The following maps show different map source/type options (except cloudmade)

ggmap quickstart The appearance of these maps may be very different depending on zoom/scale

For more functionality, see ggmap documentation and If you cant get the
https://dl.dropboxusercontent.com/u/24648660/ggmap%20u
stamen: watercolor stamen: toner stamen: terrain
map you want by
seR%202012.pdf adjusting the
There are 2 basic steps to making a map using ggmap: location/zoom
variables, the
Part 1: Download Part 2: Plot raster functions designed
map raster and overlay data for the different map
Part 1: Downloading the map raster sources provide more
Start by loading the package: library(ggmap) options:
get_googlemap,
1. Define location: 3 ways get_openstreetmap,
- location/address get_stamenmap,
myLocation <- "University of Washington get_cloudmademap
- lat/long
myLocation <- c(lon = -95.3632715, lat = 29.7632836) google: terrain google: satellite google: roadmap google: hybrid
- bounding box lowerleftlon, lowerleftlat, upperrightlon, upperrightlat
(a little glitchy for google maps)
myLocation <- c(-130, 30, -105, 50)

Convert location/address its lat/long coordinates:

geocode(University of Washington)
2. Define map source, type, and color
The get_map function provides a general approach for quickly
obtaining maps from multiple sources. I like this option for exploring
different styles of maps.
There are 4 map sources to obtain a map raster, and each of these osm*
sources has multiple map types (displayed on right). All maps can be displayed If you use Rstudio:
- stamen: maptype = c(terrain, toner, watercolor) in black and white Sometimes a plot will not
- google: maptype = c(roadmap, terrain, satellite, hybrid) display. Increasing the size
color = bw
- osm: open street map of the plot window may
- cloudmade: 1000s of maps, but an api key must be obtained help. dev.off() prior to
from http://cloudmade.com plotting may also help.
myMap <- get_map(location=myLocation, The urlonly = TRUE will
source="stamen", maptype=watercolor", crop=FALSE) return a url that will
display your map in a web
ggmap(myMap)
browser. Which is pretty
This will produce a map that looks something like this cool and may be handy!
NOTE: crop = FALSE because otherwise, with stamen plots, the map is legend=topleft will inset
slightly shifted when I overlay data. the legend on the top left
*Open street maps may return of the map is data is
3. Fine tune the scale of the map using zoom overlayed (page 2).
a '503 Service Unavailable
The get_map function takes a guess at the zoom level, but
error. This means their server
you can alter it:
is unavailable, and you must
zoom = integer from 3-21 myMap <-
wait for it to become available
3 = continent, 10=city, 21=building get_map(location=myLocation, Page 1 ggmap,
again.
(openstreetmap limit of 18) source=osm", color=bw)) Melanie Frazier
 3. Add polygons from shp file
ggmap quickstart The shp file is imported into R using the rgdal package, and must be transformed to geographic
coordinates (latitude/longitude) on the World Geodetic System of 1984 (WGS84) datum using
the rgdal package:
Part 2: Plotting the maps and data library(rgdal)
shpData <- readOGR(dsn="C:\\Documents and Settings\\Watershed", layer="WS")
1 . Plot the raster: proj4string(shpData) # describes datas current coordinate reference system
ggmap(myMap) # to change to correct projection:
2 . Add points with latitude/longitude coordinates: shpData <- spTransform(shpData,
ggmap(myMap)+ CRS("+proj=longlat +datum=WGS84"))
geom_point(aes(x = Longitude, y = Latitude), data = data,
alpha = .5, color="darkred", size = 3) To plot the data:
alpha = transparency geom_polygon(aes(x = long, y = lat, group=id),
color = color data = shpData, color ="white", fill ="orangered4",
size = size of points alpha = .4, size = .2)
color= outline color alpha = transparency
The size, color, alpha, and shape of the points can be scaled relative fill = polygon color size = outline size
to another variable (in this case estArea) within the aes function:
ggmap(myMap)+ 4. Annotate figure
geom_point(aes(x = Longitude, y = Latitude, size=sqrt(estArea)), baylor <- get_map('baylor university', zoom = 15, maptype = 'satellite')
data = data, alpha = .5, color="darkred") ggmap(baylor) +
annotate('rect', xmin=-97.11, ymin=31.54, xmax=-97.12, ymax=31.55, col="red, fill=white)+
annotate('text', x=-97.12, y=31.54, label = 'Statistical Sciences', colour = I(red'), size = 8)+
annotate('segment', x=-97.12, xend=-97.12, y=31.55, yend=31.55,
colour=I('red'), arrow = arrow(length=unit(0.3,"cm")), size = 1.5) +
labs(x = 'Longitude', y = 'Latitude') + ggtitle('Baylor University')

Controlling size and color

scale_size(range = c(3, 20))
But, the following is better for display because
size
it is based on area (rather than radius)
scale_area(range=c(3,20))

Continuous variables: color gradient between

n colors, e.g.:
scale_colour_gradientn(colours =
rainbow_hcl(7))
Discrete variables, e.g.:
color scale_colour_manual(values=rainbow_hcl(7))
Additional functions can be added to control scaling, e.g.: scale_colour_manual(values=c("8" = "red",
ggmap(myMap)+ "4" = "blue","6" = "green)
geom_point(aes(x = Longitude, y = Latitude, size=sqrt(estArea)), *Use colorspace and RColorBrewer to choose
data = data, alpha = .5, color="darkred")+ color combinations
scale_size(range=c(3,20)) Page 2 ggmap,
Melanie Frazier