v0.5.1

kmedoids 0.5.1 (2024-03-14)

• DynMSC: best loss reported incorrectly if best k=2
• add minimum k parameter
• bump rayon version (no changes)
• use pointer-sized np.uintp type for medoids, for wasm32 target
  that should match Rust usize.

v0.5.0

• add DynMSC, Silhouette clustering with optimal number of clusters
• update dependency versions

v0.4.3

Bug fix for silhouette evaluation

kmedoids 0.4.2

kmedoids 0.4.2 (2023-03-07)

• fix predict for non-precomputed distances
• add CITATION.cff for github
• update dependency versions

v0.4.1

Added FastMSC, FasterMSC, PAMMEDSIL and PAMSIL clustering,
to optimize the (Medoid) Silhouette directly.

For details on these methods, please see the upcoming publication
Lars Lenssen, Erich Schubert:
Clustering by Direct Optimization of the Medoid Silhouette
In: 15th International Conference on Similarity Search and Applications (SISAP 2022)
https://doi.org/10.1007/978-3-031-17849-8_15

v0.3.3

Minor bug fix in BUILD (not used by fasterpam)
Improved Python wheels support (by David Muhr)

v0.3.2

Documentation improvements, small bug fixes (in Rust part)
Auto-builds wheels for many more platforms (thank you, @davnn)

v0.3.1

v0.3.1

Version 0.2.0

kmedoids 0.2.0 (2022-01-07)

• make KMedoidsResult a native Python object, pass a tuple from Rust
• always use i64/f64 loss, even for i32/f32 input
• add random shuffling support
• add parallelization support
• add silhouette

See also: https://github.com/kno10/rust-kmedoids/blob/v0.2.0/CHANGELOG.md