Bayesian Predictive Stacking for Scalable Geospatial Transfer Learning

Provides functions for Bayesian Predictive Stacking within the Bayesian transfer learning framework for geospatial artificial systems, as introduced in "Bayesian Transfer Learning for Artificially Intelligent Geospatial Systems: A Predictive Stacking Approach" (Presicce and Banerjee, 2025) . This methodology enables efficient Bayesian geostatistical modeling, utilizing predictive stacking to improve inference across spatial datasets. The core functions leverage 'C++' for high-performance computation, making the framework well-suited for large-scale spatial data analysis in parallel and distributed computing environments. Designed for scalability, it allows seamless application in computationally demanding scenarios.

spBPS

Overview

This package provides the principal functions to perform accelerated modeling for univariate and multivariate spatial regressions. The package is used mostly within the novel working paper "Bayesian Transfer Learning for Artificially Intelligent Geospatial Systems: A Predictive Stacking Approach" (Luca Presicce and Sudipto Banerjee, 2024+)". To guarantee the reproducibility of scientific results, in the Bayesian-Transfer-Learning-for-GeoAI repository are also available all the scripts of code used for simulations, data analysis, and results presented in the Manuscript and its Supplemental material.

Installation

If installing from CRAN, use the following.

install.packages("spBPS")

For a quick installation of the development version, run the following command in R. We use the devtools R package to install. Then, check for its presence on your device, otherwise install it:

if (!require(devtools)) {
  install.packages("devtools", dependencies = TRUE)
}

Once you have installed devtools, we can proceed. Let's install the spBPS package!

devtools::install_github("lucapresicce/spBPS")

Usage

Once successfully installed, load the library in R.

library(spBPS)

Cool! You are ready to start, now you too could perform fast & feasible Bayesian geostatistical modeling!

Contacts