jgtml 0.0.330

JGT Ml - Machine Learning Tools

🐊 JGTML - Trading Signal Analysis Platform

A Python-based trading signal analysis system focused on fractal patterns, Alligator indicators, and multi-timeframe confluence detection.

🎯 Core Purpose

JGTML analyzes the effectiveness of trading signals within larger market structure contexts, providing tools for:

• Signal Validation: Analyze FDB (Fractal Divergent Bar) and Alligator-based signals
• Multi-Timeframe Analysis: Process signals across H1, H4, D1, W1, M1 timeframes
• Performance Metrics: Calculate win rates, profit/loss ratios, and signal quality
• Trade Lifecycle Management: From entry validation through exit strategies

🏗️ Architecture

Core Dependencies

• jgtpy: Market data acquisition wrapper from jgtfxcon and indicator calculations and signals generation.
• jgtutils: Common utilities and constants. Contains the logics for reading $HOME/.jgt/config.json and $HOME/.jgt/settings.json for all app/cli to have all the context they need. jgtutils also uses jgtcore which start become the library for settings and configurations (separating jgutils that has utilities from actual libraries which is in progress)
• pandas: Data manipulation and analysis
• numpy: Numerical computations

Key Components

📊 Signal Processing

• jgtml/SignalOrderingHelper.py: Signal validation and risk calculation
• jgtml/jtc.py: Target calculation and signal analysis core
• jgtml/TideAlligatorAnalysis.py: Unified Alligator analysis (Regular, Big, Tide) - Potential Usage not that obvious yet, it was unified by agents so we could probably use it into further code such as the FDBScan before we enter the market
• jgtml/alligator_cli.py: 🐊 NEW Unified Alligator CLI with graceful TTF pattern handling
• TODO add TTF (TTF != Time-To-Future but more like feature of multiple timeframe) probably ttfcli.py

🚀 Command Line Tools

• jgtml/jgtmlcli.py: Main CLI for data processing
• jgtml/mxcli.py: Matrix generation and analysis - (Probably the next CLI Wrapper of what jgtml/jgtmlcli.py was/is doing which is generating the mxdata which contains targets for the fdb signals (buy/sell) so we could do machine learning feature design etc). It will probably be integrated with fdb_signal_quality_predictor.py, fdb_pattern_intelligence.py, unified_discovery_dataset_generator.py,...
• jgtml/jgtapp.py: Trading operation management (includes legacy tide command wrapper) - IT was designed to contains all commands that jgtml offers along with what jgwill/jgtpy offers such as idscli and cdscli so we can access generating indicators and signals data into our logics. This is the app used to create entryOrder in the market, look an existing trade to trail the alligator's line for exit strategie, etc. TODO: Document way further and integrate the whole values developped with alligator_cli,
• jgtml/alligator_cli.py: 🐊 Unified Alligator Analysis CLI - Replaces ptojgtmltidealligator/ptojgtmlbigalligator

🧬 Memory & Persistence

• garden_one/trading_echo_lattice/: Signal crystallization and memory storage - That is just a prototype done by our agents that is not quite obvious yet, that would not be a priority to work with that except if really relevant.
• Integration with Upstash Redis for persistent analysis results

🚀 Quick Start for developer

Installation

# Install dependencies
pip install jgtpy jgtutils pandas numpy python-dateutil

# Install JGTML
pip install -e .

Basic Usage

# Process signals for an instrument
jgtmlcli -i SPX500 -t D1 --full --fresh

# Analyze signal performance  
python -m garden_one.trading_echo_lattice.cli process -i SPX500 -t D1 -d S

# Generate analysis matrix
mxcli -i EUR/USD -t H4 --fresh

📈 Trading Strategies

Five Dimensions + Triple Alligator Confluence

Multi-indicator alignment detection using:

1. Alligator Lines: Jaw, Teeth, Lips confluence
2. Fractal Signals: FDB breakout validation
3. Awesome Oscillator: Momentum confirmation
4. Multi-Timeframe: Higher TF bias confirmation
5. Volume Analysis: MFI integration

Implementation: TradingEchoLattice.detect_breakouts()

Green Dragon Breakout

FDB-based breakout detection with Alligator mouth validation.

Implementation: fdb_scanner_2408.py

🔧 CLI Reference

See CLI_HELP.md for complete command documentation.

Core Commands

# Data Processing
jgtmlcli -i INSTRUMENT -t TIMEFRAME [--full] [--fresh]
mxcli -i INSTRUMENT -t TIMEFRAME [--fresh]

# Unified Alligator Analysis ✨ NEW ✨
python -m jgtml.alligator_cli -i SPX500 -t D1 -d S --type tide    # Single Alligator
python -m jgtml.alligator_cli -i EUR/USD -t H4 -d B --type all    # Multi-Alligator convergence
python -m jgtml.alligator_cli -i GBPUSD -t D1 -d S --generate-spec # Generate .jgtml-spec

# Legacy Support (redirects to unified CLI)
jgtapp tide -i SPX500 -t D1 B  # Legacy wrapper → unified Alligator CLI

# Trading Operations  
jgtapp fxaddorder -i EUR/USD -n 0.1 -r 1.0950 -d B -x 1.0900
jgtapp fxmvstopgator -i EUR/USD -t H4 -tid TRADE_ID --lips

# Signal Analysis
python -m garden_one.trading_echo_lattice.cli process -i SPX500 -t D1,H4 -d S
python -m garden_one.trading_echo_lattice.cli search --min-win-rate 60

📊 Data Flow

Market Data (jgtpy) → Signal Processing (jtc) → Analysis (CLI tools) → Memory Lattice (Redis)

1. Data Acquisition: Pull OHLC data via jgtpy
2. Indicator Calculation: Generate Alligator, AO, Fractals, MFI
3. Signal Detection: Identify valid entry/exit signals
4. Performance Analysis: Calculate win rates and profitability
5. Memory Storage: Crystallize results in Redis for pattern recognition

🧪 Development

Running Tests

python -m pytest tests/

Contributing

1. Focus on signal accuracy and performance metrics
2. Maintain compatibility with jgtpy data structures
3. Document new indicators and validation logic
4. Test across multiple timeframes and instruments

🔄 Recursive Architecture

While JGTML operates as a practical trading platform, it embodies recursive principles:

• Memory Patterns: Each analysis builds upon previous signal history
• Multi-Scale Awareness: Signals are validated across multiple timeframes
• Adaptive Learning: Performance metrics inform future signal weighting

The system grows more intelligent through iteration, not just accumulation.

---

🧠 Technical Foundation: Precise signal analysis with mathematical rigor
🌸 Intuitive Interface: Clear CLI flows that make complex analysis accessible
🎵 Rhythmic Patterns: Market timing encoded in fractal mathematics

Built for traders who understand that the best signals emerge from the intersection of technical precision and pattern recognition.

🐊 Unified Alligator Analysis

Multi-Timeframe Convergence System ✨ NEW ✨

The unified Alligator CLI consolidates three powerful analysis frameworks into a single, graceful interface:

🔍 Regular Alligator (5-8-13 periods)

• Purpose: Quick market direction detection and entry signals
• Best For: Day trading, scalping, short-term momentum
• Signals: Immediate price action around Alligator mouth

🌊 Big Alligator (34-55-89 periods)

• Purpose: Intermediate cycle analysis and trend validation
• Best For: Swing trading, weekly positioning
• Signals: Higher timeframe context and cycle turns

🌀 Tide Alligator (144-233-377 periods)

• Purpose: Macro trend identification and major support/resistance
• Best For: Position trading, monthly strategic positioning
• Signals: Long-term trend direction and major reversals

Key Features

• 🔄 Graceful Pattern Handling: Automatically handles missing TTF patterns (zonesq, mfi, ttf)
• 🎯 Intent-Driven Analysis: Generates .jgtml-spec files for agentic integration
• 🌐 Self-Contained: No external bash script dependencies
• ⚡ Multi-Type Convergence: Analyze all three Alligator types simultaneously
• 🔧 Legacy Compatible: Seamless integration with existing jgtapp tide workflows

Usage Examples

# Single Alligator Analysis
python -m jgtml.alligator_cli -i SPX500 -t D1 -d S --type tide

# Multi-Alligator Convergence (recommended)
python -m jgtml.alligator_cli -i EUR/USD -t H4 -d B --type all

# Generate .jgtml-spec for agentic workflows
python -m jgtml.alligator_cli -i GBPUSD -t D1 -d S --type all --generate-spec

# Legacy support (automatically redirects to unified CLI)
jgtapp tide -i SPX500 -t D1 B

🔄 Migration from Legacy Commands

Important: The following legacy commands have been deprecated and replaced by the unified Alligator CLI:

Deprecated Commands ❌

• ptojgtmltidealligator → Use python -m jgtml.alligator_cli --type tide
• ptojgtmlbigalligator → Use python -m jgtml.alligator_cli --type big
• Bash function jgtml_ptojgtmltidealligator_by_instrument_tf_21 → Use unified CLI

Migration Benefits ✅

• Self-contained operation (no bash script dependencies)
• Graceful error handling (TTF pattern failures don't crash analysis)
• Multi-Alligator convergence analysis capability
• Enhanced .jgtml-spec generation for agentic workflows
• Backward compatibility (legacy jgtapp tide still works)

📡 Intent Capture API (Draft)

See docs/trading_intent_api.md for the proposed HTTP flow capturing narrated observations and generating .jgtml-spec files.