dgpsi

The R package dgpsi provides R interface to Python package dgpsi for deep and linked Gaussian process emulations.

Hassle-free Python Setup
You don’t need prior knowledge of Python to start using the package, all you need is a single click in R (see Installation section below) that automatically installs and activates the required Python environment for you!

Features

dgpsi currently has following features:

• Gaussian process emulations with separable or non-separable squared exponential and Matérn-2.5 kernels.
• Deep Gaussian process emulations with flexible structures including:
  • multiple layers;
  • multiple GP nodes;
  • separable or non-separable squared exponential and Matérn-2.5 kernels;
  • global input connections;
  • non-Gaussian likelihoods (Poisson, Negative-Binomial, and heteroskedastic Gaussian).
• Linked emulations of feed-forward systems of computer models by linking (D)GP emulators of deterministic individual computer models.
• Fast Leave-One-Out (LOO) and Out-Of-Sample (OOS) validations for GP, DGP, and linked (D)GP emulators.
• Multi-core predictions and validations for GP, DGP, and Linked (D)GP emulators.
• Sequential designs for (D)GP emulators and bundles of (D)GP emulators.

Getting started

• Check A Quick Guide to dgpsi to get started with the package.
• For experimental features, check out our website for the development version.

Installation

You can install the package from CRAN:

install.packages('dgpsi')

or its development version from GitHub:

devtools::install_github('mingdeyu/dgpsi-R')

After the installation, run

library(dgpsi)
init_py()

to install and activate the required Python environment. That’s it, the package is now ready to use!

Note
Always run init_py() after library(dgpsi), telling R to invoke the required Python environment.

If you experience issues while running init_py(), please try to reinstall the Python environment:

dgpsi::init_py(reinstall = T)

or uninstall completely the Python environment:

dgpsi::init_py(uninstall = T)

And then restart the R and rerun:

library(dgpsi)
init_py()

References

Ming, D., Williamson, D., and Guillas, S. (2022) Deep Gaussian process emulation using stochastic imputation. Technometrics. 0(0), 1-12.

Ming, D. and Guillas, S. (2021) Linked Gaussian process emulation for systems of computer models using Matérn kernels and adaptive design, SIAM/ASA Journal on Uncertainty Quantification. 9(4), 1615-1642.