Skip to content

[pull] develop from photoprism:develop - #1772

Merged
pull[bot] merged 3 commits into
Uncodedtech:developfrom
photoprism:develop
Aug 30, 2026
Merged

[pull] develop from photoprism:develop#1772
pull[bot] merged 3 commits into
Uncodedtech:developfrom
photoprism:develop

Conversation

@pull

@pull pull Bot commented Aug 30, 2026

Copy link
Copy Markdown

See Commits and Changes for more details.


Created by pull[bot] (v2.0.0-alpha.4)

Can you help keep this open source service alive? 💖 Please sponsor : )

DBSCAN takes a single link distance, so a set holding groups of different
densities has no value that suits all of them: whatever connects the sparsest
group also chains the closest pair together.

OPTICS orders points by reachability and extracts clusters from that ordering,
either at one link distance or as valleys by the xi-steep method. HDBSCAN builds
the hierarchy of all density levels, keeps the clusters that persist longest,
and yields a per-point membership and GLOSH outlier score from the same tree.
Neither implements HardClusterer, since neither produces one clustering.

Both agree with scikit-learn on generated sets: core distances and the DBSCAN
extraction exactly, and the labels except where tied distances leave more than
one valid answer. Two properties the README states for callers: a membership
probability is scaled per cluster and is not comparable between clusters, and a
point whose core distance dominates its neighborhood has several equally short
edges, so tie-breaking rather than the data decides which cluster takes it.
@pull pull Bot locked and limited conversation to collaborators Aug 30, 2026
@pull pull Bot added the ⤵️ pull label Aug 30, 2026
@pull
pull Bot merged commit 8d15207 into Uncodedtech:develop Aug 30, 2026
2 checks passed
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant