Generates synthetic incidence data if required, sets up MCMC parameters and priors, defines model comparison functions, and runs the MCMC process.
Usage
inf_run(
model,
param_inputs,
control_params,
params_to_estimate,
proposal_matrix,
adaptive_params,
start_values,
noise = FALSE,
seed = 24,
month_unequal_days = FALSE,
dates,
synthetic = TRUE,
incidence_df = NULL,
save_trajectories = TRUE,
rerun_n = Inf,
rerun_random = FALSE,
param_priors = NULL,
override_priors = NULL,
n_years_warmup = 3,
obs_config
)Arguments
- model
The model object used for simulation and inference.
- param_inputs
List of input parameters for the model, including temperature, decay, and coverage details.
- control_params
Named list of control parameters for MCMC (e.g., number of steps, burn-in, chains).
- params_to_estimate
Character vector of parameter names to be estimated.
- proposal_matrix
Covariance matrix used for the proposal distribution in MCMC.
- adaptive_params
List of parameters controlling adaptation in MCMC (e.g., update interval, scale factor).
- start_values
Named vector of initial values for MCMC chains.
- noise
Logical; if TRUE, adds noise to the synthetic incidence data.
- seed
Numeric; random seed for reproducibility of simulations.
- month_unequal_days
Logical; if TRUE, accounts for different numbers of days in each month.
- dates
Character or Date vector of length two specifying the start and end date of the simulation period.
- synthetic
Logical; if TRUE, synthetic data is generated using the model and parameters.
- incidence_df
Data frame of observed incidence data. Required if
synthetic = FALSE.- save_trajectories
Logical; if TRUE, saves intermediate state trajectories during MCMC.
- rerun_n
Numeric; frequency (in iterations) at which to re-run the deterministic model (default is Inf).
- rerun_random
Logical; if TRUE, re-run times are randomized instead of fixed interval.
- param_priors
Optional named list of prior distributions for parameters being estimated.
- override_priors
Optional list to override specific priors in
param_priors.- n_years_warmup
Number of years to run the model prior to the start date (default is 3).
- obs_config
A named list specifying configuration of the observation model. Use
make_obs_config()to construct this.