MetStaT (version 1.0)

ASCA.DoPermutationTest: Permutation test for ASCA

Description

Does a permutation test on the results from ASCA.Calculate by repeating the ASCA analysis many times with permutated samples.

Usage

ASCA.DoPermutationTest(asca, perm = 1000)

Arguments

asca

a previously done ASCA analysis should be supplied.

perm

the number of permutations to be performed.

Value

An array is returned that contains the p-value per factor or interaction of the ASCA.

Details

The significance of treatment effects or of interactions between treatment effects can be evaluated by considering the p-values that are returned by ASCA.DoPermutationTest. The p-values are determined by the fraction permutations that have a larger value for the test statistic than the test statistic of the data matrix. The test statistic used is the sum of squares of the treatment level averages.

References

Gooitzen Zwanenburg, Huub C.J. Hoefsloot, Johan A. Westerhuis, Jeroen J. Jansen and Age K. Smilde, ANOVA principal component analysis and ANOVA simultaneous component analysis: a comparison. J Chemometrics, 25, (2011), p. 561 - 567

MARTI J. ANDERSON, and CAJO J. F. TER BRAAK, PERMUTATION TESTS FOR MULTI-FACTORIAL ANALYSIS OF VARIANCE. Journal of Statistical Computation and Simulation, 73(2), (2003) p. 85 - 113

Examples

## Do ASCA on all (both) factors and the interaction between the two factors
data(ASCAdata)
ASCA <- ASCA.Calculate(ASCAX, ASCAF, equation.elements = "1,2,12", scaling = TRUE)

## Do a permutation test to evaluate the significance to the two factors and the interaction.
ASCA.DoPermutationTest(ASCA, perm=1000)