v2.4.2

Added

• **kwargs for fit functions

Changed

• Changed the signature of the fit method in the DIPLS and GCTPLS classes to accept **kwargs.

Fixed

• N/A

Removed

• N/A

Breaking Changes

• N/A

v2.4.0 - (epsilon, delta) Differentially Private Partial Least Squares Regression (EDPLS)

Summary

This release introduces a new model class EDPLS for (epsilon, delta) differentially private partial least squares regression.

Added

• New model class EDPLS for (epsilon, delta) differentially private partial least squares regression.
• edpls() function added to the functions module
• calibrateAnalyticGaussianMechanism() function to estimate noise variance added to the utils module
• Tests for the new model class added
• Jupyter notebook examples for the new model class added
• Documentation for the new model class and function added

Changed

N/A

Fixed

N/A

Removed

N/A

Breaking Changes

N/A

Upgrade Notes

None

v2.3.0 - Model Selection

v2.3.0 - Model Selection

Summary

This release introduces new features for model selection, additional unit tests, and various improvements and bug fixes.

Added

• New feature for model selection using cross-validation by GridSearchCV from scikit-learn.
• Additional unit tests for new features.
• Documentation for the new model selection feature.

Changed

• Refactored code for better readability and maintainability.
• Updated dependencies to the latest versions.

Fixed

N/A

Removed

N/A

Breaking Changes

• Refactored fit() method signature in DIPLS and GCTPLS classes.

Upgrade Notes

• Ensure to update your code to accommodate the refactored fit() method signature.

v2.2.1 - Bug Fixes

v2.2.1 - Bug Fixes

Summary

This release fixes a bug in the DIPLS class

Fixed:

• Fixed a bug in the DIPLS class where the fit() function could not handle multiple target domains for extraction of nt (number of target samples).
• Tested correct behavior in the notebooks.

v2.2.0 - Scikit-learn Integration

v2.2.0 - Scikit-learn Integration

Summary

This release introduces scikit-learn integration of diPLSlib models

Added

• Unit tests for models and functions

Features

• DIPLS and GCTPLS classes now extend Scikit-learn's BaseEstimator and RegressorMixin classes with consistent fit() and predict() functions.

Changed:

N/A

Fixed:

N/A

Removed:

N/A

Bug Fixes

N/A

Breaking Changes

N/A

Upgrade Notes

None

v2.1.0 - Restructuring and Documentation

v2.1.0 - Restructuring and Documentation

Summary

This pull request introduces significant restructuring and documentation enhancements, The key changes and improvements are as follows:

Summary

This pull request introduces significant restructuring and documentation. The key changes and improvements are as follows:

Added

• Utils Submodule: Added a new utils submodule to outsource utility functions, improving code organization and maintainability.
• Documentation: Comprehensive documentation has been added for all modules, classes, and functions using Sphinx. This includes detailed docstrings, examples, and references.

Features

• diPLSlib.utils submodule with some helper functions

Changed:

N/A

Fixed:

N/A

Removed:

N/A

Bug Fixes

N/A

Breaking Changes

• gengaus(), hellipse() and rmse() functions have been migrated to diPLSlib.utils submodule

Upgrade Notes

• None

v2.0.0 - GCT-PLS

v2.0.0 - GCT-PLS

Summary

This release includes a major overhaul of the project architecture, introduces the new GCTPLS class for Calibration Transfer, and adds a demo notebook and data repository for GCT-PLS.

Features

• New 'GCTPLS' class for Calibration Transfer.
• Demo notebook for GCT-PLS.
• Data Repository for the demo notebook.
• Changelog added.

Bug Fixes

N/A

Breaking Changes

• Renaming of the 'Model' class to 'DIPLS'
• Renaming of the 'dipals.py' submodule from diPLSlib into 'models.py'

Upgrade Notes

• None

v1.0.2 - Official Release

Bug Fixes

• Installation guide updated
• Installation of diPLSlib package from PyPI tested

v1.0.0 - Initial Release

v1.0.0 - Initial Release

Summary

This is the initial release of diPLSlib, a Python Package for Domain Adaptation in Multivariate Regression.

Features

• 'Model' class with 'fit' and 'predict' methods.
• Support for domain adaptation scenarios with multiple domains.

Bug Fixes

N/A

Breaking Changes

N/A

Upgrade Notes

• None

Note: Future versions will introduce additional functions related calibration transfer and transferability assessment.