Package: outlierMBC 0.0.1.0003

outlierMBC: Sequential Outlier Identification for Model-Based Clustering

Sequential outlier identification for Gaussian mixture models using the distribution of Mahalanobis distances. The optimal number of outliers is chosen based on the dissimilarity between the theoretical and observed distributions of the scaled squared sample Mahalanobis distances. Also includes an extension for Gaussian linear cluster-weighted models using the distribution of studentized residuals. Doherty, McNicholas, and White (2025) <doi:10.48550/arXiv.2505.11668>.

Authors:Ultán P. Doherty [aut, cre, cph], Paul D. McNicholas [aut], Andrea Cappozzo [aut], Arthur White [aut]

outlierMBC_0.0.1.0003.tar.gz
outlierMBC_0.0.1.0003.zip(r-4.7-any)outlierMBC_0.0.1.0003.zip(r-4.6-any)outlierMBC_0.0.1.0003.zip(r-4.5-any)
outlierMBC_0.0.1.0003.tgz(r-4.6-any)outlierMBC_0.0.1.0003.tgz(r-4.5-any)
outlierMBC_0.0.1.0003.tar.gz(r-4.7-any)outlierMBC_0.0.1.0003.tar.gz(r-4.6-any)
outlierMBC_0.0.1.0003.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION
card.svg |card.png
outlierMBC/json (API)

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

Bug tracker:https://github.com/ultanpdoherty/outliermbc/issues

Datasets:
  • gmm_k3n1000o10 - Simulated data set consisting of 1000 observations from 3 Gaussian components and 10 outliers.
  • gmm_k3n2000o20 - Simulated data set consisting of 2000 observations from 3 Gaussian components and 20 outliers.
  • gmm_k3n4000o40 - Simulated data set consisting of 4000 observations from 3 Gaussian components and 40 outliers.
  • lcwm_k3n1000o10 - Simulated data set consisting of 1000 observations from 3 Gaussian components and 10 outliers.
  • lcwm_k3n2000o20 - Simulated data set consisting of 2000 observations from 3 Gaussian components and 20 outliers.
  • lcwm_k3n4000o40 - Simulated data set consisting of 4000 observations from 3 Gaussian components and 40 outliers.

On CRAN:

Conda:

3.18 score 1 scripts 320 downloads 14 exports 86 dependencies

Last updated from:e6c3773d94. Checks:9 OK. Indexed: yes.

TargetResultTimeFilesSyslog
linux-develOK136
source / vignettesOK209
linux-releaseOK134
macos-releaseOK126
macos-oldrelOK98
windows-develOK110
windows-releaseOK106
windows-oldrelOK115
wasm-releaseOK108

Exports:backtrackbacktrack_gmmbacktrack_lcwmfind_grossombc_gmmombc_lcwmplot_backtrackplot_comparisonplot_comparison_bicplot_curveplot_lcwm_dualplot_selectionsimulate_gmmsimulate_lcwm

Dependencies:BHcaretclasscliclockClusterRcodetoolsContaminatedMixtcpp11data.tabledbscandiagramdigestdplyre1071farverforeachfuturefuture.applygenericsggplot2globalsgluegmpgowergtablehardhatipredisobanditeratorsKernSmoothlabelinglatticelavalifecyclelistenvlubridatemagrittrMASSMatrixmclustmixturemnormtModelMetricsmvtnormnlmennetnumDerivparallellypillarpkgconfigplyrpROCprodlimprogressrproxypurrrR6RColorBrewerRcppRcppArmadilloRcppGSLrecipesreshape2rlangrpartS7scalesshapesparsevctrsspatstat.univarspatstat.utilsSQUAREMstringistringrsurvivaltibbletidyrtidyselecttimechangetimeDatetzdbutf8vctrsviridisLitewithr

Readme and manuals

Help Manual

Help pageTopics
Move backwards from the minimum to a more conservative solution.backtrack
Fit a Gaussian mixture model to the backtrack solution.backtrack_gmm
Fit a linear cluster-weighted model to the backtrack solution.backtrack_lcwm
'cwm' function from Angelo Mazza's flexCWM packagecwm
Compute the dissimilarity for a Gaussian mixture model and identify the lowest density observation.distrib_diff_gmm
Compute the dissimilarity for a linear cluster-weighted model and identify the lowest density observation.distrib_diff_lcwm
Compute the dissimilarity for a single component of a Linear CWM.distrib_diff_lcwm_g
Compute the dissimilarity for a single multivariate Gaussian distribution.distrib_diff_mahalanobis
Compute the response dissimilarity for a single component of a Linear CWM.distrib_diff_residual
Find gross outliers.find_gross
Obtain an initial clustering as a component assignment matrix.get_init_z
Simulated data set consisting of 1000 observations from 3 Gaussian components and 10 outliers.gmm_k3n1000o10
Simulated data set consisting of 2000 observations from 3 Gaussian components and 20 outliers.gmm_k3n2000o20
Simulated data set consisting of 4000 observations from 3 Gaussian components and 40 outliers.gmm_k3n4000o40
Simulated data set consisting of 1000 observations from 3 Gaussian components and 10 outliers.lcwm_k3n1000o10
Simulated data set consisting of 2000 observations from 3 Gaussian components and 20 outliers.lcwm_k3n2000o20
Simulated data set consisting of 4000 observations from 3 Gaussian components and 40 outliers.lcwm_k3n4000o40
Constructor for '"outliermbc_gmm"' S3 class.new_outliermbc_gmm
Constructor for "outliermbc_lcwm" S3 object.new_outliermbc_lcwm
Sequentially identify outliers while fitting a Gaussian mixture model.ombc_gmm
Sequentially identify outliers while fitting a linear cluster-weighted model.ombc_lcwm
Plot the dissimilarity curve showing the backtrack solution.plot_backtrack
Plot multiple dissimilarity curves.plot_comparison
Plot multiple dissimilarity curves.plot_comparison_bic
Plot the dissimilarity curve.plot_curve
Plot three dissimilarity curves for 'outlierMBC-LCWM'.plot_lcwm_dual
Plot dissimilarity values for multiple solutions.plot_selection
plot method for '"outliermbc_gmm"' S3 class.plot.outliermbc_gmm
plot method for '"outliermbc_lcwm"' S3 class.plot.outliermbc_lcwm
print method for '"outliermbc_gmm"' S3 class.print.outliermbc_gmm
print method for '"outliermbc_lcwm"' S3 class.print.outliermbc_lcwm
Simulate data from a Gaussian mixture model with outliers.simulate_gmm
Simulate data from a linear cluster-weighted model with outliers.simulate_lcwm
Check if a new sample satisfies the outlier criteria.test_outlier_ombc
Run 'mixture::gpcm' and try alternative covariance structures or initialisations if necessary.try_mixture_gpcm
Produce a single sample that passes the outlier checks.uniform_outlier_ombc
Sample a potential outlier.uniform_sample_lcwm
Obtain the span of the observations for each component.uniform_spans_lcwm
Validator for '"outliermbc_gmm"' S3 class.validate_outliermbc_gmm
Validator for '"outliermbc_lcwm"' S3 class.validate_outliermbc_lcwm