Lenskit

There are a few more things to add and test before rolling 2025.1, but with the new data model changes I'm going to call it betea.

What's Changed
* Separate PipelineBuilder from Pipeline and update run api by mdekstrand in https://github.com/lenskit/lkpy/pull/606
* Re-add unnormalized DCG metric by mdekstrand in https://github.com/lenskit/lkpy/pull/608
* Add ARM and Windows vanilla Python to CI runs by mdekstrand in https://github.com/lenskit/lkpy/pull/609
* Implement data set builders and the new entity/relationship model by mdekstrand in https://github.com/lenskit/lkpy/pull/610
* Simplify training history components with new dataset by mdekstrand in https://github.com/lenskit/lkpy/pull/612
* Make vocabularies immutable and stop using copy() by mdekstrand in https://github.com/lenskit/lkpy/pull/611
* Tweaks for live class demo by mdekstrand in https://github.com/lenskit/lkpy/pull/613


**Full Changelog**: https://github.com/lenskit/lkpy/compare/v2025.0.0a6...v2025.1.0b1

This is a very small bump that adds better component base class typing.

What's Changed
* Update component configuration type warnings by mdekstrand in https://github.com/lenskit/lkpy/pull/603


**Full Changelog**: https://github.com/lenskit/lkpy/compare/v2025.0.0a5...v2025.0.0a6

Mostly changes around the component API:

- use Pydantic models for configuration
- update Trainable interface to take training options
- add iterative training interface

What's Changed
* Use Pydantic models for component configurations by mdekstrand in https://github.com/lenskit/lkpy/pull/596
* Re-add return type variable for Component by mdekstrand in https://github.com/lenskit/lkpy/pull/598
* Migrate HPF to configuration object and fix BiasConfig serialization by mdekstrand in https://github.com/lenskit/lkpy/pull/600
* Move Trainable interface and add TrainingOptions by mdekstrand in https://github.com/lenskit/lkpy/pull/599
* Add iterative training API by mdekstrand in https://github.com/lenskit/lkpy/pull/601
* rename "Lenskit" to LensKit for consistency by mdekstrand in https://github.com/lenskit/lkpy/pull/602


**Full Changelog**: https://github.com/lenskit/lkpy/compare/v2025.0.0a4...v2025.0.0a5

Mostly some performance and logging / debugging improvements, along with some refactors.

What's Changed
* Update pre-commit hook versions by mdekstrand in https://github.com/lenskit/lkpy/pull/594
* Add directory check in save_parquet by samiravaez in https://github.com/lenskit/lkpy/pull/593
* Refactor Pipeline into private submodule and improve documentation by mdekstrand in https://github.com/lenskit/lkpy/pull/595
* Speed up popularity scoring and top-N ranking by mdekstrand in https://github.com/lenskit/lkpy/pull/597

New Contributors
* samiravaez made their first contribution in https://github.com/lenskit/lkpy/pull/593

**Full Changelog**: https://github.com/lenskit/lkpy/compare/v2025.0.0a3...v2025.0.0a4

This improves some error messages, logging, and versioning; and tags a prerelease before doing data and configuration improvements.

What's Changed
* emit higher-quality error message on node failure by mdekstrand in https://github.com/lenskit/lkpy/pull/584
* Add support for 12Factor-style JSON logging by mdekstrand in https://github.com/lenskit/lkpy/pull/591
* Fix NumPy/PyTorch version problem and item KNN type problem w/ NumPy array by mdekstrand in https://github.com/lenskit/lkpy/pull/592


**Full Changelog**: https://github.com/lenskit/lkpy/compare/v2025.0.0a2...v2025.0.0a3

This rolls up a number of new changes and bugfixes in the run up to 2025.1.

What's Changed
* Move LensKit test support into lenskit.testing package by mdekstrand in https://github.com/lenskit/lkpy/pull/575
* Add user/item vocab options and subsetting to from_interactions_df by mdekstrand in https://github.com/lenskit/lkpy/pull/577
* Implement and use common component test suites by mdekstrand in https://github.com/lenskit/lkpy/pull/576
* Accelerate item-item k-NN algorithms by mdekstrand in https://github.com/lenskit/lkpy/pull/578
* Modest user-KNN speedups + item KNN logging by mdekstrand in https://github.com/lenskit/lkpy/pull/579
* Add basic_logging function and use it in the GettingStarted notebook by mdekstrand in https://github.com/lenskit/lkpy/pull/580
* Add stability declarations to documentation by mdekstrand in https://github.com/lenskit/lkpy/pull/581
* Add predict_pipeline, remove data.bulk module, and clean up batch functions by mdekstrand in https://github.com/lenskit/lkpy/pull/582
* Implement a proxy logger that suppresses debug & info messages by default by mdekstrand in https://github.com/lenskit/lkpy/pull/583


**Full Changelog**: https://github.com/lenskit/lkpy/compare/v2025.0.0a1...v2025.0.0a2