bilisa.perm: Bivariate Local Moran Statistic

In spdyn: Exploratory Space-Time Data Analysis

Bivariate Local Moran Statistic

Description

This function computes a permutation test to assess the statistical significance of the Local Moran Statistics.

Usage

bilisa.perm(x,y,listw,perm=999,FUN=mean,DESV=sd,...)

Arguments

x	A numeric vector
y	A numeric vector. This is the variable to be spatially lagged
listw	A listw object as produced by nb2listw or mat2listw. A list of spatial weights
perm	Number of permutations
FUN	By defautl FUN=mean. In general, a function to compute the central tendency of the data. It should return a scalar. See notes
DESV	Dispersion statistic. Defaults to sd
...	Further arguments passed to FUN

Details

This function implements a conditional permutation test to assess the statistical significance of the bivariate LISA. This is a distribution free test that overcomes the limitations of the standard assumption of normality.

Value

Returns a numeric matrix

Ii	Values of the Local Moran Statistics
p-val	Pseudo p-values
Zx	Standardized values of x
Zly	Standardized values of the spatial lag of y

Note

There is no theoretical nor simulation work assessing the behavior of the Local Moran Staistic when the measure of central tendency is different from the mean.

Author(s)

Osmar Leandro Loaiza Quintero

See Also

lisa.perm, plot.lisaPerm

Examples

data(usinc)

require(spdep)

lw<-nb2listw(
		poly2nb(usinc,queen=TRUE),style='W'
	)

my_bilisa <- bilisa.perm(x=usinc$X1929, y=usinc$X2009, listw=lw)


plot(my_bilisa, usinc)
title(sub='State per capita income in 1929 vs. spatially lagged state per capita income in 2009')

spdyn documentation built on Oct. 28, 2024, 3 p.m.