Fit a stochastic epidemic model using the linear noise approximation or ordinary differential equations to approximate the latent epidemic process.

fit_stem(
  stem_object,
  method,
  mcmc_kern,
  iterations,
  initialization_attempts = 500,
  ess_warmup = 50,
  thinning_interval = 100,
  return_adapt_rec = FALSE,
  return_ess_rec = FALSE,
  print_progress = 0,
  status_filename = NULL
)

Arguments

stem_object: a stochastic epidemic model object containing the dataset, model dynamics, and measurement process.
method: either "lna" or "ode".
mcmc_kern: MCMC transition kernel generated by a call to the mcmc_kernel function.
iterations: number of iterations
initialization_attempts: number of initialization attempts
ess_warmup: number of preliminary ESS iterations for the LNA, initial conditions, and time varying parameters prior to starting MCMC
thinning_interval: thinning interval for posterior samples, defaults to saving every 100th sample
return_adapt_rec: should the MCMC samples be returned during adaptation? defaults to FALSE.
return_ess_rec: should elliptical slice sampling steps and angles be returned? defaults to FALSE
print_progress: interval at which to print progress to a text file. If 0 (default) progress is not printed.
status_filename: string to pre-append to status files, defaults to LNA or ODE depending on the method used.

Value

list with posterior samples for the parameters and the latent process, along with MCMC diagnostics.