v0.1.5 - Research Workflow Notebooks

Added

Research Workflow Notebooks

• Complete systematic research workflow notebooks included in PyPI distribution
• notebooks/01_data_download.ipynb - Bloomberg data download (21 cells)
  • Automated download for all configured securities (CDX, VIX, ETF)
  • BloombergSource integration with graceful error handling
  • Cache directory management and validation
  • Sample data generation for missing securities
  • Comprehensive logging and progress tracking
• notebooks/02_signal_computation.ipynb - Signal computation (20 cells)
  • Batch signal computation from cached Bloomberg data
  • SignalRegistry integration for catalog-driven signal generation
  • Comprehensive validation (z-score normalization, alignment, correlations)
  • Plotly visualizations (individual signals, 3-panel subplot, correlation heatmap)
  • Metadata persistence for reproducibility
  • Production-ready workflow with error handling
• notebooks/__init__.py - Package initialization with workflow documentation
  • Notebook sequence documentation
  • Workflow position markers for each step
  • Prerequisites and outputs clearly defined

Distribution Improvements

• Research notebooks now included in PyPI package (aponyx/notebooks/)
• Notebooks accessible after installation for copy/modification
• Updated packaging to include notebook JSON files

Examples Enhancements

• Signal suitability evaluation added to end_to_end_demo.ipynb
  • Pre-backtest quality gate demonstration
  • PASS/HOLD/FAIL decision workflow
  • Component scoring breakdown visualization

Changed

• Updated .gitignore to preserve notebooks/ directory structure
• Enhanced README.md with research notebooks section and workflow documentation
• Updated PROJECT_STATUS.md to reflect notebook distribution
• Copilot instructions updated with notebook workflow context
• pyproject.toml now includes notebook files in package data

Fixed

• Bloomberg demo error handling for missing Terminal connection
• Suitability report path references in demo notebook