Objectives of this lecture
Spatial Autocorrelation Statistics¶
It is about the values of spatial features
What we have learnt: The patterns of statistical values and the patterns of locations.
What we have learnt: The intersection---location of values, values of locations.
Spatial Autocorrelation Statistics captures both attribute similarity and locational similarity
Attribute similarity:
how close are the values
measurements of (dis)similarities between values
squared differences:
absolute differences:
Locational similarity:
how close are the locations
relationship of neighborhood between locations
distance-based metrics (Euclidean, Manhattan)
adjacency-based spatial weights (rook, queen)
network-based spatial weights