AI Elliott Wave Counter

An ultra-high-speed quantitative engine for automated Elliott Wave Theory decomposition, pattern recognition, and probabilistic scenario ranking.

Designed for quantitative analysts, algorithmic traders, hedge funds, and technical chartists who need rigorous, rule-based Elliott Wave analysis delivered with sub-second execution.

---

What It Does

The AI Elliott Wave Counter scans financial market candlestick data, identifies critical price swings, and systematically evaluates multi-degree fractal wave counts according to classical Elliott Wave rules and modern quantitative extensions.

• Multi-Degree Recursive Decomposition: Deconstructs complex price trends into hierarchical wave structures across degrees (Supercycle, Cycle, Primary, Intermediate, Minor, Minute, Minuette, Sub-Minuette).
• Comprehensive Pattern Recognition:
  • Motive Structures: 5-wave Impulses, Contracting Diagonals, and Expanding Diagonals.
  • Corrective Structures: Zig-Zags (5-3-5), Flats (3-3-5), Contracting & Expanding Triangles (3-3-3-3-3), and Complex Combinations (Double/Triple Zigzags, W-X-Y).
• Alternative Counts & Probabilistic Ranking: Instead of presenting a single subjective interpretation, the engine generates all mathematically valid alternative counts and ranks them using continuous Gaussian Fibonacci tolerances and machine learning conviction scores.
• Incomplete / Developing Wave Projections: Identifies developing waves in real time to project continuation targets, wave 3 extensions, and corrective terminal zones.
• Interactive Visualizer: Generates a self-contained interactive web chart that lets you visually inspect and step through every detected wave scenario directly in your browser.

---

Processing Capacity & Limits

[!IMPORTANT] Maximum Bar Limit: The analyzer accepts up to 2,000 Japanese candlesticks (bars) per execution.

To ensure ultra-fast processing and deterministic execution times, input datasets exceeding 2,000 bars must be trimmed to the most recent 2,000 bars.

---

Inputs

Configure the analysis using simple JSON parameters in the actor input:

Parameter	Type	Default	Description
candles	Array<Object>	Optional	Array of Japanese candlestick objects (open, high, low, close, time). Maximum 2,000 bars.
symbol	String	"BTCUSDT"	Optional: If direct candles are omitted, specify a crypto pair to auto-fetch the latest market candles.
timeframe	String	"1h"	Candlestick interval when auto-fetching (1m, 5m, 15m, 30m, 1h, 4h, 1d).
topCounts	Integer	50	Maximum number of top alternative wave counts to return (1 to 200). If fewer alternative counts are detected, all available counts are returned.
swingSize	Integer	1	Pivot detection sensitivity window. Smaller values (1–2) detect micro-swings; higher values (3+) isolate major macro pivots.
maxDepth	Integer	4	Maximum depth for recursive wave decomposition (e.g., Primary down to Sub-Minuette).
analysisMode	String	"Both"	Analysis scope: "Both" (Complete & Developing), "Complete" (only finished structures), or "Incomplete" (only active waves).
scoreFormula	String	"ml_calibrated"	Scoring model: "ml_calibrated" (Gaussian + Machine Learning probability), "average", or "bayesian".
allowTriangleOverlap	Boolean	false	Enables structural sub-wave overlap flexibility in triangle patterns.
renderInteractiveHtml	Boolean	true	Generates a standalone interactive chart stored in the Key-Value store under CHART_HTML.

Candlestick Format Example

You can pass standard OHLCV objects or compact arrays:

{
  "topCounts": 50,
  "swingSize": 1,
  "maxDepth": 4,
  "candles": [
    { "time": 1700000000, "open": 64200.5, "high": 64850.0, "low": 63900.0, "close": 64600.0, "volume": 1420.5 },
    { "time": 1700003600, "open": 64600.0, "high": 65300.0, "low": 64450.0, "close": 65150.0, "volume": 1890.2 },
    { "time": 1700007200, "open": 65150.0, "high": 65400.0, "low": 64700.0, "close": 64800.0, "volume": 1210.8 }
  ]
}

Note: Compact arrays [timestamp, open, high, low, close] or [open, high, low, close] are also supported.

---

Outputs

Every execution produces clean, structured outputs across both the Default Dataset and the Default Key-Value Store:

1. Default Dataset (Tabular Scenarios)

Each alternative wave scenario is stored as an individual record in the dataset, perfectly formatted for export to JSON, CSV, or Excel:

{
  "rank": 1,
  "pattern": "Impulse",
  "degree": "Primary",
  "direction": "bullish",
  "total_score": 14.85,
  "ml_probability_pct": 82.4,
  "start_bar_index": 12,
  "end_bar_index": 118,
  "start_price": 58200.0,
  "end_price": 68450.0,
  "source": "full",
  "summary": {
    "idx": 0,
    "pattern": "Impulse",
    "degree": "Primary",
    "score": 14.85,
    "direction": "bullish",
    "c_ratio_a": 0.0,
    "y_ratio_w": 0.0,
    "ml_prob": 82.4
  },
  "tree": {
    "pt": "Impulse",
    "role": "root",
    "deg": 2,
    "si": 0,
    "ei": 8,
    "sp": 58200.0,
    "ep": 68450.0,
    "dir": "bullish",
    "sc": 5.2,
    "ts": 14.85,
    "ch": [
      { "pt": "Impulse", "role": "1", "deg": 3, "sp": 58200.0, "ep": 61000.0, "dir": "bullish" },
      { "pt": "Zig-Zag", "role": "2", "deg": 3, "sp": 61000.0, "ep": 59300.0, "dir": "bearish" },
      { "pt": "Impulse", "role": "3", "deg": 3, "sp": 59300.0, "ep": 66200.0, "dir": "bullish" },
      { "pt": "Flat",    "role": "4", "deg": 3, "sp": 66200.0, "ep": 64100.0, "dir": "bearish" },
      { "pt": "Impulse", "role": "5", "deg": 3, "sp": 64100.0, "ep": 68450.0, "dir": "bullish" }
    ]
  }
}

2. Key-Value Store Records

• SUMMARY: A concise executive summary containing analysis performance statistics, total counts discovered, and top pattern metrics.
• OUTPUT: The complete hierarchical analysis payload, including full wave trees and index-to-bar mappings.
• CHART_HTML: An interactive, standalone visualization dashboard. Open the URL to inspect candlesticks, view degree color-coded wave labels, toggle between alternative counts, and examine sub-wave trees in real time.

---

Practical Applications

• Algorithmic Trading: Ingest structured wave counts directly into automated execution systems for high-probability trend entries and invalidation stop-loss placement.
• TradingView & Chart Integration: Feed detected wave labels and pivot coordinates into custom indicators or chart overlays via webhook.
• Consensus Analysis: Measure structural consensus across multiple timeframes to determine whether the market is in an impulsive trend or complex corrective chop.
• Backtesting: Rigorously validate Elliott Wave trading strategies against historical market data without subjective manual charting bias.