Skip to content

README.md

Latest commit

 

History

History
173 lines (125 loc) · 5.03 KB

README.md

File metadata and controls

173 lines (125 loc) · 5.03 KB
Raw

Crypto Oracle - Advanced Trading Analysis System

A comprehensive cryptocurrency trading analysis system integrating technical analysis, machine learning, sentiment analysis, on-chain metrics, and AI-driven strategy generation.

Overview

Crypto Oracle is a powerful trading analysis system designed to help traders make more informed decisions by combining multiple data sources and analytical approaches:

  • Technical Analysis: Comprehensive indicators including RSI, MACD, Bollinger Bands, and more
  • Machine Learning: LSTM neural networks for price forecasting and GARCH models for volatility prediction
  • Multi-Timeframe Analysis: Examining patterns across different timeframes (1h, 4h, 1d)
  • Sentiment Analysis: Social media, news, and market sentiment integration
  • On-Chain Metrics: Blockchain data analysis for supported cryptocurrencies
  • GPT-Powered Strategy Generation: AI-generated trading strategies based on comprehensive market data
  • Backtesting: Historical performance validation of trading strategies
  • Portfolio Management: Position tracking and risk management

Features

Market Analysis

  • Fetch and analyze crypto data from major exchanges
  • Compute 20+ technical indicators
  • Multi-timeframe analysis to identify trends across different time periods
  • Identify support/resistance levels and market phases
  • Market scanning to discover trading opportunities

Machine Learning Models

  • LSTM price prediction with confidence scoring
  • GARCH volatility modeling
  • Value at Risk (VaR) calculations
  • Automated model caching for improved performance

Advanced Data Integration

  • Social media sentiment from Twitter, Reddit, and more
  • News sentiment analysis and aggregation
  • Fear & Greed Index integration
  • On-chain metrics for major cryptocurrencies
  • Exchange inflow/outflow monitoring

Trading Strategy Generation

  • AI-driven strategy creation using GPT-4
  • Entry, exit, stop-loss, and take-profit recommendations
  • Position sizing suggestions based on risk parameters
  • Strategy confidence scoring

Backtesting and Validation

  • Historical performance testing
  • Strategy quality scoring
  • Risk/reward evaluation
  • Win rate, profit factor, and drawdown analysis

Portfolio Management

  • Position tracking and performance monitoring
  • Risk management
  • Portfolio performance statistics

Installation

Prerequisites

  • Python 3.8+
  • Required packages (install via pip): 
    pip install -r requirements.txt

Configuration

  1. Create a .env file with the following variables:

    # Exchange API keys
BINANCE_API_KEY=your_binance_api_key
BINANCE_API_SECRET=your_binance_api_secret

# OpenAI API key for GPT-4 integration
OPENAI_API_KEY=your_openai_api_key

# Optional: For sentiment and on-chain metrics
ENABLE_SENTIMENT=true
ENABLE_ONCHAIN=true
TWITTER_BEARER_TOKEN=your_twitter_token
GLASSNODE_API_KEY=your_glassnode_key
CRYPTOQUANT_API_KEY=your_cryptoquant_key

  2. Customize settings in the configuration section of the main script if needed.

Usage

API Server

Run the API server to access all functionality via REST endpoints:

python crypto_oracle.py api --port 8000

Command Line Interface

Analyze a specific cryptocurrency:

python crypto_oracle.py analyze BTC/USDT --timeframes 1h 4h 1d

Scan the market for opportunities:

python crypto_oracle.py scan --max-symbols 30 --min-confidence 0.7

API Endpoints

Endpoint Method Description
/ GET API information
/tickers GET Available trading pairs
/timeframes GET Available timeframes
/analyze POST Full analysis for a symbol
/scan POST Market scan for opportunities
/sentiment/{symbol} GET Sentiment data for a symbol
/onchain/{symbol} GET On-chain metrics for a symbol
/backtest POST Backtest a strategy
/status GET System status

Example

Analyzing Bitcoin:

import requests
import json

response = requests.post('http://localhost:8000/analyze', json={
    'symbol': 'BTC/USDT',
    'timeframes': ['1h', '4h', '1d'],
    'lookback': 1000,
    'include_sentiment': True,
    'include_onchain': True
})

result = response.json()
print(json.dumps(result, indent=2))

# Access the strategy recommendation
strategy = result['data']['strategy']
print(f"Trade direction: {strategy['trade_direction']}")
print(f"Entry price: {strategy['optimal_entry']}")
print(f"Confidence score: {strategy['confidence_score']}")

Dependencies

  • CCXT: Exchange connectivity
  • TensorFlow: Machine learning models
  • FastAPI: API server
  • Pandas/NumPy: Data processing
  • TA-Lib (optional): Technical indicators
  • OpenAI: GPT strategy generation

Disclaimer

This software is for educational and research purposes only. It is not financial advice, and you should not make trading or investment decisions based solely on its output. Always do your own research and consult with a qualified financial advisor before trading.

License

MIT License