13  Merge Shapefiles With External Data

13.1 Introduction

A common task in mapping is that we have a shapefile of map data, but we want to merge in some external data from another source so that we can map that external data. Usually we want to use different colors to map that external data.

Here, I use a shapefile of countries of the world, and map it with some made up data of arbitrary and made up x values for certain countries.

This tutorial builds upon another tutorial on Mapping with ggplot(Chapter 12)

13.2 Call Libraries

library(sf) # simple (spatial) features

library(ggplot2) # beautiful plots

library(dplyr) # data wrangling and joins

library(tibble) # library for updated dataframes

library(readr) # read CSV files

library(countrycode) # work with country data

13.3 Get Shapefile

Here I am using the wrld_simpl dataset of country outlines that is available in several R packages. I read this data into an object called world.

world <- read_sf("./shapefiles/wrld_simpl/wrld_simpl.shp")

13.4 Make a Map Without Data

I map the data with ggplot, and the special geom, geom_sf.

ggplot(world) + 
  geom_sf() +
  labs(title = "Demonstration Map With No Data")

13.5 Get External Data

Here I read in a completely made up data set of arbitrary values. The file that I am reading in is a csv file, but could be a file in any number of formats.

externaldata <- read_csv("./external-data/externaldata.csv")

externaldata # replay data set

# A tibble: 6 × 3
  ISO3  arbitrary_x name     
  <chr>       <dbl> <chr>    
1 COL            10 Colombia 
2 CHL             9 Chile    
3 ARG             8 Argentina
4 GHA             7 Ghana    
5 MLI             6 Mali     
6 AUS             5 Australia

13.6 Join Data to Shapefile

I use left_join from the dplyr package to merge the spatial data in world with externaldata.

left_join is a function that keeps all observations in the data on the left (the shapefile), and only those matching observations in the data on the right (the external data), which is usually what I want in mapping.

I need a unique identifier for my rows of data, so here I use ISO3, a unique 3 letter identifier for countries of the world.

newdata <- left_join(world, externaldata, by = "ISO3")

13.7 Make a Map With The Data

Once I have the merged data, it is easy to map it with ggplot and geom_sf. Note that I need to specify an aesthetic for geom_sf. Here arbitrary_x is the fill color for countries on the map.

Data could also be mapped with another package like leaflet (Section 16.1).

ggplot(newdata) +
  geom_sf(aes(fill = arbitrary_x)) + # adding a fill aesthetic
  scale_fill_viridis_c(na.value = "grey97", # value for NA
                       option = "turbo") + # viridis colors
  labs(title = "Demonstration Map With Merged Data") +
  theme_minimal() # better theme