The phylogenetic ANOVA

Share:

Description

This function performs the variance analysis of a trait on eigenvectors associated to a phylogenetic tree.

Usage

1
2
variance.phylog(phylog, z, bynames = TRUE,
 na.action = c("fail", "mean"))

Arguments

phylog

: an object of class phylog

z

: a numeric vector of the values corresponding to the variable

bynames

: if TRUE checks if z labels are the same as phylog leaves label, possibly in a different order. If FALSE the check is not made and z labels must be in the same order than phylog leaves label

na.action

: if 'fail' stops the execution of the current expression when z contains any missing value. If 'mean' replaces any missing values by mean(z)

Details

phylog$Amat defines a set of orthonormal vectors associated the each nodes of the phylogenetic tree.
phylog$Adim defines the dimension of the subspace A defined by the first phylog$Adim vectors of phylog$Amat that corresponds to phylogenetic inertia.
variance.phylog performs the linear regression of z on A.

Value

Returns a list containing

lm

: an object of class lm that corresponds to the linear regression of z on A.

anova

: an object of class anova that corresponds to the anova of the precedent model.

smry

: an object of class anova that is a summary of the precedent object.

Author(s)

Sébastien Ollier sebastien.ollier@u-psud.fr
Daniel Chessel

References

Grafen, A. (1989) The phylogenetic regression. Philosophical Transactions of the Royal Society London B, 326, 119–156.

Diniz-Filho, J. A. F., Sant'Ana, C.E.R. and Bini, L.M. (1998) An eigenvector method for estimating phylogenetic inertia. Evolution, 52, 1247–1262.

See Also

phylog, lm

Examples

1
2
3
4
5
6
7
8
data(njplot)
njplot.phy <- newick2phylog(njplot$tre)
variance.phylog(njplot.phy,njplot$tauxcg)
par(mfrow = c(1,2))
table.phylog(njplot.phy$Ascores, njplot.phy, clabel.row = 0,
 clabel.col = 0.1, clabel.nod = 0.6, csize = 1)
dotchart.phylog(njplot.phy, njplot$tauxcg, clabel.nodes = 0.6)
orthogram(njplot$tauxcg, njplot.phy$Ascores)

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.