Interpret-community

Latest version: v0.31.0

Safety actively analyzes 678576 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 7 of 12

0.11.2

Not secure
* Cohort functionality added to the dashboard
* Improvement in PFI behavior when passing true_labels as a kwarg
* Improve type handling in the list sorting util

0.11.1

Not secure
* add support for passing in a theme object to the dashboard
* fixes for the latest version of scikit-learn
* improve warning/error messaging in the dashboard
* optimize `get_ranked_per_class_names()` memory usage
* improve PFI error messaging
* improvements to documentation
* workaround for XGBoost and TreeExplainer

0.11.0

Not secure
* v2 version of the dashboard has been added!
* theme improvement for sphinx API documentation
* improvement for port handling within the dashboard
* improvement to scipy sparse import syntax

0.10.2

Not secure
* Dashboard auto focus update
* Bugfix for sparse data on when using MimicExplainer with a linear surrogate for binary classification
* Set up port forwarding so that the dashboard can be shown in Azure environments
* Bugfix for true ys that aren't lists

0.10.1

Not secure
* Bugfix for dashboard, should now work on remote compute
* Bugfix for raw explanation generation with feature maps - engineered data should no longer be attached
* Documentation fixes

0.10.0

Not secure
- updated interpret-core to 0.1.21
- import ABC from collections.abc instead of collections for Python 3 compatibility
- fixed timeseries error where index column not removed for surrogate model when using reset_teacher on evaluation

Page 7 of 12

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.