LinearGillespie: Run linear Gillespie simulations. 'LinearGillespie'

In MorrisseyLab/CryptDriftR: Analyse crypt lineage tracing data

Description

Use this function to run gillespie simulations. The function is used internally and is not strictly necessary to use directly.

Usage

LinearGillespie(numSim, nu, a_mat, x0, time_vec)

Arguments

numSim	Number of times to run the simulation. The output is the sum of all sims.
nu	Matrix with a column per reaction and the same number of rows as states. Each column
a_mat	Two column matrix with the same number of rows as reactions. Each row contains first the rate of that reaction and second the
x0	The initial state of the system.
time_vec	A vector of time points at which to report the simulation status. Simulate a stochastic process with the Gillespie algorithm.

Value

A matrix with a column per data point and the same number of rows as states.

Examples

# This is the code used to simulate continous labelling
mu     = 1e-4
lambda = 0.1
N      = 5
Pr     = 0.5
time_points = 100:150
numSim = 1e7

beta  <- 2*lambda*Pr
alpha <- 2*lambda*(1-Pr)
#   Number of chain reactions
M      <- N
# Initial state vector
x0             <- c(numSim, rep(0,M)) 
names(x0)      <- paste("x",seq(M+1),sep="") 
nu             <- matrix(rep(0,(M*(M+1))),ncol=M)
diag(nu)       <- -1
diag(nu[2:M,]) <- +1
nu[M+1,M]      <- +1
nu             <- cbind(nu, -1*nu)
nu             <- nu[,-1*c(ncol(nu))]
a_mat          <- cbind(c(lambda*N*mu, rep(beta,M-1), rep(alpha,M-1)), c(seq(M),2:M))
sims <- LinearGillespie(numSim=1, nu, a_mat, x0, time_points) 

MorrisseyLab/CryptDriftR documentation built on Aug. 7, 2021, 8:22 p.m.