randtest.pcaiv {ade4}R Documentation

Monte-Carlo Test on the percentage of explained (i.e. constrained) inertia

Description

Performs a Monte-Carlo test on on the percentage of explained (i.e. constrained) inertia. The statistic is the ratio of the inertia (sum of eigenvalues) of the constrained analysis divided by the inertia of the unconstrained analysis.

Usage

## S3 method for class 'pcaiv':
randtest(xtest, nrepet = 99, ...)
## S3 method for class 'cca':
randtest(xtest, nrepet = 99, ...)
## S3 method for class 'pcaivortho':
randtest(xtest, nrepet = 99, ...)

Arguments

xtest an object of class pcaiv, pcaivortho or cca
nrepet the number of permutations
... further arguments passed to or from other methods

Value

a list of the class randtest

Author(s)

Stephane Dray dray@biomserv.univ-lyon1.fr, original code by Raphael Pelissier

Examples

data(rpjdl)
millog <- log(rpjdl$mil + 1)
iv1 <- cca(rpjdl$fau, millog, scan = FALSE)
randtest(iv1)

Worked out examples


> library(ade4)
> ### Name: randtest.pcaiv
> ### Title: Monte-Carlo Test on the percentage of explained (i.e.
> ###   constrained) inertia
> ### Aliases: randtest.pcaiv randtest.cca randtest.pcaivortho
> ### Keywords: multivariate nonparametric
> 
> ### ** Examples
> 
> data(rpjdl)
> millog <- log(rpjdl$mil + 1)
> iv1 <- cca(rpjdl$fau, millog, scan = FALSE)
> randtest(iv1)
Monte-Carlo test
Call: randtest.cca(xtest = iv1)

Observation: 0.2520234 

Based on 99 replicates
Simulated p-value: 0.01 
Alternative hypothesis: greater 

     Std.Obs  Expectation     Variance 
4.439441e+01 4.495246e-02 2.175614e-05 
> 
> 
> 
>
[Package ade4 Index]