Basic Cartogram with R

A cartogram is a map in which the geometry of regions is distorted in order to convey the information of an alternate variable. In this post, we are going to draw a map of Africa where the size of each country is distorted proportionally to its population. A cartogram is frequently a choropleth map in which regions are colored according to a numeric value (not necessarily the one used to build the cartogram). To understand the cartogram more easily, one can refer to what a geospatial object is. 

Preparing the Map data and creating Basic Cartogram

To perform a cartogram map we will create map data using wrld_simpl.

R

# library for the map data
library(maptools)
data(wrld_simpl)

# extracting the African map data
# this data contains the latitude and longitude of
# region boundaries and regions
afr = wrld_simpl[wrld_simpl$REGION==2,]

# plotting outline map
plot(afr)

Output:

Spatial Polygon Features in Cartogram

We have a spatial object and a numeric value associated with each region, it is possible to color each region according to its value.

R

library(sf) # library to get the SpatialPolygon
afr_sf = st_as_sf(afr)
plot(afr_sf)

Output:

Cartogram and its customization

Here we will create cartogram and then we will customize for this we will use wrld_simpl and then extract the information like shape, coordinates and then transform geometry polygon objects to new geospatial objects.

R

library(tidyverse)

# Get the shape file of Africa
library(maptools)

# retrieve the data for the countries
data(wrld_simpl)

# extract the data for Africa
# afr_data holds the data about the African polygon shape
# and information like coordinates, latitude and longitude
afr_data = wrld_simpl[wrld_simpl$REGION == 2,]
cat ("Type of afr_data: ", typeof(afr_data), "\n\n")

# We work with the cartogram library
library(cartogram)

# s4 is an system for object oriented programming
# convert this foreign object to sf object
sfno = st_as_sf(afr)
# transform geometry polygon objects to a new coordinate reference system
sfnoproj = st_transform(sfno, crs = 23038)

# construct a cartogram using the population in 2005
afr_cartogram <- cartogram_cont(sfnoproj, "POP2005", itermax = 7)

# This is a new geospatial object, we can visualise it!
ggplot() +
  geom_sf(data = afr_cartogram, aes(fill = POP2005))

Output:

Type of afr_data:  S4
Mean size error for iteration 1: 6.65362274402649 
Mean size error for iteration 2: 5.77932593731838 
Mean size error for iteration 3: 5.11349284302375 
Mean size error for iteration 4: 4.56333360251358 
Mean size error for iteration 5: 4.09786807153034 
Mean size error for iteration 6: 3.69769020137334 
Mean size error for iteration 7: 3.3519304681889