Process 30 Day Map Challenge

Author

Malcolm Campbell

Process: 30 Day Map Challenge

This document is for Day 23. Process. Show how you make a map. This could be a tutorial, a step-by-step graphic, a blog post, a video, or a screenshot of your work environment. Combine it with a map from another day! See the full list here https://30daymapchallenge.com/. This is based on the previous day Data challenge: Natural Earth.

Making a ‘simple’ map

To make a figure of the North South Divide in the UK, we will need to find some administrative boundaries from an authoritative source. Firstly, lets load some key packages we will need using the pacman library and the p_load function which can help us by installing any library that is missing. We will want to use rnaturalearth - for the boundaries with sf and tmap - for the spatial elements and visualisation alongside tidyverse - for data wrangling.

pacman::p_load(rnaturalearth, sf, tmap, tidyverse)

Next, lets get the administrative boundaries of the United Kingdom to use in our figure. To do this we use the ne_states() command from {rnaturalearth} library which gives us sub national boundaries. We need to input the country required, for example the United Kingdom, the file format we want back, in this case an sf format.

UK <- ne_states(country="United Kingdom", returnclass = "sf")

A quick test to see if this has downloaded as desired is to produce a simple map using tmap, or whichever spatial visualisation works for you. Other possibilities include ggplot2 and leaflet among others. For tmap, we need to specify the file (UK) within the tm_shape function and then use either tm_polygons, tm_borders or tm_fill to create a map. Have a look at the three maps below to see the subtle differences and then query ?tm_borders using the question mark to understand the arguments needed for each.

tm_shape(UK) + tm_polygons()

tm_shape(UK) + tm_borders()

tm_shape(UK) + tm_fill()

Next, lets decide which places are in the North, and which are in the south. We can do this by seeing what regions we have available to put into categories.

unique(UK$region)

 [1] "Northern Ireland"           "West Wales and the Valleys"
 [3] "North West"                 "West Midlands"             
 [5] "South West"                 "Eastern"                   
 [7] "North East"                 "South Western"             
 [9] "Yorkshire and the Humber"   "East Midlands"             
[11] "East"                       "South East"                
[13] "East Wales"                 "Highlands and Islands"     
[15] "North Eastern"              "Greater London"

After seeing that there are 16 regions, we want to assign a value to each in order to be able to map the North-South divide. Using the tidyverse package to do some basic data wrangling and variable creation with the mutate and the case_when functions will help us get there. We will create a variable called DIVIDE to use as a category for future mapping by creating a binary north or south set of responses for each region in the spatial file of the UK.

UK <- UK |>
  mutate ( DIVIDE = case_when(
    region =="South East" ~ "South",
    region =="East" ~ "South",
    region =="South West" ~ "South",
    region =="Greater London" ~ "South",
    TRUE ~ "North" )) # Default category - for the other 12

Now for the slightly more complicated part. We can now set up a custom palette to use in mapping. To do this we select our desired colours for the north and the south.

NS_palette <- c("South"="dodgerblue2", North="firebrick")

Next, we will make a map using the tmap package. It is easier to do this in steps, to try and ease understanding of the different commands and options, which are myriad!

Firstly, lets just colour in map polygons and add a title using tm_polygons and tm_title commands. A title is fairly easy to explain, we simply add the text “North-South Divide” within the command so that tm_title(text="North-South Divide", size = 2.5) becomes the title and we can adjust the size also, lets go for 2.5 times bigger.

We use our shape, the UK, within the tm_shape command as before. This time however, we want to fill in the polygons using the DIVIDE variable, fill="DIVIDE" that we created. The critical part is to use fill.scale to set the values from our previously created custom palette. To do this we use tm_scale(values = NS_palette) which will pull either the north or south colours that we set up in our custom palette and assign these to each region in the map. Finally, we set the border of the polygon colours to NULL col=NULL, as we haven’t properly merged the underlying polygons and we want to hide this. The last step is then to suppress the legend using the tm_legend_hide() option. I’ve also added a command to stop tmap autoscaling using tmap_options(component.autoscale = FALSE).

NSmap <-
  tm_shape(UK) +
  tm_polygons(fill="DIVIDE",
              fill.scale = tm_scale(values = NS_palette),
              col=NULL, fill.legend = tm_legend_hide())+
  tm_title(text="North-South Divide", size = 2.5) +
  tmap_options(component.autoscale = FALSE)
NSmap

After a reasonably simple map, we need to add a few more parameters and arguments to further customise the figure further. The command tm_credits allows us to add in text relating to data sources and other items using the \n adds a new line and the position it where it needs to go, so the center and bottom of the figure. I have also chosen a specific text fontface of bold and fontfamily of mono and added a background that is black using bg.color with white text using color="white" and again, the size parameter changes the size of the text. Our final tweak is to use the tm_layout command to adjust the background colour of the map and drop the frame, using frame=F to get rid of the standard box around the plot.

NSmap <-
  tm_shape(UK) +
  tm_polygons("DIVIDE",
              fill.scale = tm_scale(values = NS_palette),
              col=NULL, fill.legend = tm_legend_hide())+
  tm_title(text="North-South Divide", size = 1.5,
           frame=F, fontface = "bold", fontfamily = "mono",
           color = "white", bg = TRUE, bg.color = "black", 
           position = tm_pos_out("center", "top", pos.h = "center") )+
  tm_layout(bg.color = "aliceblue", Frame=F)  +
  tm_credits(text ="Created by Prof. Malcolm Campbell \nData: rnaturalearth boundaries",
             position = tm_pos_out("center", "bottom", pos.h = "center"),
             size=0.75, fontface = "bold", fontfamily = "mono",
             color = "white", bg = TRUE, bg.color = "black", ) +
  tmap_options(component.autoscale = FALSE)
NSmap

Lastly, we might want to save our map and share it using the tmap_save command. We need to put in our map object, called NSmap, choose a filename and format, for example .png alongside a height, width and dots per inch (dpi) value.

tmap_save(NSmap, filename = "./maps/Day22_NS.png", height = 12, width = 12, dpi=400)

There you have it. A ‘simple’ map.