v0.6.0

What's Changed

• Mark type aliases explicitly by @thomkeh in #475
• Try to turn datasets into dataclasses by @thomkeh in #474
• aif converter by @olliethomas in #476
• add acs employment dataset, and abstract some common elements of the acs datasets by @olliethomas in #477
• Make the new dataset dataclasses hydra-compatible by @thomkeh in #478
• Replace custom async code with ray by @thomkeh in #479
• Exclude imports from docs building by @thomkeh in #480
• Fix more docs problems by @thomkeh in #482
• Turn all datasets into dataclasses by @thomkeh in #483
• Disentangle Kamishima into fit()/predict() by @thomkeh in #484
• Fix problems that were uncovered by the slow tests by @thomkeh in #486
• Mark more tests as slow by @thomkeh in #487
• Define labels for confusion-matrix metrics by @olliethomas in #489
• Use fit()/predict() split for Zafar by @thomkeh in #490
• Replace count_true with np.count_nonzero by @thomkeh in #493
• Create CITATION.cff by @olliethomas in #495
• Add diagram to README.md by @thomkeh in #497
• Remove __init__ from algorithm base classes by @thomkeh in #500
• Add tests for the algorithms that now are dataclasses by @thomkeh in #501
• Create dependabot.yml by @thomkeh in #502
• Update black to non-beta version by @thomkeh in #508
• Turn metrics into dataclasses by @thomkeh in #518

Full Changelog: v0.5.1...v0.6.0

v0.5.0

What's Changed

Slight change to the API with a new feature. Separate fit and predict methods for the models.

• add separate fit and test functions by @olliethomas in #472

Full Changelog: v0.4.1...v0.5.0

v0.4.0

What's Changed

New Dataset - Added ACS Income dataset from Retiring Adult using the excellent folktables by @olliethomas in #469

Also, updated the dependencies.

Full Changelog: v0.3.5...v0.4.0

Release 0.3.5

What's Changed

• Delete the old docs subdir by @thomkeh in #466
• Wrap 3.10 in quotes by @thomkeh in #467
• add a flag to invert the s value by @olliethomas in #468

Full Changelog: v0.3.4...v0.3.5

Release 0.3.4

A couple of touch-ups. Updated dependencies; Made seed a parameter to all models; The docs now auto build on update.

Release 0.3.3

We now have the option to just refer to the S variable in the produced reulsts as just S instead of this changing per dataset. This is useful if you're looking at runs across multiple datasets and want the produced .csv to be easier to navigate.

Release 0.3.2

Minor update of dependencies and tabular datasets now also accepting enums.

Release 0.3.1

Added the LSAC LAw Admissions Dataset form Counterfactual Fairness by Kusner et al.

Release 0.3.0

it's about time we went to 0.3.
This is mostly just updates to dependencies and getting some parity across datasets (all datasets now have discrete feature groups in case you want more information about how a decoder should reconstruct them).

Also, this release was made with our lovely new release script - thanks @thomkeh 🙏 , so releasing smaller updates more often should be easier.

EthicML 0.2.7

Changes

• Add labels_as_featurse option to dataset.load()
• Expose constants that determine where the celeba dataset is stored