R package convoSPAT: Convolution-Based Nonstationary Spatial Modeling. Fits convolution-based nonstationary Gaussian process models to point-referenced spatial data. The nonstationary covariance function allows the user to specify the underlying correlation structure and which spatial dependence parameters should be allowed to vary over space: the anisotropy, nugget variance, and process variance. The parameters are estimated via maximum likelihood, using a local likelihood approach. Also provided are functions to fit stationary spatial models for comparison, calculate the kriging predictor and standard errors, and create various plots to visualize nonstationarity.
Keywords for this software
References in zbMATH (referenced in 4 articles , 1 standard article )
Showing results 1 to 4 of 4.
- Castro-Camilo, Daniela; Huser, Raphaël: Local likelihood estimation of complex tail dependence structures, applied to U.S. precipitation extremes (2020)
- Monterrubio-Gómez, Karla; Roininen, Lassi; Wade, Sara; Damoulas, Theodoros; Girolami, Mark: Posterior inference for sparse hierarchical non-stationary models (2020)
- Daniel Turek, Mark Risser: Bayesian nonstationary Gaussian process modeling: the BayesNSGP package for R (2019) arXiv
- Mark D. Risser, Catherine A. Calder: Local likelihood estimation for covariance functions with spatially-varying parameters: the convoSPAT package for R (2015) arXiv