Kernelized Stein Discrepancy for Goodness-of-Fit Tests and Stein Sampling

Provides Stein-discrepancy goodness-of-fit tests and Stein-method-based sampling tools. The tests include kernel Stein discrepancy U- and V-statistics following Liu et al. (2016) and Chwialkowski et al. (2016) , plus the finite set Stein discrepancy test of Jitkrittum et al. (2017) . The sampling tools include Stein thinning, Stein Points, Stein Point Markov chain Monte Carlo, and Stein variational gradient descent following Riabiz et al. (2022) , Chen et al. (2018) , Chen et al. (2019) , and Liu and Wang (2016) . Gaussian mixture utilities are included for constructing example targets, simulation, density evaluation, and score callbacks.

steinsampling

steinsampling provides a unified R workflow for kernel Stein goodness-of-fit tests, Stein sampling, point construction, and sample compression. It includes KSD U- and V-statistic tests, FSSD tests, SVGD, Stein Points, SP-MCMC, and Stein Thinning, together with reusable score, kernel, calibration, optimization, and transition components.

Installation

Install the released version from CRAN:

install.packages("steinsampling")

Paper

The accompanying paper is available on arXiv.

Citation

If you use steinsampling in published work, please cite the accompanying paper. The recommended citation is also available from R:

citation("steinsampling")

@misc{gao2026kernelizedsteindiscrepancygoodnessoffit,
  title         = {Kernelized Stein Discrepancy for Goodness-of-Fit Tests and
                   Stein Sampling in R},
  author        = {Junhao Gao and Ery Arias-Castro},
  year          = {2026},
  eprint        = {2608.26450},
  archivePrefix = {arXiv},
  primaryClass  = {stat.CO},
  url           = {https://arxiv.org/abs/2608.26450}
}

After journal publication, the package citation will be updated to the version of record and its journal DOI.

Links

• Package source and issue tracker: https://github.com/junhao7622/steinsampling
• CRAN package: https://CRAN.R-project.org/package=steinsampling