Skip to contents

lumen: Tests and Distributions for DescToolsX

Source: R/lumen-package.R
lumen-package.Rd

The lumen package provides a comprehensive collection of statistical hypothesis tests and probability distributions designed to complement the functionality of DescToolsX.

The package focuses on a consistent user interface, extended methodological coverage, and seamless integration of inferential procedures and distribution functions within a unified framework.

Hypothesis Tests

The package implements a wide range of statistical tests:

Goodness-of-Fit Tests

  • andersonDarlingTest, cramerVonMisesTest

  • lillieTest, jarqueBeraTest, shapiroFranciaTest

  • pearsonTest

Nonparametric Tests

  • signTest, jonckheereTerpstraTest, pageTest

  • mosesTest, siegelTukeyTest, vanWaerdenTest

Post-hoc Procedures

  • dunnTest, conoverTest, nemenyiTest

  • scheffeTest, postHoc

Parametric Tests

  • tTestA, yuenTTest, zTest, varTest

  • hotellingsT2Test, leveneTest

Contingency Table Tests

  • barnardTest, bhapkarTest, breslowDayTest

  • cochranArmitageTest, cochranQTest

  • mantelTrendTest, woolfTest, stuartMaxwellTest

Time Series Tests

  • adfTest, kpssTest

  • durbinWatsonTest, breuschGodfreyTest

Randomness and Independence Tests

  • runsTest, BartelsRankTest, vonNeumannTest

Probability Distributions

The package provides density (d*), distribution (p*), quantile (q*) and random generation (r*) functions for a range of distributions, following the standard R conventions.

Extreme Value Distributions

  • Generalized Extreme Value: dgev, pgev, qgev, rgev

  • Generalized Pareto: dgpd, pgpd, qgpd, rgpd

  • Gumbel and extended Gumbel: dgumbel, dgumbelx, ...

  • Frechet and reverse Weibull: dfrechet, drevweibull, ...

  • Maxima/minima distributions: dextreme, pextreme, ...

Special Distributions

  • Benford distribution: dbenford, pbenford, qbenford, rbenford

  • Order and triangular distributions: dorder, dtri, ...

Utilities

  • scores – Score generation for ordinal contingency tables

  • corTest – Fast correlation testing for matrices

Design Principles

  • Unified interface across hypothesis tests and distributions

  • Standard d/p/q/r naming for distributions

  • Clear classification via @family and @concept

  • Separation of statistical procedures and data transformation utilities