circlus: Clustering and Simulation of Spherical Cauchy and PKBD Models in R

Key Features • Project Overview • Data and Analysis • Dependencies • Citation • Links

circlus is an R package for the estimation and clustering of spherical data, seamlessly integrated with the flexmix package. It includes the necessary M-step implementations for both Poisson Kernel-Based Distribution (PKBD) and spherical Cauchy distribution, and provides random number generators for both distributions.

• Clustering: Supports clustering with spherical Cauchy and PKBD models.
• Random Number Generation: Tools for generating random numbers from PKBD and spherical Cauchy distributions.
• Integration with flexmix: Easily integrates with the flexmix package to support flexible, model-based clustering workflows.

You can install the latest release of circlus from CRAN:

install.packages("circlus")

Or install the development version from GitHub:

# install.packages("devtools")
devtools::install_github("lsablica/circlus")

Load the circlus package and explore its core functionalities:

library(circlus)

For a full list of functions and their usage, refer to the circlus reference manual.

circlus requires:

• R (≥ 3.1.0)
• Rcpp (≥ 0.12.18)
• Tinflex (≥ 1.8)
• flexmix
• torch
• methods

It also links to Rcpp and RcppArmadillo.

circlus is licensed under the GPL-3 license.

If you use circlus in your research, please cite the following paper:

@article{,
  author = {Lukas Sablica and Kurt Hornik and Bettina Gr{\"u}n},
  title = {circlus: An {R} Package for Circular and Spherical
                  Clustering Using {Poisson} Kernel-Based and Spherical
                  {Cauchy} Distribution},
  journal = {Austrian Journal of Statistics},
  year = 2025,
  volume = 54,
  number = 3,
  pages = {27--42},
  doi = {10.17713/ajs.v54i3.2061}
}

• CRAN Package Page