5. 🌌 Spatial Clustering

Spatial Constrained Clustering methods are used to divide an original set of n spatial units into p internally connected regions that maximize within similarity.

PostGeoDa offers two popular spatial constrained clustering algorithms: SKATER (Assunção et al. 2006) and REDCAP (Guo 2008; Guo and Wang 2011). SKATER stands for Spatial `K’luster Analysis by Tree Edge Removal, and obtains regionalization through a graph partitioning approach. REDCAP stands for REgionalization with Dynamically Constrained Agglomerative clustering and Partitioning, and consists of a family of six hierarchical regionalization methods.

Contents

• 5.1 SKATER

• 5.2 REDCAP