Preface

In contrast to the materials covered in Volume 1, this second volume has no precedent in an earlier workbook. Much of its contents have been added in recent years to the GeoDa documentation pages, as the topics were gradually included into my Introduction to Spatial Data Science course and implemented in GeoDa. At one point, the material became too much to constitute a single course and was split off into a separate Spatial Clustering course. The division of the content between the two volumes follows this organization.

In contrast to the first volume, where the focus is almost exclusively on data exploration, here attention switches to the delineation of groupings of observations, i.e., clusters. Both traditional and spatially constrained methods are considered. Again, the emphasis is on how a spatial perspective can contribute to additional insight, both by considering the spatial aspects explicitly (as in spatially constrained clustering) as well as through spatializing classic techniques.

Compared to Volume 1, the treatment is slightly more mathematical and familiarity with the methods covered in the first volume is assumed. As before, extensive references are provided. However, in contrast to the first volume, several methods included here are new and have not been treated extensively in earlier publications. They were typically introduced as part of the documentation of new features in GeoDa.

The empirical illustrations use the same sample data sets as in Volume 1. These are included in the software.

All applications are based on Version 1.22 of the software, available in Summer 2023. Later versions may include slight changes as well as additional features, but the treatment provided here should remain valid. The software is free, cross-platform and open source, and can be downloaded from https://geodacenter.github.io/download.html.