Skip to content

v0.1.6 - Complete Research Workflow with Synthetic Data Support

Choose a tag to compare

@stabilefrisur stabilefrisur released this 09 Nov 13:43
· 208 commits to master since this release

Added

Backtest Workflow Notebook

  • Complete backtest execution notebook (notebooks/04_backtest.ipynb) with 48 cells
    • Strategy registry integration for catalog-driven backtesting
    • Multi-strategy batch execution with error handling
    • Comprehensive performance metrics visualization
    • Signal-strategy alignment validation
    • Metadata persistence for reproducibility
    • Production-ready workflow completing the 4-step research cycle

Synthetic Data Support

  • Synthetic data generation utilities for development workflow (notebooks/generate_synthetic_data.py)
    • CDX, VIX, and ETF synthetic time series generation
    • Configurable market regimes and correlations
    • Documentation for synthetic data usage (notebooks/README_SYNTHETIC_DATA.md)
    • Enables full workflow execution without Bloomberg Terminal access

Changed

  • Standardized product_id naming to lowercase format across all modules
    • Registry schema updated to use lowercase identifiers (e.g., cdx_ig_5y)
    • Strategy catalog updated with lowercase product references
    • All notebooks and examples updated for consistency
  • Enhanced signal computation notebook with improved visualizations
    • Better subplot layouts and formatting
    • Clearer correlation heatmap presentation
  • Refactored signal computation notebook to use fetch functions instead of hardcoded cache paths
  • Updated documentation to reflect current codebase state and workflow completeness

Fixed

  • Added missing imports to signal computation notebook
  • Removed deprecated tenor field from DataRegistry schema
  • Improved suitability evaluation workflow notebook organization