Skip to content

release_notes.md

Latest commit

 

History

History
76 lines (60 loc) · 3.18 KB

release_notes.md

File metadata and controls

76 lines (60 loc) · 3.18 KB

Release Notes

v2.0.1 (August 2025)

Summary

Production-ready release with comprehensive statistical validation and professional code organization. This version transforms TSCSMethods.jl from an experimental implementation into robust statistical software suitable for causal inference research.

Key Features

  • Statistical Validation: Complete validation suite with coverage (96%), placebo tests (6.87%), and noiseless recovery verification
  • Comprehensive Testing: 8,146 tests with 99.4% success rate across all subsystems
  • Professional Architecture: Clean modular design with 6 logical subsystems
  • Complete Documentation: Full API documentation, tutorials, methodology explanations, and validation reports
  • Bootstrap Inference: Robust weighted block-bootstrap for uncertainty quantification
  • Advanced Features: Calipers, stratification, refinement, auto-balancing

Breaking Changes

  • Julia Version: Minimum requirement updated to Julia 1.10+
  • API Standardization: Function signatures standardized across the package
  • Parameter Validation: Enhanced input validation with informative error messages

New Features

  • Automatic Balancing: autobalance() function for p-value optimization
  • Enhanced Matching: Improved distance calculation and ranking utilities
  • Stratified Models: Full support for stratified causal inference
  • Multiple Outcomes: Analyze several dependent variables simultaneously
  • Memory Optimization: Efficient algorithms for large time-series cross-sectional datasets

Improvements

  • Error Messages: More informative validation and error reporting
  • Performance: Optimized matching and estimation algorithms
  • Type Safety: Comprehensive type validation throughout the package
  • Documentation: Visual diagrams, interactive workflows, and comprehensive tutorials

Dependencies

  • R Dependencies Optional: Bayesian factor calculation can be disabled with dobayesfactor=false
  • Minimal External Dependencies: Core functionality works with Julia standard library and carefully selected packages

Migration Guide

  • Update Julia to version 1.10 or later
  • Use dobayesfactor=false in estimate!() to avoid R dependencies
  • Review function call syntax for any custom scripts (standardized parameter passing)

Validation Results

  • Coverage Test: 96% confidence interval coverage ✓
  • Placebo Test: 6.87% Type I error rate ✓
  • Noiseless Recovery: Exact ATT recovery in synthetic data ✓
  • Test Suite: 8,146 tests passing (99.4% success rate) ✓

Citation

If you use TSCSMethods.jl v2.0.1 in your research, please cite:

@misc{feltham_tscsmethods_2023,
  title={TSCSMethods.jl: Matching methods for causal inference with time-series cross-sectional data},
  author={Feltham, Eric Martin},
  year={2023},
  url={https://github.com/human-nature-lab/TSCSMethods.jl}
}

Contributors

  • Eric Martin Feltham (Primary Developer)

Previous Versions

v2.0.0

  • Initial production release
  • Basic statistical validation
  • Core matching methodology implementation

v1.x

  • Experimental development versions
  • Research prototype implementations
  • Limited validation and testing