- 激活: 20
Fingerprint Pattern Recognition
Non‑trading utility for MetaTrader 5.
Load a “fingerprint” CSV template (OHLC), scan historical charts for similar price action, and export every match to a results CSV for later analysis and labeling. Includes a secondary formation detector (classic chart patterns) as a bonus feature.
What this utility does
Fingerprint Pattern Recognition helps you build datasets of recurring price action:
- Fingerprint mode (primary):
Compares recent closed candles against a user‑provided template CSV (OHLC). When a match is found, the EA marks it on the chart and writes a row to a results CSV, including forward bars after the match (for outcome analysis).
- Formation library (secondary / bonus):
Detects a small set of common formations and draws clean annotations (pivots/levels + breakout candle marker).
Note: This is a helper feature. It’s not a perfect “AI chartist”; it is designed for fast, transparent rule‑based detection.
Does not trade.No orders are opened or managed.
Works on any symbol and timeframe.
Uses closed candles (stable signals; no repainting on the current forming bar).
Typical workflow (recommended)
1. Capture a fingerprint template using the companion “Fingerprint Capture Script”.
- The script exports a CSV template into the terminal’s Common Files folder.
2. Attach Fingerprint Pattern Recognition to a chart.
3. Select:
- Mode: Fingerprint Recognition (Primary)
- Template CSV: your fingerprint file
- Run Tag: name your dataset run (symbol / timeframe / experiment name)
4. Let it run in Strategy Tester. Every match is written to a results CSV (Common Files).
5. Analyze / label the exported matches:
- In Excel/Python, or
- In Patternica Desktop App (recommended, link in seller´s profile ).
Files and locations
- Template CSV input: written to /read from Common Files
- Results CSV output: written to Common Files
- This design makes it easy to run multiple terminals or multiple symbols/timeframes and keep files in one place.
What gets exported (results CSV)
Each match writes one row containing:
- A compact OHLC summary of the matched window
- The match score and flags
- OHLC for N forward candles after the match (N = Forward bars)
This is specifically designed for post‑analysis (win-rate, distribution, clustering, labeling, etc.).
Important notes / disclaimer
- This utility does not provide trading signals or profitability promises.
- Pattern matching is probabilistic by nature; results depend on market regime, symbol liquidity, spread, and the fingerprint template quality.
- Use for research, dataset building, and visual study.
