position.leverage {distory}

Position Leverage

Description

Provides a rough heuristic for determining the degree to which each position in the DNA matrix affects the resulting tree.

Usage

position.leverage(data, F, to = NULL, rep = 50, by = 1)

Arguments

data

A DNA matrix in DNAbin format.

F

A tree estimation function, accepting a DNA matrix of class DNAbin and returning a tree of class phylo.

to

The tree with which distances are measured in respect to, or NULL to indicate the tree estimated by F for the starting DNA matrix.

rep

The number of times to replicate the position in question.

by

The function will perform the calculation on every by-th position (that is, on seq(1, N, by)).

Details

This function takes a DNA matrix and, for every by-th position, replicates that position rep times, randomly removing rep other positions to keep all sequences the same length other positions to keep all sequences the same length. For each new DNA matrix created in this way, F is used to estimate the corresponding tree, and the distance to tree to is computed and stored. This distance can be thought of as somewhat analogous to the leverage of that position.

Value

Returns a numeric vector of distances from tree to for each position sampled.

Author(s)

John Chakerian

References

Chakerian, J. and Holmes, S. P. Computational Tools for Evaluating Phylogenetic and Heirarchical Clustering Trees. arXiv:1006.1015v1.

See Also

dist.multiPhylo

Examples

data(woodmouse)
f.est <- function(x) root(nj(dist.dna(x)), "No305", resolve.root = TRUE)
position.leverage(woodmouse, f.est, by = 10)