Package: steinsampling 0.1.0

steinsampling: Kernel Stein Discrepancy Goodness-of-Fit and Stein Sampling Tools

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) <doi:10.48550/arXiv.1602.03253> and Chwialkowski et al. (2016) <doi:10.48550/arXiv.1602.02964>, plus the finite set Stein discrepancy test of Jitkrittum et al. (2017) <doi:10.48550/arXiv.1705.07673>. The sampling tools include Stein thinning, Stein Points, Stein Point Markov chain Monte Carlo, and Stein variational gradient descent following Riabiz et al. (2022) <doi:10.48550/arXiv.2005.03952>, Chen et al. (2018) <doi:10.48550/arXiv.1803.10161>, Chen et al. (2019) <doi:10.48550/arXiv.1905.03673>, and Liu and Wang (2016) <doi:10.48550/arXiv.1608.04471>. Gaussian mixture utilities are included for simulation, likelihoods, posterior probabilities, scores, and plots.

Authors:Junhao Gao [aut, cre], Ery Arias-Castro [aut]

steinsampling_0.1.0.tar.gz
steinsampling_0.1.0.zip(r-4.7-any)steinsampling_0.1.0.zip(r-4.6-any)steinsampling_0.1.0.zip(r-4.5-any)
steinsampling_0.1.0.tgz(r-4.6-any)steinsampling_0.1.0.tgz(r-4.5-any)
steinsampling_0.1.0.tar.gz(r-4.7-any)steinsampling_0.1.0.tar.gz(r-4.6-any)
steinsampling_0.1.0.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION
card.svg |card.png
steinsampling/json (API)

# Install 'steinsampling' in R:
install.packages('steinsampling', repos = c('https://junhao7622.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/junhao7622/steinsampling/issues

On CRAN:

Conda:

goodness-of-fitmachine-learningsamplingstatistics

2.30 score 51 exports 1 dependencies

Last updated from:4ba336054a. Checks:7 NOTE, 2 OK. Indexed: yes.

TargetResultTimeFilesSyslog
linux-devel-x86_64NOTE122
source / vignettesOK189
linux-release-x86_64NOTE123
macos-release-arm64NOTE133
macos-oldrel-arm64NOTE214
windows-develNOTE79
windows-releaseNOTE78
windows-oldrelNOTE73
wasm-releaseOK107

Exports:compute_fssd_null_pvaluecompute_fssd_unbiased_statcompute_taucross_kernelcustom_adjusted_gradientcustom_stein_kerneleval_kernelfind_median_distancefmin_gridfmin_mcfmin_nmfssd_opt_testfssd_rand_testfssd_testget_score_evaluatorgmmgrad_theta_v_kernelgrad_x_kernelgrwgrwmetropksd_u_bootstrapksd_u_statisticksd_u_testksd_uq_matrixksd_v_bootstrapksd_v_statisticksd_v_testksd_vq_matrixlikelihoodgmmmalaperturbgmmplotgmmposteriorgmmrgmmrwmscorefunctiongmmsp_mcmcsp_mcmc_criterionsp_mcmc_eval_candidatessp_mcmc_select_startsp_mcmc_statestein_codescentstein_kernelstein_kernel_imq_scorestein_kernel_inverse_logstein_kernel_matrixstein_pointsstein_thinningsvgdtrace_mixed_kernelupdate_svgd

Dependencies:mvtnorm

Readme and manuals

Help Manual

Help pageTopics
steinsampling: Stein tests and Stein sampling toolssteinsampling-package steinsampling
Simulate the FSSD null distributioncompute_fssd_null_pvalue
Estimate squared FSSD from a feature matrixcompute_fssd_unbiased_stat
Compute the FSSD feature matrixcompute_tau
Apply a custom SVGD direction adjustmentcustom_adjusted_gradient
Create a custom Stein kernelcustom_stein_kernel
Median squared distance between sample pairsfind_median_distance
Create the grid search used by Stein Pointsfmin_grid
Create the Monte Carlo search used by Stein Pointsfmin_mc
Create the multi-start Nelder-Mead search used by Stein Pointsfmin_nm
FSSD test with optimized test locationsfssd_opt_test
FSSD test with random test locationsfssd_rand_test
Finite Set Stein Discrepancy goodness-of-fit testfssd_test
Create a score function for a fixed Gaussian mixture modelget_score_evaluator
Create a Gaussian mixture modelgmm
Run the Gaussian random-walk transition used by SP-MCMCgrw
Run a Gaussian random-walk Metropolis chaingrwmetrop
Building blocks for Stein kernelscross_kernel eval_kernel grad_theta_v_kernel grad_x_kernel kernel_generics trace_mixed_kernel
Bootstrap the KSD U-statistic from a Stein-kernel matrixksd_u_bootstrap
Compute the KSD U-statistic from its Stein-kernel matrixksd_u_statistic
KSD goodness-of-fit test using the Liu et al. U-statisticksd_u_test
Build the Stein-kernel matrix for the KSD U-statisticksd_uq_matrix
Wild-bootstrap the KSD V-statisticksd_v_bootstrap
Compute the KSD V-statistic from its Stein-kernel matrixksd_v_statistic
KSD goodness-of-fit test using the Chwialkowski et al. V-statisticksd_v_test
Build the Stein-kernel matrix for the KSD V-statisticksd_vq_matrix
Compute the mixture density for sampleslikelihoodgmm
Run a Metropolis-adjusted Langevin chainmala
Perturb the component means of a Gaussian mixture modelperturbgmm
Plot a one-dimensional Gaussian mixture sampleplotgmm
Compute posterior component probabilities for a Gaussian mixtureposteriorgmm
Sample from a Gaussian mixture modelrgmm
Alias for the Gaussian random-walk transitionrwm
Compute the score of a Gaussian mixture densityscorefunctiongmm
Generate Stein Points from short MCMC candidate pathssp_mcmc
Create an SP-MCMC start-point rulesp_mcmc_criterion
Score SP-MCMC candidate pointssp_mcmc_eval_candidates
Choose the next SP-MCMC chain startsp_mcmc_select_start
Store the current SP-MCMC state for a start rulesp_mcmc_state
Refine Stein Points by coordinate descentstein_codescent
Create a built-in Stein kernelstein_kernel
Create a score-distance IMQ Stein kernelstein_kernel_imq_score
Create an inverse-log Stein kernelstein_kernel_inverse_log
Compute pairwise Stein kernel valuesstein_kernel_matrix
Build a Stein Points sequence by greedy KSD minimizationstein_points
Compress existing samples with Stein thinningstein_thinning
Create an SVGD updatersvgd
Run SVGD particle updatesupdate_svgd