expoTree (version 1.0.1)

runExpoTree: Density-dependent likelihood caluculation

Description

Calculates the density dependent likelihood of a phylogenetic tree. It takes branching and sampling times as an argument and integrates the likelihoood function over the whole tree.

Usage

runExpoTree(pars,times,ttypes,survival=TRUE,shifts=NULL,vflag=0, return.full=FALSE,rescale=TRUE,root.lineages=0)

Arguments

pars

Matrix of parameters. If the number of columns is larger than 5, only the first 5 columns are used. Each row is of the form c(N,beta,mu,psi,rho). The rows correspond to rate shifts.

times

Vector of event times.

ttypes

Vector of event types.

survival

Condition on the likelihood of observing the tree.

shifts

Times of rate shifts.

vflag

Verbosity level.

return.full

Return full probability vector rather than just the entry for I=1.

rescale

Rescale probability vector in the calculation. Avoids some numerical issues.

root.lineages

Number of lineages that are extant at the 'root' of the tree.

Value

Log-Likelihood or vector of log-likelihoods.

Details

Parameter matrix:

N = Total population size. Must be at least the maximal number of extant lineages at any given time.

beta = Branching rate. For epidemic trees, this is the infection rate. For species trees, this is the speciation rate.

mu = Extinction rate. For epidemic trees, this is the death/recovery rate. For species trees, this is the extinction rate.

psi = Sampling rate.

rho = Initial sampling rate.

References

Leventhal, Guenthard, Bonhoeffer & Stadler, 2013

See Also

expoTree

Examples

times <- cumsum(runif(10))
ttypes <- rep(1,10)

N <- 15
beta <- 1
mu <- 0.1
psi <- 0
rho <- 1
pars <- matrix(c(N,beta,mu,psi,rho),nrow=1)
lik <- runExpoTree(pars,times,ttypes)