geoModelSources: Extract latitude and longitude of points identified as...

In bobverity/Rgeoprofile: Geographic profiling in R

Description

This function takes the output of geoMCMC() and, for each 'crime', extracts the group to which it is assigned with the highest probability. For each group, the model returns the mean lat/long of all crimes assigned to that group.

Usage

geoModelSources(mcmc, data)

Arguments

mcmc	Model output in the format produced by geoMCMC().
data	Crime site data, in the format produced by geoData().

Examples

# simulated data
sim <-rDPM(50, priorMean_longitude = -0.04217491, priorMean_latitude = 
51.5235505, alpha=10, sigma=1, tau=3)
d <- geoData(sim$longitude, sim $latitude)
s <- geoDataSource(sim$source_lon, sim$source_lat)
p <- geoParams(data = d, sigma_mean = 1.0, sigma_squared_shape = 2)
m <- geoMCMC(data = d, params = p)
# extract sources identified by the model
ms <- geoModelSources(mcmc = m, data = d)
# plot data showing the sources identified by the model (note: NOT the actual suspect sites)
geoPlotMap(data = d, source = ms, params = p, breakPercent = seq(0, 10, 1), 
                  mapType = "roadmap", surfaceCols =c("red", "orange","yellow","white"),
                  crimeCol = "black", crimeCex = 2, sourceCol = "red", sourceCex = 2,
                  surface = m$geoProfile, gpLegend=TRUE, opacity = 0.4)

bobverity/Rgeoprofile documentation built on May 12, 2019, 11:28 p.m.