spatial_clusters: Identify and plot spatial clusters from point data

In reconhub/epimaps: Mapping Tools for Epidemics

Description

This function identifies spatial clusters from spatial coordinates (longitude / latitude) using Kmeans clustering. The resulting object is a spatial_clusters object which can be plotted with several options.

Usage

spatial_clusters(x, k)

## S3 method for class 'spatial_clusters'
print(x, ...)

## S3 method for class 'spatial_clusters'
plot(x, show_clusters = TRUE,
  show_points = FALSE, show_tiles = FALSE, show_density = FALSE,
  pt_col = "red", pt_alpha = 0.3, cl_col = "red", cl_alpha = 0.7,
  dens_alpha = 0.3, dens_bw = 0.005, leg_posi = c(0.1, 0.8), zoom = 13,
  ...)

Arguments

x	A matrix oar data.frame with two columns indicating longitude and latitude.
k	An integer indicating the number of clusters to be found.q
show_clusters	A logical indicating if clusters should be plotted.
show_points	A logical indicating if individual points should be plotted.
show_tiles	A logical indicating if Voronoi tesselation should be plotted.
show_density	A logical indicating if case density should be plotted.
pt_col	The color to be used for individual points.
pt_alpha	The transparency level to be used for individual points.
cl_col	The color to be used for clusters.
cl_alpha	The transparency level to be used for clusters.
dens_alpha	The transparency level to be used for case density.
dens_bw	The bandwidth to be used for case density estimation; larger values correspond to a wider kernel.
leg_posi	The (x,y) position of the legend, on a relative scale from 0 to 1.
zoom	The zoom level to be used for ggmap.
x	A spatial_clusters object.

Value

The function returns a spatial_cluster object, which is a list with the following items:

• lonlat: a matrix of lon/lat for individual points

• clusters_coords: a matrix of lon/lat/size for clusters

• clusters: cluster membership for the point locations

• voronoi: voronoi tesselation for the clusters

Author(s)

Thibaut Jombart ([e-mail])

Examples

## Not run: 
if (require("outbreaks")) {

lonlat <- ebola_sim$linelist[, c(10, 11)]
head(lonlat)
nrow(lonlat)


res <- spatial_clusters(lonlat, 30)
res
plot(res)

plot(res, show_density = TRUE, show_tiles = TRUE,
     cl_alpha = .2, cl_col = "blue")

}

## End(Not run)

reconhub/epimaps documentation built on May 7, 2019, 1:30 p.m.