Localized Ensemble of Approximate Gaussian Processes

An emulator designed for rapid sequential emulation (e.g., Markov chain Monte Carlo applications). Works via extension of the 'laGP' approach by Gramacy and Apley (2015 ). Details are given in Rumsey et al. (2023 ).

leapgp

This logo was designed by Imagine AI Art Studio

Description

leapgp is an R package designed for fast sequential emulation of computer models. Sequential emulation is common in applications requiring MCMC. The leapgp package is based on Rumsey et al. (2023) and can be viewed as a global model extension of the laGP method of Apley and Gramacy 2015.

To install the package, use

# install.packages("devtools")
devtools::install_github("knrumsey/leapgp")

The method is described in Rumsey et al. (2023) and explicit code examples are given in the vignette associated with this package.

References

Rumsey, Kellin N., Gabriel Huerta, and J. Derek Tucker. “A localized ensemble of approximate Gaussian processes for fast sequential emulation.” Stat (2023): e576.

Gramacy, Robert B., and Daniel W. Apley. “Local Gaussian process approximation for large computer experiments.” Journal of Computational and Graphical Statistics 24.2 (2015): 561-578.