repliclust 0.0.4

Generate synthetic data with clusters.

██████  ███████ ██████  ██      ██  ██████ ██      ██    ██ ███████ ████████ 
██   ██ ██      ██   ██ ██      ██ ██      ██      ██    ██ ██         ██    
██████  █████   ██████  ██      ██ ██      ██      ██    ██ ███████    ██    
██   ██ ██      ██      ██      ██ ██      ██      ██    ██      ██    ██    
██   ██ ███████ ██      ███████ ██  ██████ ███████  ██████  ███████    ██    

Description

repliclust is a Python package for generating synthetic datasets with clusters. It allows you to generate many different datasets that are geometrically similar, but without ever touching low-level parameters like cluster centroids or covariance matrices.

Features

• Reproducibly generate clusters with defined geometric characteristics
• Manage cluster overlaps, shapes, and probability distributions through intuitive, high-level controls
• Define custom dataset archetypes to power reproducible and informative benchmarks

Installation

pip install repliclust

Quickstart

from repliclust import Archetype, DataGenerator

# Create archetype for 5 oblong clusters with typical "aspect ratio" of 3
oblong_clusters = Archetype(n_clusters=5, dim=2, n_samples=500,
                            aspect_ref=3, aspect_maxmin=1.5,
                            name="oblong")
# Define the data generator
data_generator = DataGenerator(archetype=oblong_clusters)

# Sample data points X and class labels y
X, y, _ = data_generator.synthesize()

User Guide / Documentation

For a full user guide and documentation, visit the project website: https://repliclust.org.

Citation

To reference repliclust in your work, please cite:

@article{Zellinger:2023,
  title   = {repliclust: Synthetic Data for Cluster Analysis},
  author  = {Zellinger, Michael J and B{\"u}hlmann, Peter},
  journal = {arXiv preprint arXiv:2303.14301},
  doi     = {10.48550/arXiv.2303.14301},
  year    = {2023}
}