Fit the VR reporting model — vrc

Fit the hierarchical VR mortality model with an explicit reporting process. The model is implemented in Stan and is fit using rstan.

Usage

vrc_fit(
  data,
  t0,
  mortality_covariates = NULL,
  reporting_covariates = NULL,
  mortality_conflict = c("fixed", "region"),
  reporting_conflict = c("fixed", "region"),
  mortality_time = c("national", "region"),
  reporting_time = c("national", "region"),
  standardise = TRUE,
  scale_binary = FALSE,
  drop_na_y = TRUE,
  duplicates = c("error", "sum"),
  algorithm = c("sampling", "meanfield", "fullrank"),
  priors = NULL,
  age_drop_cause = NULL,
  prior_PD = FALSE,
  backend = c("rstan"),
  stan_model = "vr_reporting_model",
  ...
)

Arguments

data: A data.frame in long format (see vrc_standata()).
t0: Conflict start time. See vrc_standata().
mortality_covariates: Optional formula for additional mortality covariates, excluding conflict.
reporting_covariates: Optional formula for additional reporting covariates, excluding conflict.
mortality_conflict: How to model the conflict effect in the mortality component. Use "fixed" (default) for one effect shared across regions, or "region" for partial pooling (random slopes) by region.
reporting_conflict: How to model the conflict effect in the reporting component. Use "fixed" (default) for one effect shared across regions, or "region" for partial pooling (random slopes) by region.
mortality_time: How to model time variation in the mortality component. "national" uses a single random walk shared across regions; "region" adds region-specific random walk deviations around the national trend.
reporting_time: How to model time variation in the reporting component. "national" uses a single random walk shared across regions; "region" adds region-specific random walk deviations around the national trend.
standardise: Logical. If TRUE (recommended), standardise conflict and covariate columns in model matrices.
scale_binary: Logical. If TRUE, also standardise binary dummy columns.
drop_na_y: Logical. If TRUE, rows with missing y are removed before fitting.
duplicates: How to handle duplicate identifier rows. One of "error" (default) or "sum".
algorithm: One of "sampling", "meanfield", or "fullrank" (passed to rstan).
priors: Optional prior specification created by vrc_priors(). If NULL (default), uses vrc_priors() with package defaults.
age_drop_cause: Which cause should the age reporting breakdown occur in. Defaults to NULL which uses 2nd cause or the first if there is only one.
prior_PD: Logical. If TRUE, samples from the prior distribution (no likelihood).
backend: Backend to use. Currently only "rstan" is supported.
stan_model: Stan model name. Defaults to "vr_reporting_model".
...: Additional arguments passed to rstan::sampling or rstan::vb.

Value

An object of class vrcfit.