v1.0.5

Added

• Added comprehensive cross-validation support with multiple strategies:
  • K-Fold Cross-Validation
  • Stratified K-Fold Cross-Validation
  • Leave-One-Out Cross-Validation
• Added cross-validation assertions for multiple metrics:
  • Accuracy Score
  • Precision Score
  • Recall Score
  • F1 Score
  • ROC AUC Score
• Added get_cv_summary function for detailed cross-validation metrics
• Added parallel processing support for faster cross-validation
• Added comprehensive documentation for cross-validation features
• Added cross-validation examples in documentation

Changed

• Updated model evaluation to support cross-validation-based assertions
• Enhanced error handling for cross-validation operations
• Improved documentation structure to include cross-validation section

Fixed

• Fixed potential memory issues in large-scale cross-validation
• Fixed documentation formatting for cross-validation examples

v1.0.4

Added

• Added comprehensive documentation for all core modules and functions
• Added detailed API reference in the documentation
• Added more examples in the documentation for common use cases

Changed

• Improved error messages for better debugging experience
• Enhanced documentation with more detailed examples and explanations
• Updated contributing guidelines with more detailed instructions

Fixed

• Fixed documentation formatting issues
• Fixed typos and inconsistencies in documentation
• Fixed minor formatting issues in error messages

1.0.3

This is the first stable, public release of ml-assert! This release marks a significant milestone, moving the library from its initial concept to a robust, tested, and published package. The core focus of this update was a complete overhaul of the validation API to be more powerful, intuitive, and developer-friendly.

✨ New Features & Major Changes

• Fluent Schema Builder (ml_assert.schema): Introduced a powerful, chainable, and fluent API for defining complex data validation rules. The new schema() builder is the centerpiece of the library, allowing for clear and expressive assertions.
• New DataFrame Assertion (.satisfies()): The old .assert_schema() method has been replaced with .satisfies(), which integrates seamlessly with the new schema builder for a more readable and powerful validation experience.

⚙️ CI/CD & Publishing

• Automated PyPI Publishing: The package is now automatically published to the official Python Package Index (PyPI) upon tagging a new version, thanks to a new GitHub Actions workflow.
• Trusted Publishing: Configured modern and secure "Trusted Publishing" for all releases to PyPI.
• Robust CI Pipeline: The CI pipeline has been completely rebuilt to validate every pull request and push by installing dependencies, running the full test suite, and building the package.

Documentation

• Updated README.md: The main README has been updated to reflect the new fluent API and installation instructions.
• New "Ultimate Guide" Notebook: The example notebook has been rewritten into a comprehensive guide (examples/ultimate_guide.ipynb) that walks through all the new features.

🐛 Bug Fixes & Refactoring

• Resolved numerous import errors and dependency issues within the src and tests directories.
• Fixed incorrect metadata in pyproject.toml that was causing installation and build failures.
• Refactored the internal project structure for clarity and maintainability.