release v0.31.0

Release notes:

• Add flake8-all-not-strings to linting by @gaugup in #576
• fix interpret-community sphinx docs build failures by specifying html_theme by @imatiach-msft in #578
• update interpret-community to shap 0.42.1 and ml-wrappers 0.5.1 by @imatiach-msft in #579
• fix failing xgboost tests in nightly build by upgrading to latest shap, which is getting pinned by econml->responsibleai to older version by @imatiach-msft in #580
• update interpret to 0.4.4, ml-wrappers to 0.5.2 and enable fixed pfi test by @imatiach-msft in #582
• Update requirements-dev.txt to pin tensorflow below 2.14.0 by @gaugup in #583
• Update README.md to add supported python versions by @gaugup in #584
• update interpret-community to shap 0.43.0 by @imatiach-msft in #586
• fix gated build failures in interpret-community due to captum installation issues from conda by @imatiach-msft in #589
• update interpret-community to shap 0.44.0 by @imatiach-msft in #588
• Update python-linting.yml to upgrade setup python from v4 to v5 by @gaugup in #587
• update interpret to 0.5.0, ml-wrappers to 0.5.4 by @imatiach-msft in #590
• release interpret-community v0.31.0 by @imatiach-msft in #591
• update the pypi release script to use github action with auth token and update github action versions by @imatiach-msft in #593

Full Changelog: v0.30.0...v0.31.0

release v0.30.0

Release notes:

• Update python-linting.yml to use python 3.11 by @gaugup in #563
• Update codecov-action@v2 to codecov-action@v3 by @gaugup in #564
• fix sklearn build error due to old dependency version installed with newer scipy by @imatiach-msft in #568
• clean up load method API doc for mimic explainer to reduce confusion by @imatiach-msft in #567
• fix gated builds due to getargspec deprecation by @imatiach-msft in #570
• fix auto-generated API sphinx docs build for interpret-community by @imatiach-msft in #569
• Update explanation_utils.py to fix flake8 error by @gaugup in #571
• Create requirements-linting.txt to separate linting dependencies by @gaugup in #572
• update interpret-community to ml-wrappers 0.5.0 and interpret-core 0.4.3 by @imatiach-msft in #573
• backcompat imports to work with older versions of interpret-core by @imatiach-msft in #574
• fix flaky codecov errors in interpret-community by adding retry logic for uploads by @imatiach-msft in #575
• release interpret-community v0.30.0 by @imatiach-msft in #577

Full Changelog: v0.29.0...v0.30.0

release v0.29.0

Release Notes:

• fix flake8 build error due to missing stacklevel argument for warn method by @imatiach-msft in #554
• update interpret-community to ml-wrappers 0.4.3 by @imatiach-msft in #553
• drop python 3.6 support and fix build failures by @imatiach-msft in #557
• update interpret-community to interpret-core v0.3.2 by @imatiach-msft in #556
• Update setup.py to pin pandas below 2.0.0 by @gaugup in #558
• Update setup.py to dependency on raiutils by @gaugup in #555
• Delete requirements.txt by @gaugup in #559
• Remove deprecated iteritems() and use items() instead by @gaugup in #560
• release interpret-community v0.29.0 by @imatiach-msft in #561

Full Changelog: v0.28.0...v0.29.0

release v0.28.0

Release notes:

• Add scheduled builds by @gaugup in #542
• update interpret-community to ml-wrappers 0.3.0 by @imatiach-msft in #543
• Update setup.py to update ml-wrappers to 0.4.0 by @gaugup in #546
• Update python-linting.yml to update setup-python to version 4 by @gaugup in #547
• update interpret-community to ml-wrappers 0.4.2 by @imatiach-msft in #548
• release interpret-community v0.28.0 by @imatiach-msft in #549
• temporary fix for nightly build in interpret-community by @imatiach-msft in #551
• update interpret-community pypi release pipeline to use python 3.7 by @imatiach-msft in #552

Full Changelog: v0.27.0...v0.28.0

release v0.27.0

Release notes:

• add tests for visualize and data methods on explanation by @imatiach-msft in #530
• upgrade interpret-community to new shap 0.41.0 release by @imatiach-msft in #532
• fix nightly pytorch error on macos by installing numpy from conda by @imatiach-msft in #535
• fix gated build failures due to incompatibility with new flake8 and old flake8-bugbear and old flake8-logging-format by @imatiach-msft in #538
• update interpret-community to ml-wrappers 0.2.1 by @imatiach-msft in #536
• fix explanation adapter error when include_local=False and add tests by @imatiach-msft in #529
• fix warnings caused by deprecated parameter in linear explainer by @imatiach-msft in #526
• fix nightly build failure on windows when installing pip by @imatiach-msft in #539
• update interpret-community to ml-wrappers 0.2.2 by @imatiach-msft in #540
• release interpret-community v0.27.0 by @imatiach-msft in #541

Full Changelog: v0.26.0...v0.27.0

release v0.26.0

Release notes:

• Update github action for checkout, upload-artifact and setup-python from version 2 to 3 by @gaugup in #515
• Add flake8-pytest-style to linting by @gaugup in #512
• Improve the documentation website for interpret-community and update the main readme by @imatiach-msft in #518
• fix aggregating sparse multiclass explanations to global in global-only streaming with batches mode by @imatiach-msft in #519
• improve readme by only leaving high-level overview and move more sections to the website with more details by @imatiach-msft in #520
• add more documentation to interpret-community website, including a contributing guide by @imatiach-msft in #522
• fix gated build due to protobuf dependency error by @imatiach-msft in #524
• rename master branch to main in interpret-community by @imatiach-msft in #525
• fix zero importances for very small data when using mimic explainer with lightgbm surrogate model by @imatiach-msft in #523
• update interpret-community to ml-wrappers 0.2.0 by @imatiach-msft in #527

Full Changelog: v0.25.0...v0.26.0

release v0.25.0

Release notes:

• Make function _get_surrogate_model_replication_measure() public
• Add linting workflow for python files
• decrease tolerance on serialization check for macos test builds
• Add python 3.7 tests
• replace numpy.array with numpy.ndarray in docs
• Update requirements-test.txt to pin markupsafe<2.1.0
• fix build failure due to old sqlalchemy dependency
• Replace load_boston() with fetch_california_housing()
• Add notebook replication measure of global surrogate model
• fix windows nightly build failures due to pywinpty dependency failing
• fix some methods and classes missing from readthedocs documentation
• update interpret-community to ml-wrappers 0.1.0
• fix PFI handling tensorflow models for tensorflow>=2.6
• fix shape check to support other types of wrapped datasets such as BatchDataset

release v0.24.2

• move PR gate to github actions
• correctly annotate python warnings
• remove devops unit tests from PR gate
• add backcompat test for saving and loading explanations and fix backcompat issue
• make function _get_surrogate_model_replication_measure() public
• add linting workflow for python files
• decrease tolerance on serialization check for macos test builds

release v0.24.1

• fix reimports from ml-wrappers package for WrappedClassificationModel, WrappedRegressionModel

release v0.24.0

• update interpret-community to use new ml-wrappers repository (https://github.com/microsoft/ml-wrappers)
• add tests for ExplanationAdapter for code coverage
• rename test directory from test/ to tests/
• suppress tensorflow warnings that may sometimes occur on import
• add support for serializing explanations using save_explanation and load_explanation with sparse data
• update interpret-community to ml-wrappers 0.0.5
• fix flake8 error on builds