Adaptivebridge

Latest version: v1.1.0

Safety actively analyzes 685507 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

1.0.1

- **1.1.0** [290124] (Netanel Eliav):
1. **Bug Fix**:
- Fixed typos and addressed an issue where build.sh wasn't generating requirements.txt when the file exists.
2. **Performance**:
- Improved model mapping (part of the .fit method) by restructuring some methods, resulting in fewer iterations for the same tasks.
3. **Feature**:
- Added a new method for adaptive mode, the new Mutually Exclusive Estimation (_mutually_exclusive_detection) for boolean and one-hot encoded features. It will now handle classification tasks better by checking mutually exclusive features that are always the same or always not the same. Please refer to the [documentation](https://inetanel.github.io/adaptivebridge/techniques-and-algorithms.html).
- Added a progress bar using tqdm lib for .fit method. It now shows progress within AdaptiveBridge for the following processes:
- Determining Feature Distribution
- Calculating Feature Importance
- Estimating Mutually Exclusive
- Mapping Models and Features
- Creating Feature Dependencies
> **NOTE**: This version is the first production release version. It contains unit tests, bug fixes, and a complete set of features to handle most ML models.

1.0.1b

- **1.0.1 - Beta** [190123] (Netanel Eliav):
1. Bug Fix: Fixed benchmark not showing all features.
2. Enhancement: Added colors to graphs; now mandatory features are in blue, and the rest are in green.

1.0.0b

**1. Feature:**
- Add uni-tests and L1 (E2E) tests meeting production ready, reaching 98% code-coverage.
- _> Note: this version is initial pre-production beta release version, please make sure to check the documentation._

**2. Enhancement:**
- Add _mode_selector method for district and continuous distributions. Now it will handled better multi-mode edge cases.
- Add _DistributionType class to manage better all type of data distributions.
- Add comments for everything for better code review.

**3. Remove:**
- Remove all unique central tendency for each data distribution. now it will have a new generic algorithm for all type of distribution.

**4. Bug Fix:**
- Fix __str__ to show Accuracy Logic name and not object name.
- Fix predict method raise an error in edge cases, now it will execute feature engineering mis-match check first before executing bridge method.
- Fix feature_importance_score method not executable, now it will required dataset as parameter.

**5. Change:**
- Change author email to be updated one.

0.9.1

1. Bug Fix - Fix Images and Documentation. nbow it will show the logos.
2. Enhancement - Improvement to the PyPi package builder (build.sh). now it will take the version fro the CHANGELOG.md dynamicly.
3. Enhancement - Added the changelog to the PyPi package page.

0.9.0

>Note: this version is initial official alpha version (3-Alpha).
1. Alpha version.
- Change - PEP 8 for all files.
- Feature - support for feature engineering.
- Remove - Pyramid benchmark was remove due to low value and accuracy.
- Enhancement - Add documentation.

Links

Releases

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.