PLAN105 Lecture Notes - Lecture 8: Spatial Analysis, Contiguity, Nomothetic
9 May 2019
School
Department
Course
Professor
Document Summary
The built environment is an example of proximal or distal: proximal obesity, distal - environment, the following figure shows nine areas with their identity number. Moran"s i scatter plot to detect spatial dependence: quantify spatial structure/autocorrelation/dependence, positive values, positive spatial autocorrelation, negative values, negative spatial autocorrelation, near to zero, no spatial autocorrelation. Defining neighbours: for each area (vector or raster dataset), which areas are its neighbours, measure of similarity between neighbouring areas using spatial statistics, which take into account the neighbourhood structure of the dataset. Defining neighbours: adjacency-based: neighbours may be defined based on adjacency or contiguity, ways to define adjacency/contiguity, share a border (common boundary line) rook-based contiguity, share a vertex (common bound point) queen-based contiguity. Difference between (global) moran"s i and local moran"s i despite same input. Local moran"s i results: cluster map + significance map: lisa: local indicators of spatial association, showing locations of clusters with their p-values.
