simulatetsir {tsiR} | R Documentation |
simulatetsir
Description
This function just simulates the forward prediction given the data and a parms list generated from estpars or mcmcestpars.
Usage
simulatetsir(
data,
nsim = 100,
IP = 2,
parms,
method = "deterministic",
epidemics = "cont",
pred = "forward",
threshold = 1,
inits.fit = FALSE,
add.noise.sd = 0,
mul.noise.sd = 0
)
Arguments
data |
The data frame containing cases and interpolated births and populations. |
nsim |
The number of simulations to do. Defaults to 100. |
IP |
The infectious period. Defaults to 2. |
parms |
Either the parameters estimated by estpars or mcmcestpars, or a list containing beta, rho, Z, sbar, alpha, X, Y, Yhat, contact, alphalow, alphahigh, loglik, pop vectors. |
method |
The type of next step prediction used. Options are 'negbin' for negative binomial, 'pois' for poisson distribution, and 'deterministic'. Defaults to 'deterministic'. |
epidemics |
The type of data splitting. Options are 'cont' which doesn't split the data up at all, and 'break' which breaks the epidemics up if there are a lot of zeros. Defaults to 'cont'. |
pred |
The type of prediction used. Options are 'forward' and 'step-ahead'. Defaults to 'forward'. |
threshold |
The cut off for a new epidemic if epidemics = 'break'. Defaults to 1. |
inits.fit |
Whether or not to fit initial conditions using simple least squares as well. Defaults to FALSE. This parameter is more necessary in more chaotic locations. |
add.noise.sd |
The sd for additive noise, defaults to zero. |
mul.noise.sd |
The sd for multiplicative noise, defaults to zero. |