probmap: Probability mapping for rates

In spdep: Spatial Dependence: Weighting Schemes, Statistics and Models

Description

The function returns a data frame of rates for counts in populations at risk with crude rates, expected counts of cases, relative risks, and Poisson probabilities.

Usage

probmap(n, x, row.names=NULL, alternative="less")

Arguments

n	a numeric vector of counts of cases
x	a numeric vector of populations at risk
row.names	row names passed through to output data frame
alternative	default “less”, may be set to “greater”

Details

The function returns a data frame, from which rates may be mapped after class intervals have been chosen. The class intervals used in the examples are mostly taken from the referenced source.

Value

raw	raw (crude) rates
expCount	expected counts of cases assuming global rate
relRisk	relative risks: ratio of observed and expected counts of cases multiplied by 100
pmap	Poisson probability map values: probablility of getting a more “extreme” count than actually observed - one-tailed, default alternative observed “less” than expected

Author(s)

Roger Bivand [e-mail]

References

Bailey T, Gatrell A (1995) Interactive Spatial Data Analysis, Harlow: Longman, pp. 300–303.

See Also

EBest, EBlocal, ppois

Examples

example(auckland)
res <- probmap(auckland$M77_85, 9*auckland$Und5_81)
rt <- sum(auckland$M77_85)/sum(9*auckland$Und5_81)
ppois_pmap <- numeric(length(auckland$Und5_81))
for (i in seq(along=ppois_pmap)) {
ppois_pmap[i] <- poisson.test(auckland$M77_85[i], r=rt,
  T=(9*auckland$Und5_81[i]), alternative="less")$p.value
}
all.equal(ppois_pmap, res$pmap)
brks <- c(-Inf,2,2.5,3,3.5,Inf)
cols <- grey(6:2/7)
plot(auckland, col=cols[findInterval(res$raw*1000, brks, all.inside=TRUE)])
legend("bottomleft", fill=cols, legend=leglabs(brks), bty="n")
title(main="Crude (raw) estimates of infant mortality per 1000 per year")
brks <- c(-Inf,47,83,118,154,190,Inf)
cols <- cm.colors(6)
plot(auckland, col=cols[findInterval(res$relRisk, brks, all.inside=TRUE)])
legend("bottomleft", fill=cols, legend=leglabs(brks), bty="n")
title(main="Standardised mortality ratios for Auckland child deaths")
brks <- c(0,0.05,0.1,0.2,0.8,0.9,0.95,1)
cols <- cm.colors(7)
plot(auckland, col=cols[findInterval(res$pmap, brks, all.inside=TRUE)])
legend("bottomleft", fill=cols, legend=leglabs(brks), bty="n")
title(main="Poisson probabilities for Auckland child mortality")

Example output

Loading required package: sp
Loading required package: Matrix

acklnd> require(maptools)
Loading required package: maptools
Checking rgeos availability: TRUE

acklnd> auckland <- readShapePoly(system.file("etc/shapes/auckland.shp",
acklnd+  package="spdep")[1])

acklnd> auckland.nb <- poly2nb(auckland)
Warning message:
use rgdal::readOGR or sf::st_read 
[1] TRUE

spdep documentation built on Aug. 19, 2017, 3:01 a.m.