Spatial Misalignment: Interpolation, Linkage, and Estimation

Provides functions to estimate, predict and interpolate areal data. For estimation and prediction we assume areal data is an average of an underlying continuous spatial process as in Moraga et al. (2017) , Johnson et al. (2020) , and Wilson and Wakefield (2020) . The interpolation methodology is (mostly) based on Goodchild and Lam (1980, ISSN:01652273).

output: github_document

smile

Spatial Misalignment: Interpolation, Linkage, and Estimation

The smile package for R simplifies the analysis of spatial data. It offers tools to:

• Estimate and predicte: Using a model-based approach, smile treats areal data as averages derived from a continuous Gaussian random field. It is possible to model point-referenced and areal data jointly.

• Interpolate: For straightforward interpolation, smile employs a non-parametric method known as simple areal interpolation.

• Quantify uncertainty: When dealing with simple areal interpolation and knowing the variance of your observations, smile helps you quantify the uncertainty associated with the interpolation.

Essentially, smile provides a user-friendly way to explore, model, and interpolate spatial data with R, offering both model-based and non-parametric approaches.

The package accompanies a web page (powered by pkgdown) and 5 vignettes.

Vignettes

1. Converting sf to spm objects;
2. Fitting models and making predictions;
3. Areal Interpolation;
4. Method;
5. Spatial covariance functions;

Installation

To install the CRAN version of the package, use

install.packages("smile")

The installation of the development version from GitHub can be done via

remotes::install_github("lcgodoy/smile")
## or devtools::install_github("lcgodoy/smile")