R: Monte-Carlo Test on a Discriminant Analysis (in C).

randtest.discrimin {ade4}

R Documentation

Monte-Carlo Test on a Discriminant Analysis (in C).

Description

Test of the sum of a discriminant analysis eigenvalues (divided by the rank). Non parametric version of the Pillai's test. It authorizes any weighting.

Usage

## S3 method for class 'discrimin':
randtest(xtest, nrepet = 999, ...)

Arguments

`xtest`	an object of class `discrimin`
`nrepet`	the number of permutations
`...`	further arguments passed to or from other methods

Value

returns a list of class randtest

Author(s)

Jean Thioulouse ade4-jt@biomserv.univ-lyon1.fr

Examples

data(meaudret)
pca1 <- dudi.pca(meaudret$mil, scan = FALSE, nf = 3)
rand1 <- randtest(discrimin(pca1, meaudret$plan$dat, scan = FALSE), 99)
rand1
#Monte-Carlo test
#Observation: 0.3035 
#Call: as.randtest(sim = sim, obs = obs)
#Based on 999 replicates
#Simulated p-value: 0.001 
plot(rand1, main = "Monte-Carlo test")
summary.manova(manova(as.matrix(meaudret$mil)~meaudret$plan$dat), "Pillai")
#                   Df Pillai approx F num Df den Df  Pr(>F)    
# meaudret$plan$dat  3   2.73    11.30     27     30 1.6e-09 ***
# Residuals         16                                          
# ---
# Signif. codes:  0 `***' 0.001 `**' 0.01 `*' 0.05 `.' 0.1 ` ' 1 
# 2.731/9 = 0.3034

Worked out examples


> library(ade4)
> ### Name: randtest.discrimin
> ### Title: Monte-Carlo Test on a Discriminant Analysis (in C).
> ### Aliases: randtest.discrimin
> ### Keywords: multivariate nonparametric
> 
> ### ** Examples
> 
> data(meaudret)
> pca1 <- dudi.pca(meaudret$mil, scan = FALSE, nf = 3)
> rand1 <- randtest(discrimin(pca1, meaudret$plan$dat, scan = FALSE), 99)
> rand1
Monte-Carlo test
Call: randtest.discrimin(xtest = discrimin(pca1, meaudret$plan$dat, 
    scan = FALSE), nrepet = 99)

Observation: 0.3034897 

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

     Std.Obs  Expectation     Variance 
6.1878007203 0.1558929353 0.0005689594 
> #Monte-Carlo test
> #Observation: 0.3035 
> #Call: as.randtest(sim = sim, obs = obs)
> #Based on 999 replicates
> #Simulated p-value: 0.001 
> plot(rand1, main = "Monte-Carlo test")

> summary.manova(manova(as.matrix(meaudret$mil)~meaudret$plan$dat), "Pillai")
                  Df Pillai approx F num Df den Df    Pr(>F)    
meaudret$plan$dat  3 2.7314   11.299     27     30 1.636e-09 ***
Residuals         16                                            
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 
> #                   Df Pillai approx F num Df den Df  Pr(>F)    
> # meaudret$plan$dat  3   2.73    11.30     27     30 1.6e-09 ***
> # Residuals         16                                          
> # ---
> # Signif. codes:  0 `***' 0.001 `**' 0.01 `*' 0.05 `.' 0.1 ` ' 1 
> # 2.731/9 = 0.3034
> 
> 
> 
>

[Package ade4 Index]