Computing with Maps: A Gentle Introduction to Spatial Statistics in R

As a technical problem, map making has never been easier. As an approach to visualization or even data analysis, however, the story rarely ends with attaching “placemarks,” specifying paths or defining regions over some territory. Instead, when we observe or explore patterns in spatial data (or any data for that matter), we are drawn somewhat naturally to questions of inference: Is there really a pattern here? Do the data provide support for a particular “theory” or explanation for the pattern? “Spatial statistics,” as a term, refers both to a collection of numerical summaries for representing and comparing patterns, as well as a set of inferential tools; and it includes a range of problems including smoothing and interpolation, the estimation of spatial autocorrelations, and the analysis of spatial point patterns (to borrow Ripley’s well-established categories). In our session, we will present an introduction to spatial statistics using R, with the goal of providing new tools for both application developers as well as suggestions for facilitating data analysis by end-users. R is both a language and an environment for statistical computing and graphics and offers a number of packages for representing, visualizing and modeling spatial data. We will assume no familiarity with R or statistics (spatial or otherwise), emphasizing computation and visualization.

Please download and install R prior to the workshop www.r-project.org. If you have difficulty, please send email to [email protected].