How disciplined-pioneer/crypto-market-screeners works
Main Components
• Data Sources:
• External Market Data: The project uses live cryptocurrency data (e.g., via the Binance API) – see data_processing.py.
• Preprocessed Data Files: The project reads sentiment CSV files stored under data/Sentiment and volatility outputs.
• Data Processing:
• data_processing.py is responsible for retrieving and formatting raw market data including price, volume, delta, and open interest.
• Analysis / Screener Modules (each represented as its own component):
• ATR Screener (atr_screener.py): Calculates and outputs market volatility (written to data/ATR/volatility_results.json).
• Isolation Forest Screener (isol_forest_screener.py): Detects volume and price anomalies using the Isolation Forest algorithm.
• Market Ratio Screener (market_ratio.py): Computes various market metrics to detect supply and demand imbalances.
• Sentiment Screener (sentiment.py): Processes CSV files from data/Sentiment and computes market sentiment.
• Volume Spikes Screener (volume_spikes.py): Identifies significant trading volume spikes.
• Data Repository:
• The “data” directory holds the output/intermediate results from various screeners.
Relationships and Interactions
• Data flows from external sources (e.g., Binance API, CSV files) into the Data Processing component.
• The standardized or processed data is then fed as input into each screener module.
• Each analysis module (ATR, Isolation Forest, etc.) produces outputs that are saved back to the “data” directory or reported via logs/visualizations.
• There is an implied dependency where the data_processing module is the initial step, supplying data for most other analyses.
Architectural Patterns / Design Principles
• Modular Design: Each Python script represents a separate module focused on a specific analysis, promoting single responsibility.
• Pipeline Flow: There is a data ingestion and processing phase that feeds different analysis pipelines.
• Separation of Concerns: Data gathering, processing, and analysis are implemented as distinct components, allowing you to upgrade or adjust one part without affecting others.
Mention the Relevant Technologies and Frameworks
• Python 3.8+ and libraries mentioned in requirements.txt (e.g., scikit-learn for Isolation Forest, seaborn for visualization).
• File-based data persistence (JSON and CSV) and data processing using custom modules.