Which Districts Had the Largest Population in Malawi in 2008?

A step-by-step cartographic walkthrough of how I built a thematic map that shows population for Malawi’s districts using labels for symbology using R, sf, and tmap — part of my 2025 #30DayMapChallenge series.
cartograpy
R
population
Author

Dickson Mbeya

Published

July 5, 2026

Introduction

I decided to visualise Malawi’s district population from the 2008 census. Instead of a traditional filled‑polygon choropleth, I used a typographic map—where the district names are scaled by their population. This technique makes the data self‑explanatory and adds an elegant, text‑heavy aesthetic that works well for social media.

Today I’ll walk you through the entire R workflow, from loading the shapefiles to adding an Africa locator inset. Let’s dive in!

1. Load the required packages

We need a handful of specialised packages:

  • sf – reads and handles spatial vector data (shapefiles).
  • tmap – our main mapping engine; it uses a grammar of graphics similar to ggplot2 but is tailored for maps.
  • dplyr – for quick data manipulation (renaming columns).
  • extrafont – to use system fonts like Roboto for a modern, clean look. If you haven’t used extrafont before, you can run font_import() and loadfonts() once to make your system fonts available in R.
library(tmap)
library(dplyr)
library(sf)
popn <- st_read("data/mw_popn.shp")
Reading layer `mw_popn' from data source 
  `C:\Users\devmbeya\Documents\blog\posts\make_popn_map\data\mw_popn.shp' 
  using driver `ESRI Shapefile'
Simple feature collection with 32 features and 23 fields
Geometry type: MULTIPOLYGON
Dimension:     XY
Bounding box:  xmin: 3636988 ymin: -1935618 xmax: 3998420 ymax: -1047139
Projected CRS: WGS 84 / Pseudo-Mercator
bnd <- st_read("data/mw_bnd.shp")
Reading layer `mw_bnd' from data source 
  `C:\Users\devmbeya\Documents\blog\posts\make_popn_map\data\mw_bnd.shp' 
  using driver `ESRI Shapefile'
Simple feature collection with 22 features and 1 field
Geometry type: POLYGON
Dimension:     XY
Bounding box:  xmin: 464479.7 ymin: 8105086 xmax: 814025.4 ymax: 8964834
Projected CRS: WGS 84 / UTM zone 36S

2. Read the spatial data

The data comes from the Malawi National Statistical Office and contains district boundaries and population counts for 2008.

In this example I store the shapefiles inside a data/ folder relative to the project root. This makes the script portable and reproducible.

First, I read the population point/centroid layer which contains the district names and population counts:

popn <- st_read("data/mw_popn.shp")
Reading layer `mw_popn' from data source 
  `C:\Users\devmbeya\Documents\blog\posts\make_popn_map\data\mw_popn.shp' 
  using driver `ESRI Shapefile'
Simple feature collection with 32 features and 23 fields
Geometry type: MULTIPOLYGON
Dimension:     XY
Bounding box:  xmin: 3636988 ymin: -1935618 xmax: 3998420 ymax: -1047139
Projected CRS: WGS 84 / Pseudo-Mercator

Next, I read the district boundary polygon layer which will serve as the background map:

bnd <- st_read("data/mw_bnd.shp")
Reading layer `mw_bnd' from data source 
  `C:\Users\devmbeya\Documents\blog\posts\make_popn_map\data\mw_bnd.shp' 
  using driver `ESRI Shapefile'
Simple feature collection with 22 features and 1 field
Geometry type: POLYGON
Dimension:     XY
Bounding box:  xmin: 464479.7 ymin: 8105086 xmax: 814025.4 ymax: 8964834
Projected CRS: WGS 84 / UTM zone 36S

Let’s inspect the population data to see what columns we have:

head(popn)
Simple feature collection with 6 features and 23 fields
Geometry type: MULTIPOLYGON
Dimension:     XY
Bounding box:  xmin: 3667131 ymin: -1884697 xmax: 3932595 ymax: -1047139
Projected CRS: WGS 84 / Pseudo-Mercator
      DIST_NAME      DISTRICT  REGION Shape__Are Shape__Len
1        Balaka        Balaka Central 2309288545  227264.32
2      Blantyre      Blantyre   South 1935900658  314395.31
3 Blantyre City Blantyre City   South  261111996   92609.81
4      Chikwawa      Chikwawa   South 5318053799  413383.01
5    Chiradzulu    Chiradzulu   South  825563720  184760.60
6       Chitipa       Chitipa   North 4499298409  627147.82
                                GlobalID CreationDa Creator   EditDate Editor
1 {7732b160-5d5f-4f3c-a381-d8e5bbe4f9c2} 2022-02-06  mbeyad 2022-02-07 mbeyad
2 {1f059ed3-7289-4a55-a6d5-9b45872048fe} 2022-02-06  mbeyad 2022-02-07 mbeyad
3 {ead7d5d2-408f-4cce-bcba-827bb9a71dab} 2022-02-06  mbeyad 2022-02-07 mbeyad
4 {3ea276d9-a990-4a77-8e3b-d2807b223417} 2022-02-06  mbeyad 2022-02-07 mbeyad
5 {fa51c606-08ea-4b41-8340-ba68833bad02} 2022-02-06  mbeyad 2022-02-07 mbeyad
6 {20595550-2501-44ae-8f91-5786859edc50} 2022-02-06  mbeyad 2022-02-07 mbeyad
  MALE1998 FEM1998 TOTAL1998 MALE2008 FEM2008 TOTAL2008 FEM2018 TOTAL2018
1   120706  132397    253103   151637  165111    316748  229105    438379
2   413429  395968    809397   164546  173501    338047  232756    451220
3        0       0    502053   337655  323789    661444  399092    800264
4   178217  178465    356682   217981  220914    438895  287794    564684
5   111376  124674    236050   137194  153752    290946  187196    356875
6    60682   66117    126799    86152   92920    179072  120535    234927
  MALE2018 POPDEN2008 POPDEN1998  AREA_KM2 DEN2018
1   209274        148        119 2133.8513     205
2   218464        189        453 1788.3027     252
3   401172       2805       2129  235.8099    3394
4   276890         90         73 4891.6565     115
5   169679        382        310  762.5165     468
6   114392         42         30 4247.7344      55
                        geometry
1 MULTIPOLYGON (((3916682 -16...
2 MULTIPOLYGON (((3908694 -17...
3 MULTIPOLYGON (((3905701 -17...
4 MULTIPOLYGON (((3822326 -17...
5 MULTIPOLYGON (((3913071 -17...
6 MULTIPOLYGON (((3673613 -10...

Looking at the output, I can see the column TOTAL2008 holds the total population for each district. I’ll rename it to something more intuitive for our mapping:

popn <- popn %>%
  rename(Population = TOTAL2008)

Now we have a clean attribute called Population that we can use in our map.

3. Design the main typographic map

This is the heart of the visual. Instead of colouring polygons, I use tm_text() to place the district names directly on the map. The size argument is mapped to the Population column—so larger populations appear in bigger text.

Let’s break down the styling choices:

Background – a very light grey ("grey99") for the districts, with thin grey borders ("grey70"). This keeps the focus on the text.

Text colour – bright orange ("orange") stands out beautifully against the light grey.

Layout – I added generous inner margins so the text doesn’t touch the frame. The frame itself is rounded (frame.r = 10) and coloured orange to tie the design together.

Legend – I place it at the top right (position c(0.7, 0.8)) with a light grey frame.

Title – a multi‑line, attention‑grabbing question. The title is placed on the right side, with a white background to improve readability.

Credits – three separate credits cover the challenge tag, my name, and the data source. I use Roboto throughout for consistency.

mypop <- tm_shape(bnd) + 
  tm_polygons(fill = "grey99", col = "grey70") +
  
  tm_shape(popn) + 
  tm_text(text = "DISTRICT", 
          size = "Population", 
          col = "orange", 
          fontface = "bold") + 
  
  tm_layout(inner.margins = c(0.02, 0.3, 0.02, 0.3),
            outer.margins = c(0.03, 0.02, 0.03, 0.02),
            frame.r = 10,
            frame.lwd = 1,
            frame.color = "orange",
            legend.frame.lwd = 1,
            legend.frame.color = "grey80",
            legend.title.fontfamily = "Roboto Light",
            legend.position = c(0.7, 0.8)) + 
  
  tm_title("Which Districts\nHad the Largest\nPopulation in Malawi\nin 2008?", 
           color = "orange",
           size = 1.2,
           bg = TRUE,
           bg.alpha = 1,
           position = c("right", "top"),
           frame = FALSE) +
  
  tm_credits("Date: 05 July 2026",
             position = c("left", "bottom"),
             size = 0.7, color = "orange",
             fontfamily = "Roboto") + 
  
  tm_credits("© Dickson Mbeya",
             position = c("center", "bottom"),
             size = 0.7, color = "orange",
             fontfamily = "Roboto") +   
  tm_credits("Data Source: Malawi\nNational Statistical Office\n\nwww.dicksonmbeya.com",
             fontfamily = "Roboto",
             position = c("right", "bottom"),
             size = 0.6, color = "orange")

Why text‑sized names instead of a classic choropleth?
Because readers can immediately identify the largest districts without needing to consult a legend. The text itself becomes the visual variable, which is both intuitive and visually striking.

4. Build a locator inset map

When you share maps online, not everyone knows exactly where Malawi is. A small locator map solves that problem.

I use an orthographic projection centered on Malawi’s coordinates (lat_0 = -13.98, lon_0 = 33.77). This creates a globe‑like effect that makes the country pop out from the rest of Africa.

First, I load the world dataset using the rnaturalearth package:

library(rnaturalearth)
Warning: package 'rnaturalearth' was built under R version 4.5.3
world <- ne_countries(scale = "medium", returnclass = "sf")

Next, I define the orthographic projection and transform both the world and Malawi boundaries:

ortho_crs <- "+proj=ortho +lat_0=-13.98 +lon_0=33.77"
world_ortho <- st_transform(world, crs = ortho_crs)
bnd_ortho <- st_transform(bnd, crs = ortho_crs)

Now I build the locator map. African countries are filled with light grey, while Malawi is outlined in bright orange:

locator_map <- tm_shape(world_ortho) +
  tm_polygons(col = "grey80", border.col = "white", lwd = 0.3) +
  tm_shape(bnd_ortho) +
  tm_borders(col = "orange", lwd = 1) +
  tm_layout(frame = FALSE, bg.color = "white")
── tmap v3 code detected ───────────────────────────────────────────────────────
[v3->v4] `tm_polygons()`: use 'fill' for the fill color of polygons/symbols
(instead of 'col'), and 'col' for the outlines (instead of 'border.col').
This message is displayed once every 8 hours.

The white background and no frame ensure the inset integrates cleanly with the main map.

5. Combine everything with tm_inset()

Now I overlay the locator map onto the main map using tm_inset().

I set the inset size to 9 which works well with the existing layout dimensions. The position c(0, 0.3) places it at the left side, slightly below the top margin.

final_map <- mypop +
  tm_inset(locator_map,
           width = 9,
           height = 9,
           position = c(0, 0.3))

final_map
Warning in graphics::strwidth(comp$title, units = "inch", cex = titleS, : font
family not found in Windows font database
Warning in graphics::strwidth(comp$text, units = "inch", family =
comp$fontfamily, : font family not found in Windows font database
Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
font family not found in Windows font database
Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
font family not found in Windows font database
Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
font family not found in Windows font database

6. Save the final output

You can export the map as a high‑resolution PNG for sharing on social media, in presentations, or for print.

#tmap_save(final_map, "2008_Population.png", width = 10, height = 8, dpi = 300)

What the map reveals

Looking at the scaled district names, Lilongwe (the capital), Blantyre (the commercial hub), and Mzimba (the northern giant) clearly dominate the population distribution. The text‑size approach makes this pattern visible instantly—without needing a separate colour legend.

The map also reveals interesting spatial patterns: - The central region (Lilongwe) has the highest population - The southern region (Blantyre) is the second most populated - The northern region (Mzimba) has large populations but more spread out

Final thoughts

This map was a fun exercise in using tmap’s typographic capabilities. By focusing on text rather than fill colours, I created a poster‑style visual that is both informative and decorative.

Key takeaways for your own maps:

  • Don’t be afraid to use tm_text() for quantitative data – size scaling works beautifully
  • Always add a locator inset when your audience might be global
  • Small typographic touches (fonts, frame colours, margin control) elevate the final product
  • Using a light background with bold text creates a clean, professional look

If you have any questions about the code or want to adapt it for your own country, feel free to reach out!

Happy mapping, everyone!
– Dickson