hlmBayes_sp: A spatial hierarchical Bayes Markov Chain Monte Carlo sampler

In arh926/spWombling: Bayesian Wombling for Spatial Processes

A spatial hierarchical Bayes Markov Chain Monte Carlo sampler

Description

Fits a univariate Gaussian spatial regression model: Y(s)=x(s)^T\beta+Z(s)+\epsilon. Parameters not listed are optional.

Usage

hlmBayes_sp(
  y = NULL,
  coords = NULL,
  niter = NULL,
  nburn = NULL,
  report = NULL,
  a.sigma = NULL,
  b.sigma = NULL,
  a.tau = NULL,
  b.tau = NULL,
  lower.phi = NULL,
  upper.phi = NULL,
  cov.type = NULL,
  verbose = TRUE,
  trgtFn.compute = FALSE,
  digits = 3
)

Arguments

y	observed response (order L x 1)
coords	coordinates for observed process (order L x 2)
niter	number of MCMC iterations
nburn	number of burn-in samples
report	batch length
a.sigma	shape parameter for inverse-gamma prior on \sigma^2
b.sigma	scale parameter for inverse-gamma prior on \sigma^2
a.tau	shape parameter for inverse-gamma prior on \tau^2
b.tau	scale parameter for inverse-gamma prior on \tau^2
lower.phi	lower limit for uniform prior on \phi
upper.phi	upper limit for uniform prior on \phi
cov.type	covariance type (three available choices: Gaussian, Mat\'ern(\nu=3/2)), Mat\'ern(\nu=5/2)
verbose	if true prints output for batches
trgtFn.compute	compute posterior
digits	rounding digits

arh926/spWombling documentation built on April 14, 2025, 4 p.m.