User Guide¶

esda covers a range of exploratory spatial data analysis methods. To help get you oriented, start with the question you are trying to answer, then follow the relevant section.

New to esda? Start with the Getting Started overview, which walks through the key concepts — spatial weights, spatial lag, autocorrelation statistics, and permutation inference — using a single dataset from beginning to end.

Guiding Questions¶

• Is there spatial autocorrelation? → Global Spatial Autocorrelation Methods: Moran’s I, Geary’s C, Getis-Ord G, Join Counts

• Where are clusters or outliers? → Local Spatial Autocorrelation Methods: Local Moran’s I (LISA), Local Geary, G_i*, LOSH, Local Join Counts, Multivariate Moran

• How does spatial dependence change across distance? → Spatial Pattern Diagnostics Methods: Spatial correlogram (distance bands, KNN, nonparametric)

• Where do point clusters form? → Spatial Clustering Methods: A-DBSCAN

• What are the shapes or geometry of features? → Shape and Geometry Analysis Methods: Shape compactness measures, geo-silhouettes

• How are units connected structurally? → Topology Methods: Isolation, Prominence