Gibbs Samplers for Discrete Bayesian Spatiotemporal Models
Add neighbors to adjacency information
Age-standardize model objects
Aggregate count arrays
Aggregate samples by non-age group
Create CAR model
Extract estimates from RSTr model object
Generate medians, credible intervals, and relative precisions
Load model
Load MCMC samples
Generate count data for RSTr object
tools:::Rd_package_title("RSTr")
Split sample groups
Age-standardize samples
Suppress estimates based on reliability criteria
Update model
Takes Poisson or Binomial discrete spatial data and runs a Gibbs sampler for a variety of Spatiotemporal Conditional Autoregressive (CAR) models. Includes measures to prevent estimate over-smoothing through a restriction of model informativeness for select models. Also provides tools to load output and get median estimates. Implements methods from Besag, York, and Mollié (1991) "Bayesian image restoration, with two applications in spatial statistics" <doi:10.1007/BF00116466>, Gelfand and Vounatsou (2003) "Proper multivariate conditional autoregressive models for spatial data analysis" <doi:10.1093/biostatistics/4.1.11>, Quick et al. (2017) "Multivariate spatiotemporal modeling of age-specific stroke mortality" <doi:10.1214/17-AOAS1068>, and Quick et al. (2021) "Evaluating the informativeness of the Besag-York-Mollié CAR model" <doi:10.1016/j.sste.2021.100420>.