v1.1.0

Changelog

[1.1.0] - 2024-07-23

---

Added

• New Method: rfe_brute_force
  • Implemented the rfe_brute_force method for exhaustive feature elimination.
  • This method iteratively adds features based on Recursive Feature Elimination (RFE) and evaluates model performance to find the best feature subset.
  • Example usage:

    best_features = fs.rfe_brute_force(estimator=RandomForestClassifier(), n_features_to_select=5, force=True)
    print("Best Features:", best_features)

Changed

• Backward Elimination and Forward Selection Enhancements
  • Added a fast parameter to both backward_elimination and forward_selection methods.
  • When fast=True, these methods use a quicker evaluation process, reducing computational time.
  • Example usage:

    selected_features = fs.backward_elimination(significance_level=0.05, fast=True)
    print("Selected Features (Fast Backward Elimination):", selected_features)

  • Example usage for forward selection:

    selected_features = fs.forward_selection(significance_level=0.05, fast=True)
    print("Selected Features (Fast Forward Selection):", selected_features)

Fixed

• Filter Method Improvements
  • Improved the filter_method to raise appropriate errors when using chi2 for regression problems or anova for classification problems.
  • Ensured better handling of data types and target variable compatibility.

v1.0.0

v1.0.0

Overview

Feature Selection Toolkit v1.0.0 adds a number of improvements and new functions, providing a more powerful and flexible toolkit for feature selection in machine learning projects. This release comes with new methods, improved documentation and bug fixes.

New Features

1. Support Vector Machines (SVC, SVR): Support Vector Machines have been added for both classification (SVC) and regression (SVR). This extends the range of models that can be used for feature selection.
2. XGBoost Models: Integrated XGBClassifier and XGBRegressor for powerful gradient boosting based feature selection.

Improvements

1. Improved Documentation: Detailed documentation and use cases have been added for all methods, making it easier for users to understand and apply feature selection techniques.
2. Alert Management: Added alert suppression to improve user experience during operation.
3. API Consistency: Ensured consistency for parameter naming and method signatures across all methods, making API usage more intuitive.
4. Code Improvements: Performance improvements and code edits have been made, resulting in a faster and more efficient working environment.

Bug Fixes

• Warnings: Fixed some warnings in the previous version.
• Model Compatibility: Fixed issues with some models not working with certain data types.

Installation and Use

pip install feature-selection-toolkit

GitHub

You can visit our GitHub page for more information and source code:
Feature Selection Toolkit GitHub

Feedback

We welcome your feedback about our library! You can share your contributions and suggestions with us to build a better and more efficient toolkit.

v0.4.2

v0.4.2

Overview

Feature Selection Toolkit v0.4.2 adds a number of improvements and new functions, providing a more powerful and flexible toolkit for feature selection in machine learning projects. This release comes with new methods, improved documentation and bug fixes.

New Features

1. Support Vector Machines (SVC, SVR): Support Vector Machines have been added for both classification (SVC) and regression (SVR). This extends the range of models that can be used for feature selection.
2. XGBoost Models: Integrated XGBClassifier and XGBRegressor for powerful gradient boosting based feature selection.

Improvements

1. Improved Documentation: Detailed documentation and use cases have been added for all methods, making it easier for users to understand and apply feature selection techniques.
2. Alert Management: Added alert suppression to improve user experience during operation.
3. API Consistency: Ensured consistency for parameter naming and method signatures across all methods, making API usage more intuitive.
4. Code Improvements: Performance improvements and code edits have been made, resulting in a faster and more efficient working environment.

Bug Fixes

• Warnings: Fixed some warnings in the previous version.
• Model Compatibility: Fixed issues with some models not working with certain data types.

Installation and Use

pip install feature-selection-toolkit

GitHub

You can visit our GitHub page for more information and source code:
Feature Selection Toolkit GitHub

Feedback

We welcome your feedback about our library! You can share your contributions and suggestions with us to build a better and more efficient toolkit.

v0.4.1

v0.4.1

Overview

Feature Selection Toolkit v0.4.1 adds a number of improvements and new functions, providing a more powerful and flexible toolkit for feature selection in machine learning projects. This release comes with new methods, improved documentation and bug fixes.

New Features

1. Support Vector Machines (SVC, SVR): Support Vector Machines have been added for both classification (SVC) and regression (SVR). This extends the range of models that can be used for feature selection.
2. XGBoost Models: Integrated XGBClassifier and XGBRegressor for powerful gradient boosting based feature selection.

Improvements

1. Improved Documentation: Detailed documentation and use cases have been added for all methods, making it easier for users to understand and apply feature selection techniques.
2. Alert Management: Added alert suppression to improve user experience during operation.
3. API Consistency: Ensured consistency for parameter naming and method signatures across all methods, making API usage more intuitive.
4. Code Improvements: Performance improvements and code edits have been made, resulting in a faster and more efficient working environment.

Bug Fixes

• Warnings: Fixed some warnings in the previous version.
• Model Compatibility: Fixed issues with some models not working with certain data types.

Installation and Use

pip install feature-selection-toolkit

GitHub

You can visit our GitHub page for more information and source code:
Feature Selection Toolkit GitHub

Feedback

We welcome your feedback about our library! You can share your contributions and suggestions with us to build a better and more efficient toolkit.

v0.3.1.

v0.3.1

Overview

Feature Selection Toolkit v0.3.1 adds a number of improvements and new functions, providing a more powerful and flexible toolkit for feature selection in machine learning projects. This release comes with new methods, improved documentation and bug fixes.

New Features

1. Support Vector Machines (SVC, SVR): Support Vector Machines have been added for both classification (SVC) and regression (SVR). This extends the range of models that can be used for feature selection.
2. XGBoost Models: Integrated XGBClassifier and XGBRegressor for powerful gradient boosting based feature selection.

Improvements

1. Improved Documentation: Detailed documentation and use cases have been added for all methods, making it easier for users to understand and apply feature selection techniques.
2. Alert Management: Added alert suppression to improve user experience during operation.
3. API Consistency: Ensured consistency for parameter naming and method signatures across all methods, making API usage more intuitive.
4. Code Improvements: Performance improvements and code edits have been made, resulting in a faster and more efficient working environment.

Bug Fixes

• Warnings: Fixed some warnings in the previous version.
• Model Compatibility: Fixed issues with some models not working with certain data types.

Installation and Use

pip install feature-selection-toolkit

GitHub

You can visit our GitHub page for more information and source code:
Feature Selection Toolkit GitHub

Feedback

We welcome your feedback about our library! You can share your contributions and suggestions with us to build a better and more efficient toolkit.

v0.3.0.

v0.3.0

Overview

Feature Selection Toolkit v0.3.0 adds a number of improvements and new functions, providing a more powerful and flexible toolkit for feature selection in machine learning projects. This release comes with new methods, improved documentation and bug fixes.

New Features

1. Support Vector Machines (SVC, SVR): Support Vector Machines have been added for both classification (SVC) and regression (SVR). This extends the range of models that can be used for feature selection.
2. XGBoost Models: Integrated XGBClassifier and XGBRegressor for powerful gradient boosting based feature selection.

Improvements

1. Improved Documentation: Detailed documentation and use cases have been added for all methods, making it easier for users to understand and apply feature selection techniques.
2. Alert Management: Added alert suppression to improve user experience during operation.
3. API Consistency: Ensured consistency for parameter naming and method signatures across all methods, making API usage more intuitive.
4. Code Improvements: Performance improvements and code edits have been made, resulting in a faster and more efficient working environment.

Bug Fixes

• Warnings: Fixed some warnings in the previous version.
• Model Compatibility: Fixed issues with some models not working with certain data types.

Installation and Use

pip install feature-selection-toolkit

GitHub

You can visit our GitHub page for more information and source code:
Feature Selection Toolkit GitHub

Feedback

We welcome your feedback about our library! You can share your contributions and suggestions with us to build a better and more efficient toolkit.