9. Maps

Like for other plots, there are different approaches in R to make maps. You can use “base plot” in many cases. Alternatively use levelplot, either via the spplot function (implemented in sp and raster) or via the rasterVis package.

Here are some brief examples about making maps. You can also look elsewhere on the Internet, like here, or this for spplot and rasterVis.

Vector data

Base plots

library(raster)
p <- shapefile(system.file("external/lux.shp", package="raster"))
plot(p)

image0

n <- length(p)
plot(p, col=rainbow(n))

image1

One colour per region (NAME_1)

u <- unique(p$NAME_1)
u
## [1] "Diekirch"     "Grevenmacher" "Luxembourg"
m <- match(p$NAME_1, u)
plot(p, col=rainbow(n)[m])
text(p, 'NAME_2', cex=.75, halo=TRUE)

image2

spplot

spplot(p, 'AREA')

image3

Raster

Example data

library(raster)
b <- brick(system.file("external/rlogo.grd", package="raster"))

r <- raster(p, res=0.01 )
values(r) <- 1:ncell(r)
r <- mask(r, p)

Several generic functions have been implemented for Raster* objects to create maps and other plot types. Use ‘plot’ to create a map of a Raster* object. When plot is used with a RasterLayer, it calls the function ‘rasterImage’ (but, by default, adds a legend; using code from fields::image.plot). It is also possible to directly call image. You can zoom in using ‘zoom’ and clicking on the map twice (to indicate where to zoom to). With click it is possible to interactively query a Raster* object by clicking once or several times on a map plot.

After plotting a RasterLayer you can add vector type spatial data (points, lines, polygons). You can do this with functions points, lines, polygons if you are using the basic R data structures or plot(object, add=TRUE) if you are using Spatial* objects as defined in the sp package. When plot is used with a multi-layer Raster* object, all layers are plotted (up to 16), unless the layers desired are indicated with an additional argument.

plot(r)
plot(p, add=TRUE)

image4

image does not provide a legend and that can be advantageous in some cases.

image(r)
plot(p, add=TRUE)

image5

Multi-layer Raster objects can be plotted as individual layers

plot(b)

image6

They can also be combined into a single image, by assigning individual layers to one of the three color channels (red, green and blue):

plotRGB(b, r=1, g=2, b=3)

image7

You can also plot Raster* objects with spplot.

bounds <- list("sp.polygons", p)
spplot(r, sp.layout=bounds)

image8

spplot(b, layout=c(3,1))

image9

The rasterVis package has several other lattice based plotting functions for Raster* objects. The rasterVis package also facilitates creating a map from a RasterLayer with the ggplot2 package.

You can also use the a number of other plotting functions with a raster object as argument, including hist, persp, contour}, and density. See the help files for more info.

Basemaps

From Google and others…

Get a google map.

library(dismo)
g <- gmap("Belgium")
# Lookup and make point for Brussels
brus <- geocode('Brussels, Belgium')
merc <- Mercator(brus[, c('longitude', 'latitude')])

Plot it

plot(g, interpolate=TRUE)

image10

points(merc, pch='*', col='red', cex=5)

image11

Specialized packages

coming soon….