R Packages for Analyzing Spatial Data: A Comparative Case Study with Areal Data

Abstract

The count of open source software packages hosted by the Comprehensive R Archive Network (CRAN) using key spatial data handling packages has now passed 1,000. Providing a comprehensive review of these packages is beyond the scope of an article. Consequently, this review takes the form of a comparative case study, reproducing some of the approach and workflow of a spatial analysis of a data set including almost all the census tracts in the coterminous United States. The case study moves from visualization and the construction of a spatial weights matrix, to exploratory spatial data analysis and spatial regression. For comparison, implementations of the same steps in PySAL and GeoDa are interwoven, and points of convergence and divergence noted and discussed. Conclusions are drawn about the usefulness of open source software, the significance of sharing contributions both in software implementation but also more broadly in reproducible research, and in opportunities for exchanging ideas and solutions with other research domains.