Fit a VR mortality + reporting model

High-level wrapper around vrc_fit() modelled after the user experience of packages like epidemia and rstanarm. You specify a mortality component and a reporting component, and vrcm() handles data preparation and model fitting.

The fitted model separates:

a latent mortality process (lambda), and
a reporting completeness process (rho).

Both processes include a conflict proxy by default.

Usage

vrcm(
  mortality = vrc_mortality(~1),
  reporting = vrc_reporting(~1),
  data,
  t0,
  priors = NULL,
  ...
)

Arguments

mortality: A vrc_mortality() specification, or a formula.
reporting: A vrc_reporting() specification, or a formula.
data: A data.frame in canonical VR long format (see vrc_standata()).
t0: Conflict start time (index or a value in data$time).
priors: Optional prior specification created by vrc_priors(). If NULL (default), uses vrc_priors() with package defaults.
...: Passed through to vrc_fit().

Value

A vrcfit object, or (if chains = 0) a standata bundle returned by vrc_standata().