R: Function to perform a test of dimensionality

testdim {ade4}

R Documentation

Function to perform a test of dimensionality

Description

This functions allow to test for the number of axes in multivariate analysis. The procedure is only implemented for principal component analysis on correlation matrix. The procedure is based on the computation of the RV coefficient.

Usage

testdim(dudi, ...)
## S3 method for class 'pca':
testdim(dudi, nrepet = 99, nbax = dudi$rank, alpha = 0.05, ...)

Arguments

`dudi`	a duality diagram (an object of class `dudi`)
`nrepet`	the number of repetitions for the permutation procedure
`nbax`	the number of axes to be tested, by default all axes
`alpha`	the significance level
`...`	other arguments

Value

An object of the class krandtest. It contains also:

nb The estimated number of axes to keep

nb.cor The number of axes to keep estimated using a sequential Bonferroni procedure

Author(s)

Stephane Dray dray@biomserv.univ-lyon1.fr

References

Dray, S. (2007) On the number of principal components: A test of dimensionality based on measurements of similarity between matrices. Computational Statistics and Data Analysis, in press.

Examples

tab <- data.frame(matrix(rnorm(200),20,10))
pca1 <- dudi.pca(tab,scannf=FALSE)
test1 <- testdim(pca1)
test1
test1$nb
test1$nb.cor
data(doubs)
pca2 <- dudi.pca(doubs$mil,scannf=FALSE)
test2 <- testdim(pca2)
test2
test2$nb
test2$nb.cor

Worked out examples


> library(ade4)
> ### Name: testdim
> ### Title: Function to perform a test of dimensionality
> ### Aliases: testdim testdim.pca
> ### Keywords: multivariate
> 
> ### ** Examples
> 
> tab <- data.frame(matrix(rnorm(200),20,10))
> pca1 <- dudi.pca(tab,scannf=FALSE)
> test1 <- testdim(pca1)
> test1
class: krandtest 
Monte-Carlo tests
Call: testdim.pca(dudi = pca1)

Test number:   10 
Permutation number:   99 
      Test       Obs    Std.Obs   Alter Pvalue
1   Axis 1 0.5754389 -0.7788928 greater   0.81
2   Axis 2 0.6644669  0.9606160 greater   0.20
3   Axis 3 0.6557210  0.1901559 greater   0.42
4   Axis 4 0.6970035 -0.1572369 greater   0.48
5   Axis 5 0.7389212  1.0312897 greater   0.15
6   Axis 6 0.7195032  1.6013217 greater   0.08
7   Axis 7 0.7520397  2.3792042 greater   0.02
8   Axis 8 0.8616320  3.6664826 greater   0.01
9   Axis 9 0.7311890  0.3059093 greater   0.35
10 Axis 10 1.0000000  4.7288302 greater   0.01

other elements: nb nb.cor 
> test1$nb
[1] 0
> test1$nb.cor
[1] 0
> data(doubs)
> pca2 <- dudi.pca(doubs$mil,scannf=FALSE)
> test2 <- testdim(pca2)
> test2
class: krandtest 
Monte-Carlo tests
Call: testdim.pca(dudi = pca2)

Test number:   11 
Permutation number:   99 
      Test       Obs    Std.Obs   Alter Pvalue
1   Axis 1 0.9276029  8.1908092 greater   0.01
2   Axis 2 0.8765561  5.1610629 greater   0.01
3   Axis 3 0.8195407 -1.1308706 greater   0.90
4   Axis 4 0.7130800  0.8359168 greater   0.23
5   Axis 5 0.7621707  1.7325389 greater   0.05
6   Axis 6 0.7781209  2.3032324 greater   0.03
7   Axis 7 0.8316557  4.7990456 greater   0.01
8   Axis 8 0.9641126  4.4190248 greater   0.01
9   Axis 9 0.7978495  1.3429964 greater   0.13
10 Axis 10 0.9701563  7.7595169 greater   0.01
11 Axis 11 1.0000000  7.8768202 greater   0.01

other elements: nb nb.cor 
> test2$nb
[1] 2
> test2$nb.cor
[1] 2
> 
> 
> 
>

[Package ade4 Index]

`nb`	The estimated number of axes to keep
`nb.cor`	The number of axes to keep estimated using a sequential Bonferroni procedure