run_DAMCMC: Run the DA-MCMC

In rmorsomme/PDSIR: Data-augmentation Marko chain Monte Carlo for Fitting the Stochastic SIR Model to Incidence Counts

Run the DA-MCMC

Description

Data-augmentation Markov chain Monte Carlo to fit the stochastic SIR model to discretely observed incidence counts with the PD-SIR algorithm.

Usage

run_DAMCMC(
  Y,
  N = 10000,
  rho = 1,
  param = "bg",
  approx = "ldp",
  iota_dist = "exponential",
  gener = FALSE,
  b = 1/2,
  thin = 1,
  theta_0,
  print_i = FALSE,
  save_SS = FALSE,
  par_prior = list(a_beta = 0.01, b_beta = 1, a_gamma = 1, b_gamma = 1, a_R0 = 2, b_R0 =
    2, a_lambda = 1, b_lambda = 1)
)

Arguments

Y	observed data
N	number of iterations of the Markov chain
rho	proportion of the latent data updated each iteration
param	c("bg", "bR"); parameterize the models in terms of (beta, gamma) or (beta, R0)
approx	c("poisson", "ldp"); whether to approximate the distribution of the infection times with a poisson process or a linear death process
iota_dist	c("exponential", "weibull"); distribution of the infection period
gener	logical; whether to use the generalized SIR of Severo (1972)
b	parameter of the generalized SIR
thin	thinning parameter for the Markov chain
theta_0	initial value for the parameters
print_i	logical; whether to print the iteration
save_SS	logical; whether to save the sufficient statistics generated each iteration
par_prior	named list; parameters of the prior distributions

Value

list with the draws for the parameters, the log likelihood, acceptance rate and size of initial susceptible population

rmorsomme/PDSIR documentation built on April 27, 2023, 2:56 p.m.