Maps in R

In this tutorial we’ll cover the following topics:

Basic Maps
Heat Maps
Leaflet Maps

Advanced topic:

Spatial Data

Basic Maps

There is no function for displaying maps in the base R functionality. To display a simple map, we use the maps package. First, let’s create an Indiana map.

library(maps)
map(database = 'state', regions = 'indiana')

Now we can add a few monitors as points. For illustration, I’ve created a simple data frame below, with latitude and longitude information.

monitors <- read.table(header=TRUE, text='
  monitorid        lat        long                       name            
          1   41.60668  -87.304729                 Gary-IITRI
          2  39.811097  -86.114469     Indpls-Washington-Park
          3  39.749019  -86.186314         Indpls-Harding-St.
          4  38.013248  -87.577856     Evansville-Buena-Vista
          5  39.159383  -86.504762                Bloomington
          6  39.997484  -86.395172                 Whitestown
                       ')

Now we can use the longitude column as the x-axis values and latitude for the y-axis values in the points() function. This function adds points to an already existing plot in R.

points(x = monitors$long, y = monitors$lat)

We can jazz it up a bit by making it a county map and changing the symbol type and color.

map(database = 'county', regions = 'indiana')
points(x = monitors$long, y = monitors$lat, pch = 19, col = 'red')

We can also select specific counties to look at, and give the map a title.

map(database = 'county', regions = c('indiana,marion', 'indiana,boone'))
points(x = monitors$long, y = monitors$lat, pch = 19, col = 'red')
title(main = "Air Monitor Locations")

If we want to label the monitors, we use the text() function.

map(database = 'county', regions = c('indiana,marion', 'indiana,boone'))
points(x = monitors$long, y = monitors$lat, pch = 19, col = 'red')
title(main = "Air Monitor Locations")
text(x = monitors$long, y = monitors$lat, labels = monitors$name, pos = 2)

Heat Maps

The choroplethr package makes it fairly easy to create heat maps in R. For illustration, here’s a simple data frame of the number of nonattainment areas for 8-hr ozone (2008 standard) by state in Region 5 (reference).

nonattain <- read.table(header=TRUE, text='
      state  areas
   illinois      2
    indiana      2
   michigan      0
       ohio      3
  wisconsin      2
                      ')

Now we’ll use the choroplethr package to make a heat map. The state_choropleth() function requires the column with state names to be labeled “region” and the column with the numeric values to be labeled “value”. To do this, we’ll introduce another handy dplyr function, rename().

library(dplyr)

nonattain <- rename(nonattain, region = state, value = areas)

Now we can use the state_choropleth() function.

library(choroplethr)
library(choroplethrMaps)

nonattain_map <- state_choropleth(nonattain, zoom = nonattain$region,
                                  title = "2008 8-hr Ozone Nonattainment",
                                  legend = "Areas")
nonattain_map

Leaflet Maps

Leaflet is an interactive map that can be created in R using the leaflet package.

library(leaflet)
m <- leaflet()
m <- addTiles(m)
m <- addMarkers(m, lng=monitors$long, lat=monitors$lat, 
                popup=monitors$name)
m

Exercises

Exercises for this tutorial can be found here: http://rpubs.com/NateByers/MapsExercises.

Advanced Topic

Spatial Data

It is possible to perform any kind of geographical analysis in R.

To illustrate how to create spatial data objects in R, let’s re-create the simple monitors data frame, with a small modification:

monitors <- read.table(header=TRUE, text='
  monitorid        long        lat     datum                    name            
          1  -87.304729   41.60668     WGS84              Gary-IITRI
          2  -86.114469  39.811097     WGS84  Indpls-Washington-Park
          3  -86.186314  39.749019     WGS84      Indpls-Harding-St.
          4  -87.577856  38.013248     NAD83  Evansville-Buena-Vista
          5  -86.504762  39.159383     NAD83             Bloomington
          6  -86.395172  39.997484     NAD83              Whitestown
                       ')

Now we have a “datum” column. We’ve also put the longitude column first, followed by the latitude column. And to be precise, before we lump all of these monitors together in a data object, we need to split them up by datum.

mon_WGS84 <- filter(monitors, datum == "WGS84")
mon_NAD83 <- filter(monitors, datum == "NAD83")

sp is the foundational package for working with spatial data in R. Once we library the package we can create SpatialPoints data objects by using the SpatialPoints() function.

library(sp)

mon_WGS84_sp <- SpatialPoints(coords = select(mon_WGS84, long, lat),
                              proj4string = CRS("+proj=longlat +ellpsWGS84"))

mon_NAD83_sp <- SpatialPoints(coords = select(mon_NAD83, long, lat),
                              proj4string = CRS("+proj=longlat +ellpsNAD83"))

Now we can plot the SpatialPoints objects.

map(database = 'state', regions = 'indiana')
plot(mon_WGS84_sp, add = TRUE)

plot(mon_NAD83_sp, add = TRUE)