seir_model_fit: SEIR model fitting method to a multiple outbreak model

In sempwn/cr0eso: Estimation of R0 For Multiple Outbreaks

SEIR model fitting method to a multiple outbreak model

Description

Create an instance of the hierarchical SEIR Stan model incorporating various data elements and sample model.

Usage

seir_model_fit(
  stan_model = NULL,
  tmax,
  n_outbreaks,
  outbreak_cases,
  outbreak_sizes,
  intervention_switch = TRUE,
  multilevel_intervention = FALSE,
  priors = prior_list,
  chains = 4,
  iter = 600,
  seed = 42,
  fit_type = "NUTS",
  data_model = "poisson"
)

Arguments

stan_model	[rstan] model object
tmax	Total number of time-points in observation
n_outbreaks	Total number of outbreaks
outbreak_cases	Number of daily reported cases by outbreak
outbreak_sizes	The total size of each facility (initial number of susceptible and exposed)
intervention_switch	Describes whether interventions occur in data (default TRUE)
multilevel_intervention	Describes whether intervention occurs
chains	Number of chains to sample
iter	number of iterations of MCMC
seed	The seed for random number generation. Set to replicate results.
fit_type	string "NUTS" or "VB" (VB quicker but less accurate).
data_model	string "poisson" or "negative_binomial". If negative_binomial selected then uses phi prior to control for overdispersion
prior_list	List of priors. See [prior_list]

Value

An object of class 'stanfit' returned by sampling

Author(s)

Mike Irvine

Examples

stan_mod <- rstan::stan_model(system.file("stan",
  "hierarchical_SEIR_incidence_model.stan", package = "cr0eso"))
tmax <- 5
pop_size <- 100
dim(pop_size) <- c(1)
example_incidence <- matrix(c(1,1,2,3,2),ncol=1)
fit <- seir_model_fit(stan_model = stan_mod, tmax,1,example_incidence,pop_size)

sempwn/cr0eso documentation built on Aug. 21, 2022, 1:35 a.m.