Monte-Carlo test on a Co-inertia analysis (in C).

Performs a Monte-Carlo test on a Co-inertia analysis.

Usage

# S3 method for class 'coinertia'
randtest(xtest, nrepet = 999, fixed=0, ...)

Arguments

xtest: an object of class coinertia
nrepet: the number of permutations
fixed: when non uniform row weights are used in the coinertia analysis, this parameter must be the number of the table that should be kept fixed in the permutations
...: further arguments passed to or from other methods

Value

a list of the class randtest

References

Dolédec, S. and Chessel, D. (1994) Co-inertia analysis: an alternative method for studying species-environment relationships. Freshwater Biology, 31, 277–294.

Author

Jean Thioulouse Jean.Thioulouse@univ-lyon1.fr modified by Stéphane Dray stephane.dray@univ-lyon1.fr

Note

A testing procedure based on the total coinertia of the analysis is available by the function randtest.coinertia. The function allows to deal with various analyses for the two tables. The test is based on random permutations of the rows of the two tables. If the row weights are not uniform, mean and variances are recomputed for each permutation (PCA); for MCA, tables are recentred and column weights are recomputed. If weights are computed using the data contained in one table (e.g. COA), you must fix this table and permute only the rows of the other table. The case of decentred PCA (PCA where centers are entered by the user) is not yet implemented. If you want to use the testing procedure for this case, you must firstly center the table and then perform a non-centered PCA on the modified table. The case where one table is treated by hill-smith analysis (mix of quantitative and qualitative variables) will be soon implemented.

Examples

data(doubs)
dudi1 <- dudi.pca(doubs$env, scale = TRUE, scan = FALSE, nf = 3)
dudi2 <- dudi.pca(doubs$fish, scale = FALSE, scan = FALSE, nf = 2)
coin1 <- coinertia(dudi1,dudi2, scan = FALSE, nf = 2)
plot(randtest(coin1))