Sequentia

Major changes

- Remove `scikit-learn` validation constraints from `IndependentFunctionTransformer`. ([237](https://github.com/eonu/sequentia/pull/237))

Minor changes

- Change default `mean_filter`/`median_filter` width to 5. ([238](https://github.com/eonu/sequentia/pull/238))
- Update repository documentation. ([239](https://github.com/eonu/sequentia/pull/239))

Major changes

- Set `max_nbytes=None` to fix read-only buffer source array error in `joblib.Parallel` (see https://github.com/scikit-learn/scikit-learn/issues/7981). ([#235](https://github.com/eonu/sequentia/pull/235))
- Added `sequentia.preprocessing` module with [`sklearn.preprocessing`](https://scikit-learn.org/stable/modules/classes.html#module-sklearn.preprocessing) compatibility. ([234](https://github.com/eonu/sequentia/pull/234))
- Added `sequentia.pipeline` module for [`sklearn.pipeline`](https://scikit-learn.org/stable/modules/classes.html#module-sklearn.pipeline) compatibility. ([234](https://github.com/eonu/sequentia/pull/234))

Minor changes

- Upgrade `sklearn` version specifier from `>=0.22` to `>=1.0`. ([234](https://github.com/eonu/sequentia/pull/234))
- Upgrade development status classifier to stable. ([233](https://github.com/eonu/sequentia/pull/233))

Major changes

- Fix `CategoricalHMM` and `GaussianMixtureHMM` parameter defaults for `params`/`init_params` being modified. ([231](https://github.com/eonu/sequentia/issues/231))
- Fix `CategoricalHMM` and `GaussianMixtureHMM` `unfreeze()` calling `super().freeze()` instead of `super().unfreeze()`. ([231](https://github.com/eonu/sequentia/issues/231))
- Fix serialization/deserialization for `_KNNMixin` when `weighting=None`. ([231](https://github.com/eonu/sequentia/issues/231))
- Add unit tests. ([231](https://github.com/eonu/sequentia/issues/231))

Minor changes

- Change `load_digits` `numbers` parameter name to `digits`. ([231](https://github.com/eonu/sequentia/issues/231))
- Change `SequentialDataset` properties to not return copies of arrays. ([231](https://github.com/eonu/sequentia/issues/231))
- Remove `SequentialDataset.__eq__`. ([231](https://github.com/eonu/sequentia/issues/231))
- Change `HMMClassifier` `prior` default to `None`. ([231](https://github.com/eonu/sequentia/issues/231))

Minor changes

- Fix broken link on README.md. ([229](https://github.com/eonu/sequentia/issues/229))

Major changes

- Rework interface to follow sklearn-like patterns. ([226](https://github.com/eonu/sequentia/issues/226))
- Remove `preprocessing` module (temporarily until design is finalized). ([226](https://github.com/eonu/sequentia/issues/226))
- Add KNN regression. ([226](https://github.com/eonu/sequentia/issues/226))
- Add HMM classifier with categorical emissions. ([226](https://github.com/eonu/sequentia/issues/226))
- Use Pydantic for better validation. ([226](https://github.com/eonu/sequentia/issues/226))
- Add `datasets` module for sample datasets. ([226](https://github.com/eonu/sequentia/issues/226))
- Split KNN logic across more functions. ([226](https://github.com/eonu/sequentia/issues/226))
- Better multi-processing for KNN. ([226](https://github.com/eonu/sequentia/issues/226))
- Documentation rework + switch Sphinx documentation theme. ([226](https://github.com/eonu/sequentia/issues/226))
- Fix Sakoe-Chiba width calculation. ([226](https://github.com/eonu/sequentia/issues/226))

Major changes

- Add `digits.npz` as package data in `setup.py`. ([221](https://github.com/eonu/sequentia/issues/221))