https://github.com/baggepinnen/hdbscan.jl
A julia wrapper for HDBSCAN
https://github.com/baggepinnen/hdbscan.jl
Last synced: 3 months ago
JSON representation
A julia wrapper for HDBSCAN
- Host: GitHub
- URL: https://github.com/baggepinnen/hdbscan.jl
- Owner: baggepinnen
- Created: 2019-10-03T02:57:59.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2021-05-09T10:02:42.000Z (over 4 years ago)
- Last Synced: 2025-05-31T10:42:50.194Z (5 months ago)
- Language: Julia
- Size: 11.7 KB
- Stars: 5
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# HDBSCAN
This package is a very simple wrapper around [hdbscan](https://github.com/scikit-learn-contrib/hdbscan) for python. It is not feature complete, but fully functional.
## Functions
```julia
using HDBSCAN, Clustering
result = hdbscan(X; min_cluster_size=5, min_samples=min_cluster_size, kwargs...)
probabilities(result)
exemplars(result) # Computed at first call, this takes long time
outlier_scores(result)
result.assigments # Contains the cluster assignments
```
where `X` is `n_features × n_points` (the convention of Clustering.jl, opposite to the convention of the python library).
The `result::HdbscanResult <: Clustering.ClusteringResult` contains the `PyObject` clusterer which can be used to access everything that is not wrapped.
The label assignments are stored in `result.assigments`. 0 values indicate noise (-1 in python version), positive values indicate a cluster assignment.
## Documentation
The original documentation is available here
https://hdbscan.readthedocs.io/en/latest/basic_hdbscan.html