hdbscan-rs
High-performance HDBSCAN clustering for Python, powered by a Rust core. Drop-in compatible with scikit-learn's API, but significantly faster -- especially on small and large datasets.
Installation
pip install hdbscan-rs
Requires Python >= 3.11 and NumPy >= 1.20. Pre-built wheels available for Linux, macOS, and Windows.
Quick start
import numpy as np
from hdbscan_rs import HDBSCAN
data = np.random.randn(10000, 2)
clusterer = HDBSCAN(min_cluster_size=15)
labels = clusterer.fit_predict(data)
print(f"Found {labels.max() + 1} clusters, {(labels == -1).sum()} noise points")
Migrating from other HDBSCAN packages
From hdbscan (standalone)
- import hdbscan
- clusterer = hdbscan.HDBSCAN(min_cluster_size=15, min_samples=2, metric='euclidean')
+ from hdbscan_rs import HDBSCAN
+ clusterer = HDBSCAN(min_cluster_size=15, min_samples=2, metric='euclidean')
labels = clusterer.fit_predict(data)
From sklearn.cluster.HDBSCAN
- from sklearn.cluster import HDBSCAN
+ from hdbscan_rs import HDBSCAN
clusterer = HDBSCAN(min_cluster_size=15)
labels = clusterer.fit_predict(data)
From BERTopic
from hdbscan_rs import HDBSCAN
from bertopic import BERTopic
topic_model = BERTopic(hdbscan_model=HDBSCAN(min_cluster_size=15))
No other code changes needed. Labels, probabilities, and outlier scores are all compatible.
API
HDBSCAN(
min_cluster_size=5, # Smallest group that counts as a cluster
min_samples=None, # Controls density estimate (default: min_cluster_size)
metric="euclidean", # "euclidean", "manhattan", "cosine", "minkowski", "precomputed"
p=None, # Minkowski p parameter
alpha=1.0, # Mutual reachability scaling factor
cluster_selection_epsilon=0.0, # Merge clusters below this distance
cluster_selection_method="eom", # "eom" (Excess of Mass) or "leaf"
allow_single_cluster=False,
)
Methods
fit_predict(X)-- Fit and return cluster labels (numpy array, -1 = noise)fit(X)-- Fit the model without returning labelsapproximate_predict(X)-- Predict labels for new points (returns labels, probabilities)
Properties (after fitting)
labels_-- Cluster labels (-1 = noise)probabilities_-- Membership strength [0, 1]outlier_scores_-- GLOSH outlier scores [0, 1]
Performance
Best-of-3 wall time on a 4-core AMD EPYC. Data is make_blobs with 5 centers, min_cluster_size=10. Numbers are from the native Rust core; the Python binding adds <5ms overhead for data conversion.
| Config | sklearn | C-hdbscan | fast-hdbscan | hdbscan-rs | vs sklearn | vs fast |
|---|---|---|---|---|---|---|
| 1Kx2D | 12.0 ms | 12.7 ms | 3.8 ms | 2.2 ms | 5.4x | 1.7x |
| 5Kx2D | 121 ms | 76.0 ms | 20.6 ms | 9.2 ms | 13.1x | 2.2x |
| 10Kx2D | 445 ms | 181 ms | 45.1 ms | 17.8 ms | 25.0x | 2.5x |
| 50Kx2D | 12,757 ms | 1,011 ms | 302 ms | 101 ms | 126x | 3.0x |
| 5Kx10D | 240 ms | 133 ms | 70.5 ms | 49 ms | 4.9x | 1.4x |
| 1Kx256D | 235 ms | 230 ms | 65.4 ms | 19 ms | 12.1x | 3.4x |
| 500x1536D | 412 ms | 439 ms | 80.7 ms | 27 ms | 15.1x | 3.0x |
Memory: 3-41 MB (Rust) vs 120-150 MB (sklearn) vs 121-169 MB (C-hdbscan) vs 457-470 MB (fast-hdbscan + Numba JIT).
Precomputed distances
from hdbscan_rs import HDBSCAN
import numpy as np
# Compute your own distance matrix
dist_matrix = np.array([[0, 1, 5], [1, 0, 3], [5, 3, 0]], dtype=np.float64)
clusterer = HDBSCAN(min_cluster_size=2, metric="precomputed")
labels = clusterer.fit_predict(dist_matrix)
License
Licensed under either of Apache License, Version 2.0 or MIT License, at your option.
1 maintainer
Metadata
1 maintainer
Release files for hdbscan-rs 0.6.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| hdbscan_rs-0.6.1.tar.gz | 620.9 kB | Details |
Built distributions (wheels)
Total release size: 5.9 MB
Release files / hdbscan_rs-0.6.1.tar.gz
| Download URL | hdbscan_rs-0.6.1.tar.gz |
|---|---|
| Size | 620.9 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
e0b8588f2ae9201e3729e7274db3fa77f98af4c4d77de860488f476b67e4dcac
|
|
BLAKE2b-256 checksum How to use checksums |
e5dc89cd1dc6b308a9ff9eeda441a6d885c724b33172b3b9252e90dc4e262d3a
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp313-cp313-win_amd64.whl
| Download URL | hdbscan_rs-0.6.1-cp313-cp313-win_amd64.whl |
|---|---|
| Size | 304.1 kB |
| Tags | CPython 3.13 Windows x86-64 |
|
SHA-256 checksum How to use checksums |
58db33dac67f99dc29eaf7e7a399ad4f317ed9beea257c3f1734c71b002ed2aa
|
|
BLAKE2b-256 checksum How to use checksums |
b1ba037b7f511a77e742e1b237eaa83ef38e5b79ccea85dc6069a48e69c54122
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
| Download URL | hdbscan_rs-0.6.1-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl |
|---|---|
| Size | 394.1 kB |
| Tags | CPython 3.13 Linux glibc 2.17+ x86-64 |
|
SHA-256 checksum How to use checksums |
892f7205c264a31b20c521fc7b4fa39163e061ce498898e53e398d00b696132c
|
|
BLAKE2b-256 checksum How to use checksums |
a1108d67f40ac1077279296f40174046da6748bea1a17ba6a95e01cea110b2b6
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
| Download URL | hdbscan_rs-0.6.1-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl |
|---|---|
| Size | 357.5 kB |
| Tags | CPython 3.13 Linux glibc 2.17+ ARM64 |
|
SHA-256 checksum How to use checksums |
ef26d4961b26ad1f380375fd0b46209463452d546379d2f32a7d6f4d2c5ca766
|
|
BLAKE2b-256 checksum How to use checksums |
0383b7f461dfa33ea24c102911b35666861a41ae504ec28b8e454408b3d225ea
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp313-cp313-macosx_11_0_arm64.whl
| Download URL | hdbscan_rs-0.6.1-cp313-cp313-macosx_11_0_arm64.whl |
|---|---|
| Size | 339.8 kB |
| Tags | CPython 3.13 macOS 11.0+ ARM64 |
|
SHA-256 checksum How to use checksums |
536397b09d7589e53f130234e38382508482d92c2399f3cdf3bd1efce918bd38
|
|
BLAKE2b-256 checksum How to use checksums |
3ed4309ff824c74bd4e41506e8030c23d88ae3658231319e870f62a86ef16080
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp313-cp313-macosx_10_12_x86_64.whl
| Download URL | hdbscan_rs-0.6.1-cp313-cp313-macosx_10_12_x86_64.whl |
|---|---|
| Size | 359.9 kB |
| Tags | CPython 3.13 macOS 10.12+ x86-64 |
|
SHA-256 checksum How to use checksums |
73455d77dccffa80e4a4e4bc301da14b95deff6074ace489d37b03920f0d9f4f
|
|
BLAKE2b-256 checksum How to use checksums |
f185ab270e0457ecf83228e1d068e5ac75c3ec0d8170822bebfc28de8852f389
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp312-cp312-win_amd64.whl
| Download URL | hdbscan_rs-0.6.1-cp312-cp312-win_amd64.whl |
|---|---|
| Size | 304.4 kB |
| Tags | CPython 3.12 Windows x86-64 |
|
SHA-256 checksum How to use checksums |
56541fa8e2b29538ec1328dd7e979fe98b66ad71043d05b210b2374d4fc6bb50
|
|
BLAKE2b-256 checksum How to use checksums |
a6cc92b1256d2f53faa26e464e0c089eabc37883080c5f050fe19165aa661731
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
| Download URL | hdbscan_rs-0.6.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl |
|---|---|
| Size | 394.5 kB |
| Tags | CPython 3.12 Linux glibc 2.17+ x86-64 |
|
SHA-256 checksum How to use checksums |
4cb3e116d611927e0010045e6f9893c33ddf1d14e673ce54b58a1fcb02f6ef60
|
|
BLAKE2b-256 checksum How to use checksums |
60f2ec98047a9717dce74f8eb74d1eff382b5354213466173bf0125ce85f2f29
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
| Download URL | hdbscan_rs-0.6.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl |
|---|---|
| Size | 358.0 kB |
| Tags | CPython 3.12 Linux glibc 2.17+ ARM64 |
|
SHA-256 checksum How to use checksums |
bead81e56d39e0e4c992cac021afebfe256ab319a1b7c5b470c2b683f29bfbff
|
|
BLAKE2b-256 checksum How to use checksums |
f56a69aa85599e2fb3764bc346ec527abd050a56db7ebf7598437d78a6881303
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp312-cp312-macosx_11_0_arm64.whl
| Download URL | hdbscan_rs-0.6.1-cp312-cp312-macosx_11_0_arm64.whl |
|---|---|
| Size | 340.2 kB |
| Tags | CPython 3.12 macOS 11.0+ ARM64 |
|
SHA-256 checksum How to use checksums |
ef0b76a8ac6ec983ec3e59616ede7d9f6130908419e9b02d05f4a94a2c6799eb
|
|
BLAKE2b-256 checksum How to use checksums |
d3052e4f3a75ef7ae1a7a5671d93d13d376e5f1d0e370250c724cb90e68c4674
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp312-cp312-macosx_10_12_x86_64.whl
| Download URL | hdbscan_rs-0.6.1-cp312-cp312-macosx_10_12_x86_64.whl |
|---|---|
| Size | 360.3 kB |
| Tags | CPython 3.12 macOS 10.12+ x86-64 |
|
SHA-256 checksum How to use checksums |
b1aaca8c1b66eb8928b4c031a91338b845fbb70f1fdb426255306394dc704331
|
|
BLAKE2b-256 checksum How to use checksums |
181371eab1f39bd8422fbff6af3cf2cffb0d26bd8bdafd0fd93cbff20864df69
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp311-cp311-win_amd64.whl
| Download URL | hdbscan_rs-0.6.1-cp311-cp311-win_amd64.whl |
|---|---|
| Size | 303.4 kB |
| Tags | CPython 3.11 Windows x86-64 |
|
SHA-256 checksum How to use checksums |
883925a3a1896255a3cb2229ef1c0b2a19571ac433073ff4a9c0d34de9c27169
|
|
BLAKE2b-256 checksum How to use checksums |
4505b25dad69bd6a4d583631f6c26e7b439678d00a69b7820a08e77711a6c427
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
| Download URL | hdbscan_rs-0.6.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl |
|---|---|
| Size | 395.5 kB |
| Tags | CPython 3.11 Linux glibc 2.17+ x86-64 |
|
SHA-256 checksum How to use checksums |
ffb65b4d571d25f536c9b8c765c968318a1208481e30087043f61549b3bbe500
|
|
BLAKE2b-256 checksum How to use checksums |
121e4c4e24ad254daf186252c13c88c167eb29d2a400ff3e705304d2c7be25cb
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
| Download URL | hdbscan_rs-0.6.1-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl |
|---|---|
| Size | 359.4 kB |
| Tags | CPython 3.11 Linux glibc 2.17+ ARM64 |
|
SHA-256 checksum How to use checksums |
159df5c2d1f6e0d77d81779bc52ab88670a70e2bafeeccff4c778b10e0b52df2
|
|
BLAKE2b-256 checksum How to use checksums |
5183eba6653c29478159991b7f15eba1a52be3dc2668e72d6be67190644f1204
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp311-cp311-macosx_11_0_arm64.whl
| Download URL | hdbscan_rs-0.6.1-cp311-cp311-macosx_11_0_arm64.whl |
|---|---|
| Size | 340.9 kB |
| Tags | CPython 3.11 macOS 11.0+ ARM64 |
|
SHA-256 checksum How to use checksums |
2072497953afc9b74de7920c3fca197cbfddafe36de52fb7a527f27c02875964
|
|
BLAKE2b-256 checksum How to use checksums |
1923a2221a169cb96f76f03ec2533f47aaff659a8a56620c0dbf66d64c257f2e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency logRelease files / hdbscan_rs-0.6.1-cp311-cp311-macosx_10_12_x86_64.whl
| Download URL | hdbscan_rs-0.6.1-cp311-cp311-macosx_10_12_x86_64.whl |
|---|---|
| Size | 360.9 kB |
| Tags | CPython 3.11 macOS 10.12+ x86-64 |
|
SHA-256 checksum How to use checksums |
286322da3a98c255084d09101406c39324af8406cb4108759935299f4f50bdb1
|
|
BLAKE2b-256 checksum How to use checksums |
91fdcd8b499a33c08a26c08eb33c0ff03defc38ae8aee9c8623d7e98a6593945
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Mar 13, 2026.
Transparency log
