Bivariate tmap

The bivariate_tmap.R script defines a simple function named bivariate_choropleth which uses the tmap library to create a bivariate choropleth map (using a color scheme fro Joshua Stevens's fantastic blog post) and plot it including a legend created using the grid library.

The bivariate_choropleth function parameters are:

• bivmap_dataset, a SpatialPoligonDataFrame;
• bivmap_vars, a vector of characters containing the name of the two variables;
• bivmap_labels, a vector of characters containing the labels for the two variables, to use in the legend (default is NA, in; which case, values from bivmap_vars are used);
• bivmap_style, classification type for the bins (default is 'quantile');
• bivmap_scale, whether to use a scale bar (default is FALSE).

The bivariate_choropleth function uses a function named get_bivariate_choropleth which simply creates the map (without the legend) and returns the related object. This function can also be used directly for further manipulation of the map.

Requirements

The function was developed using R version 3.5.1, and it requires the following R libraries:

• sp
• spdep
• tmap
• classInt
• grid
• gridExtra
• lattice

Example

Example script.

# Import function
source("bivariate_tmap.R")

# Import example data
library(tmap)
data("World")
# Transform data to SpatialPolygonDataFrame
library(sf)
world_spdf <- as(World, "Spatial")

# Plot bivariate choroplet map
bivariate_choropleth(world_spdf, c("gdp_cap_est", "inequality"))

Console output (categories).

gdp_cap_est breaks (x-axis):
style: quantile
[300.4693,3356.704) [3356.704,13587.93)    [13587.93,2e+05] 
                 59                  58                  59 
inequality breaks (y-axis):
style: quantile
  one of 9,045 possible partitions of this variable into 3 classes
[0.0432207,0.1672642) [0.1672642,0.2736572) [0.2736572,0.5073423] 
                   45                    45                    46 

Output map.

Acknowledments

I would like to acknowledge the contribution by Andrea Ballatore, who collaborated on the paper for which I originally wrote this script, helped me testing it, and encouraged me to publish it.