PCScoreCorrelation function

PC Score Correlation Test

Given a set of covariance matrices and means for terminals, test the hypothesis that observed divergence is larger/smaller than expected by drift alone using the correlation on principal component scores.

PCScoreCorrelation(
  means,
  cov.matrix,
  taxons = names(means),
  show.plots = FALSE
)

Arguments

• means: list or array of species means being compared. array must have means in the rows.
• cov.matrix: ancestral covariance matrix for all populations
• taxons: names of taxons being compared. Must be in the same order of the means.
• show.plots: Logical. If TRUE, plot of eigenvalues of ancestral matrix by between group variance is showed.

Returns

list of results containing:

correlation matrix of principal component scores and p.values for each correlation. Lower triangle of output are correlations, and upper triangle are p.values.

if show.plots is TRUE, also returns a list of plots of all projections of the nth PCs, where n is the number of taxons.

Examples

#Input can be an array with means in each row or a list of mean vectors
means = array(rnorm(40*10), c(10, 40)) 
cov.matrix = RandomMatrix(40, 1, 1, 10)
taxons = LETTERS[1:10]
PCScoreCorrelation(means, cov.matrix, taxons)

## Not run:

##Plots list can be displayed using plot_grid()
library(cowplot)
pc.score.output <- PCScoreCorrelation(means, cov.matrix, taxons, TRUE)
plot_grid(plotlist = pc.score.output$plots)
## End(Not run)

References

Marroig, G., and Cheverud, J. M. (2004). Did natural selection or genetic drift produce the cranial diversification of neotropical monkeys? The American Naturalist, 163(3), 417-428. doi:10.1086/381693

Author(s)

Ana Paula Assis, Diogo Melo