Click the Original, Code and Reconstruction tabs to read about the issues and how they were fixed.

Original


Source: Howmuch.net


Objective

The objective of the visualization is to be informative and to provide insights on the global trade of major conventional weapons*. By providing a snapshot of the magnitude of the trade and the key players in the market, the visualization aims at making arms’ trade data more accessible to the general public. The targetted audience is likely composed of university students, researchers, journalists, and professionals in fields such as business, politics and finance.

*(such as aircraft, artillery, radar systems, missiles, and ships designed for military use)

The visualization chosen had the following three main issues:

  • Visual bombardment: This issue stems from the fact that the overall design is cluttered and lacks a clear hierarchy. There are visual elements that are unnecessary and not organized in a meaningful way. For example, flags and the countries depicted in them do not add any value; rather, they are a source of distraction and make the visualization have too many colors. As a result, even though the original visualization eventually enables the viewer to capture the main message (i.e., who are the biggest exporters and importers), visual bombardment makes it aesthetically unpleasing and potentially inaccessible to people with certain visual impairments.
  • Poor use of the source data: The underlying dataset contains additional variables (besides the values of imports and exports) that were excluded and not depicted. However, these variables (1) influence the countries’ ability to export or import weapons and (2) would improve the comparability between countries. Examples include the % of GDP devoted to military expenses and which income group the country belongs to. As a consequence, the visualization has poor context and might provide an incomplete picture of the arms trade.
  • Information redundancy: the same information is displayed more than once through multiple visual elements. Both the size, the colour and the text of the bubbles show the value of arms’ trade. As a result the legend is inaccurate and inconsistent. For example the use of the size visual element is particularly problematic. The largest bubble in the legend corresponds to a value of 1B$ but the map also contains countries with higher values. All in all the legend is unclear and hardly readable.

Reference

Code

The following code was used to fix the issues identified in the original.

Importing the data

After downloading the source data from The World Bank website I save them in Excel format. I perform some very preliminary operations in Excel in order to clean the data and I subsequently import it in R.

# Importing data 

library(readxl)

Exports_data <- read_excel("Arms_Trade_Data_v1.xlsx", 
    sheet = "Data_EX")

Imports_data <- read_excel("Arms_Trade_Data_v1.xlsx", 
    sheet = "Data_IM")

Mil_Exp_data <- read_excel("Arms_Trade_Data_v1.xlsx", 
    sheet = "Data_MIL_EXP_%GDP")
Preparing the data
library(tidyverse)
library(dplyr)

Mil_Exp <- Mil_Exp_data |>
  dplyr::mutate(Mil_Exp_2016 = round(Mil_Exp_2016, digits=2)) 

Exports <- Exports_data |>
  # Show the data in Billion
  dplyr::mutate(Exp_2016 = round(Exp_2016/10^9, 2)) |>
  # Add the data on military expenditures
  dplyr::left_join(Mil_Exp, by = 'Country_Code') |>
  dplyr::select(c(1,2,3,5)) |>
  dplyr::rename(Country_Name = Country_Name.x)
  
Imports <- Imports_data |>
  # Show the data in Billion
  dplyr::mutate(Imp_2016 = round(Imp_2016/10^9, 2)) |>
  # Add the data on military expenditures
  dplyr::left_join(Mil_Exp, by = 'Country_Code') |>
  dplyr::select(c(1,2,3,5)) |>
  dplyr::rename(Country_Name = Country_Name.x)
Map of World’s biggest importers
library(ggiraph)
library(sf)
library(tidyverse)
library(ggthemes)
library(patchwork)

# Building the pipe for ggplot ####
imports_gg <- Imports |>
  dplyr::inner_join(rnaturalearth::ne_countries(returnclass = "sf") |>
                      dplyr::select(iso_a3, geometry),
                    by=c("Country_Code"="iso_a3")) |>
  # Plot
  ggplot()+
  labs(title = 'Importers of Military Weapons', 
       caption = "Note: White countries correspond to NA values")+
  # Using the interactive option and custom tooltip
  geom_sf_interactive(aes(geometry=geometry, fill=Imp_2016, 
                          tooltip=paste0(Country_Name,
                                         "<br><b>Value:", Imp_2016,
                                         "<br><i>Military Expenditures (%GDP):", Mil_Exp_2016 |> format(big.mark=" "))))+
    scale_fill_stepsn_interactive(name ='US$ Bil', n.breaks=8, colors = c('#fff0f3', '#ffccd5', '#fbb1bd','#ff8fa3', '#ff5c8a','#c9184a','#B13356','#a4133c'), limits=c(0,3.5), na.value = 'white')+
  theme_map()+
  theme(plot.title = element_text(face='bold'))
                      
# ggiraph ####  
imports_girafe <- ggiraph::girafe(
  ggobj = imports_gg)
Map of World’s biggest exporters
#Importing packages 
library(sf)
library(rnaturalearth)
library(ggiraph)
library(tidyverse)
library(ggthemes)
library(patchwork)
library(ggpubr)

# Building the pipe for ggplot ####
exports_gg <- Exports |>
  dplyr::inner_join(rnaturalearth::ne_countries(returnclass = "sf") |>
                      dplyr::select(iso_a3, geometry),
                    by=c("Country_Code"="iso_a3")) |>
  # Plot
  ggplot()+
  labs(title = 'Exporters of Military Weapons')+
  # Using the interactive option and custom tooltip
  geom_sf_interactive(aes(geometry=geometry, fill=Exp_2016, 
                          tooltip=paste0(Country_Name,
                                         "<br><b>Value:", Exp_2016,
                                         "<br><i>Military Expenditures (%GDP):",Mil_Exp_2016 |> format(big.mark=" "))))+
    scale_fill_stepsn_interactive(
      name = "US$ Bil", n.breaks=10, limits=c(0, 10), colours = c('#a9d6e5', '#89c2d9','#61a5c2', '#468faf', '#2c7da0', '#2a6f97', '#014f86', '#01497c', '#013a63', '#012a4a'), na.value='white')+
  theme_map()+
  theme(plot.title = element_text(face='bold'))
                      
# ggiraph ####  
exports_girafe <- ggiraph::girafe(
  ggobj = exports_gg)
World Map of both importers and exporters
# Combining the two plots into a unique figure 
g <- wrap_plots(list(exports_gg, imports_gg), ncol = 1)
Military Expenditures (%GDP)

-By Country and Income Group

library(ggplot2)

mil_exp_plot <- Mil_Exp |>
  ggplot(aes(x=Mil_Exp_2016, y=reorder(Country_Name, Mil_Exp_2016), fill=Mil_Exp_2016)) +
  geom_bar(stat = 'identity')+
  geom_text(aes(label=round(Mil_Exp_2016,1)), stat = 'identity', hjust = -0.5, size = 2)+
  # Facet by Income Group
  facet_wrap(~factor(Income_Group, levels= c('High income', 'Upper middle income', 'Lower middle income', 'Low income')), scales = 'free_y')+
  scale_x_continuous(breaks=seq(0,13,1))+
  labs(title='Military Expenditures (%GDP) by Country and Income Group*', caption = 'Note: Income Group is based on the World Bank classification', x='Military Expenditures (%GDP)', y=NULL)+
  scale_fill_gradientn(colours = c('#d8f3dc','#b7e4c7','#95d5b2', '#74c69d', '#52b788', '#40916c', '#2d6a4f', '#1b4332'))+
  theme_bw()+
  theme(legend.position = 'none', plot.title = element_text(face='bold'))

Data Reference

Reconstruction

The following plot fixes the main issues in the original.

❗Hover over each Country to find out the value of arms’ trade as well as the % of GDP spent in military expenses 🌎

Additional

I have additionaly constructed the faceted plot shown below. Countries are categorized by income group and are shown in decreasing order depending on their military expenditures (%GDP).

This plot can be used as a supplement to the main visualization. In fact it provides a clear and comparative view of how much each country is spending on their military in relation to their economic output -which can be useful to put data into a more informative context.