<b>spNNGP</b> <i>R</i> Package for Nearest Neighbor Gaussian Process Models

This paper describes and illustrates functionality of the spNNGP R package. The package provides a suite spatial regression models for Gaussian non-Gaussian pointreferenced outcomes that are spatially indexed. implements several Markov chain Monte Carlo (MCMC) MCMC-free nearest neighbor process (NNGP) inference about large data. Non-Gaussian modeled using NNGP Pólya-Gamma latent variable. OpenMP parallelization options provided to take advantage multiprocessor systems. Package features illustrated simulated real data sets.