galotalp/gpMCMC: Generate MCMC or FBI samples from the posterior distribution of Gaussian Process hyper-parameters

Gaussian Process posterior distributions are often difficult to sample from. We offer a series of MCMC sampling methods that use the Metropolis algorithm to sample from posterior distributions that use a Gaussian Correlation function and an Exponential prior. Essentially, our MCMC functions sample from the posterior distribution of the correlation hyperparameters in the gaussian correlation function.

install.packages("remotes")
remotes::install_github("galotalp/gpMCMC")