Handbook of Environmental and Ecological Statistics.

Abstract

models using MCMC the book turns to WinBUGS, a wellestablished piece of software, for the integrated nested Laplace approximation (INLA) uses R-INLA. In the years since the book’s publication, the computational toolbox has continued to expand. The sp package has been augmented with the sf package, which provides a simple and flexible interface that will be familiar to users of the tidyverse. Mapping in R has continued to improve through the development of tmap and leaflet. The probabilistic programming language Stan and the associated R package rstan allows more efficient Bayesian inferences using Hamiltonian Monte Carlo. These packages would be a welcome addition to a second edition of this book. Although the methodology presented in the book is well explained, the code to reproduce most of the approaches is not provided. This is an unfortunate omission, as it would have provided means for researchers and practitioners to apply the methods presented to their own investigations easily and unambiguously. It also would have aided instructors who wish to pull chapters from this book for use in an applied course. Overall, this book provides an overview of the rich variety of spatial epidemiology methods and their application. For readers coming from general statistics, the attention paid to issues such as misalignment, ecological bias, and the interpretation of results will be particularly useful. In the years since the publication of this handbook, the field has continued to evolve. A present survey of the discipline would include a discussion of new data sources, machine learning approaches, and modern software packages for spatial analysis, as well as a supplement containing the code necessary to apply the methods presented. With these additions, this handbook stands as a valuable resource for researchers and professionals interested in solving new challenges in spatial epidemiology and public health surveillance.