scikit-clarans 0.1.3

A scikit-learn compatible implementation of CLARANS clustering algorithm

scikit-clarans

A scikit-learn compatible implementation of the CLARANS (Clustering Large Applications based on RANdomized Search) algorithm.

CLARANS acts as a bridge between the high quality of PAM (Partition Around Medoids) and the speed required for large datasets. By using randomized search instead of exhaustive search, it finds high-quality medoids efficiently without exploring the entire graph of solutions.

---

Features

• Scikit-Learn Native: Use it just like KMeans or DBSCAN. Drop-in compatibility for pipelines and cross-validation.
• Scalable: Designed to handle datasets where standard PAM/k-medoids is too slow.
• Flexible: Choose from multiple initialization strategies (k-medoids++, build, etc.) and distance metrics (euclidean, manhattan, cosine, etc.).

Installation

Install simply via pip:

pip install scikit-clarans

Or install from source:

pip install .

For development

pip install -e ".[dev]"

Quick Start

CLARANS

from clarans import CLARANS
from sklearn.datasets import make_blobs

# 1. Create dummy data
X, _ = make_blobs(n_samples=1000, centers=5, random_state=42)

# 2. Initialize CLARANS
#    - n_clusters: 5 clusters
#    - numlocal: 3 restarts for better quality
#    - init: 'k-medoids++' for smart starting points
clarans = CLARANS(n_clusters=5, numlocal=3, init='k-medoids++', random_state=42)

# 3. Fit
clarans.fit(X)

# 4. Results
print("Medoid Indices:", clarans.medoid_indices_)
print("Labels:", clarans.labels_)

FastCLARANS

For datasets that fit in memory, FastCLARANS can provide significant speedups by caching pairwise distances:

from clarans import FastCLARANS

fast_model = FastCLARANS(n_clusters=5, numlocal=3, random_state=42)
fast_model.fit(X)

Examples

This repository includes a number of runnable examples in the examples/ folder showing common usage patterns, integrations and a Jupyter notebook (examples/clarans_examples.ipynb) with many interactive recipes. Run any example with::

python examples/01_quick_start.py

Documentation

For full API reference and usage guides, please see the Documentation.

Contributing

Contributions are welcome! Please check out CONTRIBUTING.md for guidelines.

License

This project is licensed under the MIT License - see the LICENSE file for details.