This package provides functions to compute permutation tests in linear models with nuisances variables. The package has several goals :

The lmperm() function

This function is constructed as an extension of the the lm() function for permutation test. It produces t statistics with univariate and bivariate p-value by permutation.

The aovperm() function

This function is constructed as an extension of the the aov() function for permutation test. It produces marginal F statistics (type III) for factorial ANOVA and ANCOVA. Moreover, repeated measures ANOVA can be perform using the same notations used in an aov() formula with +Error(id/within) to specify the random effects.

The clusterlm() function

This function compute cluster-mass statistics for multiple comparisons. It is designed for ERP analysis of unichannel EEG data. The left part of formula object must be a matrix or dataframe which columns represents multiple responses tested on the same experimental design (specified by right part of the formula). This function provides several methods to handle nuisance variables, a F or t statistics, an extension for repeated measure anova and several methods for the multiple comparisons lit the threshold-free cluster enhancement.


If you need help to use the package or want to report errors, contact Jaromil Frossard at


For permutation tests with nuisance variables :

For permutation test in repeated measures ANOVA :

For cluster-mass statistics for the muliple comparison problems :

For the threshold-free cluster-enhancement method :

Academic works using permuco