MoransI()

Summary

Computes statistics for Moran's I, a measure of spatial autocorrelation.

Added to Version 2017.

array = MoransI(string layer_bar_set, string field_spec, array options)

Argument	Contents
layer_bar_set	The layer\|set of an area layer
field_spec	The full field spec of the field of interest

Options	Type	Contents
Clustering	String	A full field spec for Adjacency clustering
Adjacency	String	The type of adjacency: "Edge" (default), "Node" or "Island"
Matrix	String	The adjacency value needed: "Adjacency" (default), "Adjacency Level", "Common Border Length" or "Inter-centroid Distance"
Search Distance	Number	A numeric value, used for Island Adjacency, in User Units (default= 1 mile)
Temp	String	A path for temporary files (default is the current temporary path)

Returns an options array with the Moran's I statistics:

Option	Contents
Observations	The number of cells
S1	The S1 statistic
S2	The S2 statistic
Sum of Weights	The sum of the spatial weights
Moran's I	The measure of spatial autocorrelation
Expected Value	The expected value, E(I) = -1 / (N - 1)
Std Error	The standard error
95% C.I. Upper	The upper bound of the 95% confidence interval
95% C.I. Lower	The lower bound of the 95% confidence interval

Moran's I is -1 for perfect dispersion (such as a checkerboard), 0 for a random arrangement, and close to 1 for high spatial autocorrelation.
The Clustering option is Group by the chosen field. For example, using the State field for Counties will compute adjacency based on their cluster (group) code.
In the Matrix option, "Common Border Length" is the sum of the lengths of the border in common between the two regions, and "Adjacency Level" is Common Border Length / Average Perimeter.

a = MoransI("County|", "County.Population",)

Error Type	Meaning
Error	Inputs are not valid or could not generate the Moran's I value