# compar.lynch

From ape v3.0-2
0th

Percentile

##### Lynch's Comparative Method

This function computes the heritable additive value and the residual deviation for continous characters, taking into account the phylogenetic relationships among species, following the comparative method described in Lynch (1991).

Keywords
regression
##### Usage
compar.lynch(x, G, eps = 1e-4)
##### Details

The parameter estimates are computed following the EM (expectation-maximization) algorithm. This algorithm usually leads to convergence but may lead to local optima of the likelihood function. It is recommended to run several times the function in order to detect these potential local optima. The optimal' value for eps depends actually on the range of the data and may be changed by the user in order to check the stability of the parameter estimates. Convergence occurs when the differences between two successive iterations of the EM algorithm leads to differences between both residual and additive values less than or equal to eps.

##### Value

A list with the following components:vareestimated residual variance-covariance matrix.varaestimated additive effect variance covariance matrix.uestimates of the phylogeny-wide means.Aaddtitive value estimates.Eresidual values estimates.liklogarithm of the likelihood for the entire set of observed taxon-specific mean.

##### Note

The present function does not perform the estimation of ancestral phentoypes as proposed by Lynch (1991). This will be implemented in a future version.

##### References

Lynch, M. (1991) Methods for the analysis of comparative data in evolutionary biology. Evolution, 45, 1065--1080.

pic, compar.gee

##### Examples
### The example in Lynch (1991)
cat("((((Homo:0.21,Pongo:0.21):0.28,",
"Macaca:0.49):0.13,Ateles:0.62):0.38,Galago:1.00);",
file = "ex.tre", sep = "")
G = vcv.phylo(tree.primates, cor = TRUE))`