Tools for Multivariate Nonparametrics

Classical multivariate methods such as Hotelling's T-test are based on the sample mean vector and the sample covariance matrix, which makes them sensitive to outliers and heavy-tailed distributions. The package provides nonparametric alternatives based on marginal ranks, spatial signs and spatial ranks, as reviewed for example in Nordhausen and Oja (2018) . It contains tests for location and independence, Hotelling's T-test as a parametric reference, the spatial median and other location estimates, and several estimates of shape. The tests can also be applied to invariant coordinates obtained with the package 'ICS'.


Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("ICSNP")

1.2-0 by Klaus Nordhausen, 2 days ago


Browse source code at https://github.com/cran/ICSNP


Authors: Klaus Nordhausen [aut, cre] (ORCID: , Seija Sirkia [aut] , Hannu Oja [aut] , David E. Tyler [aut]


Documentation:   PDF Manual  


GPL (>= 2) license


Depends on mvtnorm, ICS

Suggests testthat


Imported by CAFT, ICSShiny, ICtest, TCIU, ellipticalsymmetry, fitHeavyTail.

Depended on by MNM, MultNonParam, OjaNP.

Suggested by ICS, JADE, SpatialNP, geex.


See at CRAN