HDBSCAN¶
The hdbscan library is a suite of tools to use unsupervised learning to find clusters, or dense regions, of a dataset. The primary algorithm is HDBSCAN* as proposed by Campello, Moulavi, and Sander. The library provides a high performance implementation of this algorithm, along with tools for analysing the resulting clustering.
http://hdbscan.readthedocs.io/en/latest/
Available modules¶
The overview below shows which HDBSCAN installations are available per target architecture in EESSI, ordered based on software version (new to old).
To start using HDBSCAN, load one of these modules using a module load command like:
(This data was automatically generated on Fri, 06 Feb 2026 at 13:12:13 UTC)
| aarch64/generic | aarch64/a64fx | aarch64/neoverse_n1 | aarch64/neoverse_v1 | aarch64/nvidia/grace | x86_64/generic | x86_64/amd/zen2 | x86_64/amd/zen3 | x86_64/amd/zen4 | x86_64/intel/cascadelake | x86_64/intel/haswell | x86_64/intel/icelake | x86_64/intel/sapphirerapids | x86_64/intel/skylake_avx512 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HDBSCAN/0.8.38.post1-foss-2023a | x | x | x | x | x | x | x | x | x | x | x | x | x | x |