Interpret-community

Latest version: v0.31.0

Safety actively analyzes 688126 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 5 of 12

0.15.4

- Fix number of features for raw explanations when using LinearExplainableModel
- Fix explanation dashboard json serialization issue

0.15.3

- fix issue with serializing the lightgbm surrogate model on mimic explainer for lightgbm >= 3.0.0

0.15.2

- add eval_data to get_raw_explanation
- remove use_cdn from explanation dashboard
- fix credentialed VM response data when using axios library
- reuse version in sphinx build

0.15.1

- fix tolist error for sparse feature importances and sparse raw_to_output_map on creating raw explanation
- move LIME from azureml-contrib-interpret package to interpret-community
- fix error during visualization on explanation with id for eval_y_predicted and eval_y_predicted_proba in datasets mixin
- add true_y parameter to ExplanationDashboard notebooks to demonstrate additional functionality

0.15.0

- remove visualization directory as it is being moved to a common respository
- update tests to support lightgbm 3.0 update
- optimize data mapper memory usage
- update CDN path for widget
- add mimic explainer test for high-dimensional dense raw features explanation
- fix lightgbm 3.0 support by densifying input matrix to reenable tests
- fix issues with importing version in setup.py before dependencies installed

0.14.4

- fix open in new tab on public VMs (specifically tested on Azure DSVM)
- handle case where classifier has predict function that returns probabilities in model wrapper
- added tests for shap_values_output for tree explainer and mimic explainer, fixed issue with XGBoost and TreeExplainer case

Page 5 of 12

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.