spseg: Integrated Functions for Spatial Segregation Analysis

In spatialkernel: Nonparametric Estimation of Spatial Segregation in a Multivariate Point Process

Description

Spatial segregation analysis to be performed by a single function and presentations by associated plot functions.

Usage

  spseg(pts, marks, h, opt = 2, ntest = 100, poly = NULL, 
    delta = min(apply(apply(pts, 2, range), 2, diff))/100, 
	proc = TRUE)
  plotcv(obj, ...)
  plotphat(obj, types = unique(obj$marks), sup = TRUE, 
    col = risk.colors(10), 
	breaks = seq(0, 1, length = length(col) + 1), ...)
  plotmc(obj, types = unique(obj$marks), quan = c(0.05, 0.95), 
    sup = FALSE, col = risk.colors(10), 
	breaks = seq(0, 1, length = length(col) + 1), ...)

Arguments

pts	matrix containing the x,y-coordinates of the point locations at which type-specific probabilities estimated.
marks	numeric/character vector of the types of the point in the data.
h	numeric vector of the kernel smoothing bandwidth at which to calculate the cross-validated log-likelihood function.
opt	integer, 1 to select bandwidth; 2 to calculate type-specific probabilities; and 3 to do the Monte Carlo segregation test.
ntest	integer with default 100, number of simulations for the Monte Carlo test.
poly	matrix containing the x,y-coordinates of the polygonal boundary of the data.
delta	spacing distance of grid points at which to calculate the estimated type-specific probabilities for image plot.
proc	logical with default TRUE to print the processing message.
obj	list of the returning value of spseg
types	numeric/character types of the marks of data points to plot the estimated type-specific probabilities, default to plot all types.
sup	logical with default FALSE, if TRUE to superimpose data points on the estimated type-specific probability surface.
quan	numeric, the pointwise significance levels to add contours to image plot of the estimated type-specific probability surface, with default of c(0.05, 0.95).
col	list of colors such as that generated by risk.colors.
breaks	a set of breakpoints for the col: must give one more breakpoint than colour.
...	other arguments concerning plot and points

Details

spseg implements a complete spatial segregation analysis by selecting bandwidth, calculating the type-specific probabilities, and then carrying out the Monte Carlo test of spatial segregation and pointwise significance. Some plot functions are also provided here so that users can easily present the results.

These functions are provided only for the convenience of users. Users can instead use individual functions to implement the analysis step by step and plot the diagrams as they wish.

Examples of how to use spseg and present results using plot functions are presented in spatialkernel-package.

Value

spseg returns a list with components

hcv	bandwidth selected by the cross-validated log-likelihood function.
gridx,gridy	x, y coordinate vectors at which the grid points are generated at which to calculate the type-specific probabilities and pointwise segregation test p-value.
p	estimated type-specific probabilities at grid points generated by vectors gridx, gridy.
pvalue	p-value of the Monte Carlo spatial segregation test.
stpvalue	pointwise p-value of the Monte Carlo spatial segregation test.
...	copy of pts, marks, h, opt.

Note

Setting h to a unique value may force spseg to skip the selecting bandwidth step, go straight to calculate the type-specific probabilities and then test the spatial segregation with this fixed value of bandwidth.

See Also

cvloglk, phat, mcseg.test, pinpoly, risk.colors, and metre

spatialkernel documentation built on May 2, 2019, 4:37 p.m.