WFP: Simulate Wright-Fisher Process

In reptalex/WrightFisher:

Description

Simulate Wright-Fisher Process

Usage

wfp(nspecies, rho = NULL, lambda = 15, X0 = rho, dt = 1e-05, Tmax = 1,
  nsamples = 1000, method = "VSM", tol = 1e-16)

Arguments

nspecies	One way to parameterize WFP by number of species. rho will be rep(1/nspecies,nspecies)
rho	Compositional vector signifying the mean of the community
lambda	Rate of migration or mean-reversion
X0	optional initial community state - must be a compositional vector of same size as rho
dt	size of timesteps for numerical integration
Tmax	Size of time window of numerical integration
nsamples	Number of timepoints between 0 and T to sample. Output will be a matrix with nrow=length(rho) and ncol=nsamples
method	Method for simulation, either 'VSM' or 'WFP'. Default is VSM - simulates volatility stabilized market and converts to relative abundances.
tol	Tolerance for VSM simulation, i.e. the reflecting lower bound for stock prices to ensure values stay positive.

Value

Outputs a two-element list. The first element is the time vector for the time of samples, and the second element is the matrix of the fluctuating community.

References

Washburne, Alex (2015) "Competition and Coexistence in an Unpredictable World". https://www.academia.edu/15517160/PhD_Thesis_-_Competition_and_Coexistence_in_an_Unpredictable_World

Examples

library(WrightFisher)
library(plotrix)

set.seed(1)
X <- wfp(nspecies=5,rho=rep(.2,5),dt=1e-3,Tmax=1)

time <- X$time
X <- X$Community
par(mfrow=c(1,1))
stackpoly(X,stack=T,xlab='time',ylab='Relative Abundance',main='Wright-Fisher Process Trajectory')

reptalex/WrightFisher documentation built on May 27, 2019, 5:54 a.m.