Nannyml

Fixed

- Fixed incorrect default thresholds in the docstrings for the univariate drift calculator. Thanks for the eagle-eyed reading [josecaloca](https://github.com/josecaloca)! [(#425)]((https://github.com/NannyML/nannyml/issues/425))

Changed

- Thorough revamp of our dependency version specifications. Dependencies are now less strict, making it easier to use NannyML as a dependency. Big credits to [davisthomas-aily](https://github.com/davisthomas-aily) and [canoadri](https://github.com/canoadri) for their contributions, thoughts and patience on this one. Much appreciated! [(#433)]((https://github.com/NannyML/nannyml/issues/433))
- Added support for Python 3.12
- Dropped support for Python 3.8

Fixed

- Fixed component filtering misbehaving for CBPE results. [(423)](https://github.com/NannyML/nannyml/issues/423)

Fixed

- Fixed broken links in usage logging docs. Cheers once more to [NeoKish](https://github.com/neokish)! [(#417)](https://github.com/NannyML/nannyml/issues/417)
- Fixed issues with runner type validation due to changes in Pydantic 2 behavior. [(421)](https://github.com/NannyML/nannyml/issues/421)
- Fixed a typo in one the plotting blueprint modules. Eagle eyes [nikml](https://github.com/nikml)! [(#418)](https://github.com/NannyML/nannyml/issues/418)

Added

- Added multiclass support for estimated and realized performance metrics `average_precision` and `business_value`. [(409)](https://github.com/NannyML/nannyml/issues/409)
- Added threshold value limits for multiclass metrics. [(411)](https://github.com/NannyML/nannyml/issues/411)

Changed

- Made the dependencies required for database access optional. Big thanks to [Duncan-Hunter](https://github.com/Duncan-Hunter)
- Improved denominator checks in CBPE base estimation functions. [(416)](https://github.com/NannyML/nannyml/issues/416)
- Relaxed constraints for the `rich` dependency. [(422)](https://github.com/NannyML/nannyml/issues/422)

Removed

- Dropped support for Python 3.7 as it was causing major issues with dependencies. [(410)](https://github.com/NannyML/nannyml/issues/410)

Changed

- Updated `Pydantic` to `^2.7.4`, `SQLModel` to `^0.0.19`. [(401)](https://github.com/NannyML/nannyml/issues/401)
- Removed the `drop_duplicates` step from the `DomainClassifier` for a further speedup. [(402)](https://github.com/NannyML/nannyml/issues/402)
- Reverted to previous working dependency configuration for `matplotlib` as the current one causes issues in `conda`. [(403)](https://github.com/NannyML/nannyml/issues/403)

Fixed

- Added `DomainClassifier` method for drift detection to be run in the CLI.
- Fixed `NaN` handling for multiclass confusion matrix estimation in CBPE. [(400)](https://github.com/NannyML/nannyml/issues/400)
- Fixed incorrect handling of columns marked as categorical in Wasserstein and Hellinger drift detection methods.
The `treat_as_categorical` value was ignored. We've also added a `treat_as_continuous` column to explicitly mark columns as continuous.
[(404)](https://github.com/NannyML/nannyml/issues/404)
- Fixed an issue with multiclass `AUROC` calculation and estimation when not all classes are available in a
reference chunk during fitting. [(405)](https://github.com/NannyML/nannyml/issues/405)

Added

- Added a new data quality calculator to check if continuous values in analysis data are within the ranges
encountered in the reference data. Big thanks to [jnesfield](https://github.com/jnesfield)! Still needs some documentation...
[(408)](https://github.com/NannyML/nannyml/issues/408)

Changed

- Optimized summary stats and overall performance by avoiding unnecessary copy operations and index resets in during chunking
[(390)](https://github.com/NannyML/nannyml/issues/390)
- Optimized performance of `nannyml.base.PerMetricPerColumnResult` filter operations by adding a short-circuit path
when only filtering on period. [(391)](https://github.com/NannyML/nannyml/issues/391)
- Optimized performance of all data quality calculators by avoiding unnecessary evaluations and avoiding copy and index reset operations
[(392)](https://github.com/NannyML/nannyml/issues/392)

Fixed

- Fixed an issue in the Wasserstein "big data heuristic" where outliers caused the binning to cause out-of-memory errors. Thanks! [nikml](https://github.com/nikml)!
[(393)](https://github.com/NannyML/nannyml/issues/393)
- Fixed a typo in the `salary_range` values of the synthetic car loan example dataset. `20K - 20K €` is now `20K - 40K €`.
[(395)](https://github.com/NannyML/nannyml/issues/395)

Changed

- Make predictions optional for performance calcuation. When not provided, only AUROC and average precision will be calculated. [(380)](https://github.com/NannyML/nannyml/issues/380)
- Small DLE docs updates
- Combed through and optimized the reconstruction error calculation with PCA resulting in a nice speedup. Cheers [nikml](https://github.com/nikml)! [(#385)](https://github.com/NannyML/nannyml/issues/385)
- Updated summary stats value limits to be in line with the rest of the library. Changed from `np.nan` to `None`. [(387)](https://github.com/NannyML/nannyml/issues/387)

Fixed

- Fixed a breaking issue in the sampling error calculation for the median summary statistic when there is only a single value for a column. [(377)](https://github.com/NannyML/nannyml/issues/377)
- Drop `identifier` column from the documentation example for reconstruction error calculation with PCA. [(382)](https://github.com/NannyML/nannyml/issues/382)
- Fix an issue where default threshold configurations would get changed when upon setting custom thresholds, bad mutables! [(386)](https://github.com/NannyML/nannyml/issues/386)