Clustering.jl Versions Save

Clustering v0.15.7

Diff since v0.15.6

Merged pull requests:

• Providing intrinsic clustering quality indices (#257) (@jaksle)
• build(deps): bump actions/checkout from 3 to 4 (#268) (@dependabot[bot])

Clustering v0.15.6

Diff since v0.15.5

Merged pull requests:

• fuzzycmeans() enhancements (#265) (@alyst)
• Add ClusteringBenchmarks.jl reference (#266) (@alyst)
• Cleanup kmeans() seeding (#267) (@alyst)

Closed issues:

• Link to benchmarks from README? (#264)

Clustering v0.15.5

Diff since v0.15.4

Merged pull requests:

• Distributed silhouette (#255) (@nomadbl)
• Fix documentation generation (#262) (@alyst)

Clustering v0.15.4

Diff since v0.15.3

Closed issues:

• Fix warnings for latest Distances.jl (#260)

Merged pull requests:

• Fix affprop() random CI failures (#246) (@alyst)
• Update to Distances.jl 0.10.9 (#259) (@alyst)

Clustering v0.15.3

Diff since v0.15.2

Closed issues:

• Unexpected behaviour of cutree - bug with :optimallayout? (#252)

Merged pull requests:

• Stop using unexported Base.permute!! (#229) (@LilithHafner)
• Fix cutree() for trees with unsorted merges (#254) (@alyst)

Clustering v0.15.2

Diff since v0.15.1

Closed issues:

• Docs not deployed for tags. (#247)

Merged pull requests:

• CompatHelper: bump compat for StatsBase to 0.34, (keep existing compat) (#251) (@github-actions[bot])

Clustering v0.15.1

Diff since v0.15.0

Closed issues:

• AssertionError in kmedoids alg (#231)

Merged pull requests:

• kmedoids(): fix duplicate medoids case (#250) (@alyst)

Clustering v0.15.0

This minor release consolidates the two implementations of DBSCAN fixing some bugs and introducing breaking changes that require updates to the user code:

• res = dbscan(point, radius; kwargs...) now returns DbscanResult instead of Vector{DbscanCluster}. To get the vector of clusters use res.clusters.
• dbscan(dists, radius, N) is deprecated. Use dbscan(dists, radius, min_neighbors=N, metric=nothing) instead.

Diff since v0.14.4

Closed issues:

• DBSCAN DenseMatrix Requirement (#153)
• DBscan Adjacency Lists have repeated clusters (#200)
• Docu for DBscan (#201)
• Readthedocs github link broken (#202)
• DBSCAN docs - consistency between 2 APIs (#221)

Merged pull requests:

• Refactor dbscan() (#248) (@alyst)
• Adjust CI setup for docs (#249) (@ararslan)

Clustering v0.14.4

Diff since v0.14.3

Closed issues:

• Adjusted Rand index inconsistency for large n (#225)
• TagBot trigger issue (#234)
• -1 silhouette score returned with empty classes (#241)

Merged pull requests:

• fixed ARI (fixes #225 & #226) (#227) (@wildart)
• Remove use of old typealiases from StatsBase (#240) (@yuyichao)
• silhouette(): fix empty clusters case (#242) (@alyst)
• randindex(): fix int overflow (#243) (@alyst)
• add dependabot.yml (#244) (@alyst)
• Update CI scripts (#245) (@alyst)

What's Changed

• Remove nonexistent export of kmeans_opts by @nalimilan in https://github.com/JuliaStats/Clustering.jl/pull/210
• CI: Switch to GitHub Actions and Codecov by @alyst in https://github.com/JuliaStats/Clustering.jl/pull/211
• tests: increase vmeasure test iterations to 200 by @alyst in https://github.com/JuliaStats/Clustering.jl/pull/212
• Fix typo by @Alex-vZyl in https://github.com/JuliaStats/Clustering.jl/pull/232
• Add rng parameter to appropriate methods by @wildart in https://github.com/JuliaStats/Clustering.jl/pull/228
• Remove test for a method from StatsBase by @ararslan in https://github.com/JuliaStats/Clustering.jl/pull/233

New Contributors

• @Alex-vZyl made their first contribution in https://github.com/JuliaStats/Clustering.jl/pull/232

Full Changelog: https://github.com/JuliaStats/Clustering.jl/compare/v0.14.2...v0.14.3