create_simulator: Create surveillance data simulator.

Description

create_simulator creates a pomp object that will run simulations of an SIR or SIS model according to Gillespie's direct method and generate simulated observations of the process.

Usage

create_simulator(times = seq(0, 9), t0 = min(times), process_model = c("SIR", "SIS"), params = c(gamma = 24, mu = 1/70, d = 1/70, eta = 1e-05, beta = 1e-04, rho = 0.1, S_0 = 1, I_0 = 0, R_0 = 0, N_0 = 1e+05), covar = data.frame(gamma_t = c(0, 0), mu_t = c(0, 0), d_t = c(0, 0), eta_t = c(0, 0), beta_t = c(0, 0), time = c(0, 1e+06)))

Arguments

times

A numeric vector of increasing times at which the state of the simulation will be sampled.

The time at which the simulation is started with state variable set to the initial conditions specified via params.

process_model

Character string giving the process model. Allowed values are '"SIR"' and '"SIS"'.

params

A named numeric vector of parameter values and initial conditions.

covar

A data frame containing values of the time-dependent components of the parameters.

Value

A pomp object with which simulations can be run via pomp::simulate.

Details

See the vignette "Getting Started with spaero" for a description of the model. The "params" argument must include all model parameters. These will become the default parameters for the model object. They can be overriden when the simulation is run via the "params" argument of pomp::simulate. The case is the same for the "times" argument. The "covar" argument should be a data frame with a column named for each of the time-dependent parameters and a column named time. This data frame describes the time series of each of the time-dependent parameters. In the simulation, interpolation based on this data frame determines the value of these parameters at specific instants in time. The user must ensure that these values result in the parameters remaining non-negative for the course of the simulation.

Examples

Run this code


foo <- create_simulator()
out <- pomp::simulate(foo, times=seq(0, 20, by=1/26))
out <- as(out, "data.frame")
head(out)

opar <- par(mfrow=c(2, 1))
plot((S/N)~time, data=out, type="l")
plot(cases~time, data=out, type="l")
par(opar)