This lesson requires the following packages:
if(!require('pacman')) install.packages('pacman')
pacman::p_load_gh('afrimapr/afrilearndata')
pacman::p_load(rnaturalearth,
ggspatial,
ggplot2,
mdsr,
tidyverse,
dplyr,
sf)Identify two types of Thematic maps (choropleth and dot maps) used by epidemiologist to visualize Geospatial data.
Create Thematic maps using {ggplot2} and the geom_sf() function.
Relate each Thematic map with a Geometry type.
• Colors or shadings represent geographic regions in relation to the value of a variable.
• E.g. Larger values could be indicated by a darker color, and smaller values by a lighter color.
• Geospatial data can be plotted using ggplot2::geom_sf().
• Colors and shapes can be depicted with the aes() function,
• Using the fill, color and size arguments.
• A Quantitative Choropleth map requires the fill argument.
• Let’s create one with the africountries dataset from {afrilearndata}!
• Administrative boundaries of all the countries in the African continent.
ggplot2::geom_sf() to plot African countries, only.fill each of them according to their population (pop_est):# WRITE YOUR CODE HERE
library(sf)
library(afrilearndata)
africountries %>%
mutate(geometry = st_as_text(geometry)) %>%
as_tibble()# A tibble: 51 × 12
name name_long pop_est gdp_md_est lastcensus income_grp iso_a3 geometry
<chr> <chr> <dbl> <dbl> <dbl> <chr> <chr> <chr>
1 Angola Angola 1.28e7 110300 1970 3. Upper … AGO MULTIPO…
2 Burundi Burundi 8.99e6 3102 2008 5. Low in… BDI MULTIPO…
3 Benin Benin 8.79e6 12830 2002 5. Low in… BEN MULTIPO…
4 Burkina F… Burkina … 1.57e7 17820 2006 5. Low in… BFA MULTIPO…
5 Botswana Botswana 1.99e6 27060 2011 3. Upper … BWA MULTIPO…
6 Central A… Central … 4.51e6 3198 2003 5. Low in… CAF MULTIPO…
7 Côte d'Iv… Côte d'I… 2.06e7 33850 1998 4. Lower … CIV MULTIPO…
8 Cameroon Cameroon 1.89e7 42750 2005 4. Lower … CMR MULTIPO…
9 Dem. Rep.… Democrat… 6.87e7 20640 1984 5. Low in… COD MULTIPO…
10 Congo Republic… 4.01e6 15350 2007 4. Lower … COG MULTIPO…
# ℹ 41 more rows
# ℹ 4 more variables: name_fr <chr>, name_pt <chr>, name_af <chr>,
# name_sw <chr>ggplot(data = africountries) +
geom_sf(mapping = aes(fill = pop_est))
• sf stands for “simple features”, an open standard used to represent a wide range of geometric shapes.
Create a Choropleth map with the world data from the {spData} package, using geom_sf(), to portrait its countries and fill them in relation to its population, available in the pop variable.
library(spData)
# Write and visualize your answer:
q1 <- world %>%
ggplot() +
geom_sf(mapping = aes(fill = pop))• Let’s keep using fill.
• Create a map with countries colored by their Economical classification (economy):
countries <- rnaturalearth::ne_countries(returnclass = "sf")
ggplot(data = countries) +
geom_sf(mapping = aes(fill = economy))
• Before using geom_sf(), verify that your Spatial data is an "sf" R object:
library(sf)
library(afrilearndata)
class(africountries)[1] "sf" "data.frame"• More about sf in the following lessons!
• Visualize how one variable changes across a defined regions.
• Region of interest (Sierra Leone)
• Partitioned into a finite number of subregions (districts)
• Number of cases aggregated at that level.
• Choropleth maps visualize a shape called Polygons.
• Collects data from an enclosed region.
• Partitioned into a finite number of areal units with defined boundaries.
• E.g., data collected by ZIP code, census tract, or the administrative boundary levels of a country (Figure 2).
• Thematic map type that uses dots to represent attribute values.
• The Dot map could use the size or color argument.
• A Quantitative Dot map requires the size argument.
• Let’s create a Dot map!
• Use the africapitals dataset, it contains the location of capital cities in the African continent.
ggplot2::geom_sf() to plot these locations,size them according to their number of inhabitants:library(sf)
library(afrilearndata)
africapitals %>% as_tibble()# A tibble: 51 × 5
capitalname countryname pop iso3c geometry
<chr> <chr> <int> <chr> <POINT [°]>
1 Abuja Nigeria 178462 NGA (7.17 9.18)
2 Accra Ghana 2029143 GHA (-0.2 5.56)
3 Addis Abeba Ethiopia 2823167 ETH (38.74 9.03)
4 Algiers Algeria 2029936 DZA (3.04 36.77)
5 Antananarivo Madagascar 1463754 MDG (47.51 -18.89)
6 Asmara Eritrea 578860 ERI (38.94 15.33)
7 Bamako Mali 1342519 MLI (-7.99 12.65)
8 Bangui Central African Republic 547668 CAF (18.56 4.36)
9 Banjul Gambia 34388 GMB (-16.6 13.46)
10 Bissau Guinea-Bissau 404119 GNB (-15.6 11.87)
# ℹ 41 more rowsggplot(data = africapitals) +
geom_sf(mapping = aes(size = pop))
We can replicate John Snow’s Dot map with the number of deaths per household from the 1854 London cholera outbreak:
cholera_deaths <-
read_rds(here("data/cholera_deaths.rds"))
ggplot(data = cholera_deaths) +
geom_sf(mapping = aes(size = Count), alpha = 0.7)• Visualize airports classified by type using the color argument:
airports <- rnaturalearth::ne_download(scale = 10, type = "airports", returnclass = "sf")
ggplot(data = airports) +
geom_sf(mapping = aes(color = type))
Create a Thematic map with the afriairports object to portrait all its airport locations, using geom_sf(), and color them in relation to the type variable.
# Write and visualize your answer:
q2 <- ggplot(data = afriairports) +
geom_sf(mapping = aes(color = type))• To visualize the scatter of your data and visually scan for clusters.
• Dot maps visualize a shape called Point.
• Collects data that register the locations of random events.
• E.g., geographical coordinates of individuals with a given diagnosis (Figure 3).
• Bothered by having just dots and no geographical context?
• That’s good! We will see how to add those using roads and rivers very soon.
• Thematic maps visualize specific Geometry types:
• Choropleth maps visualize Polygons.
• Dot maps visualize Points.
Which of the following options of Thematic map types:
"choropleth_map""dot_distribution_map"…corresponds to each of these Epidemic map figures seen below?
# Write your answers here as comments:
# Malaria cases in Africa : a
# COVID-19 cases in the world: b
• We learned about Thematic maps,
• How to create them using ggplot2::geom_sf(),
• Which type of Geometry they visualize.
• But, how can we complement Thematic maps with geographic context?
• We are going to learn about how to add Physical features to our maps.
The following team members contributed to this lesson:
Motivated by reproducible science and education
A firm believer in science for good, striving to ally programming, health and education
Passionate about education
Some material in this lesson was adapted from the following sources:
Batra, Neale, et al. (2021). The Epidemiologist R Handbook. Chapter 28: GIS Basics. (2021). Retrieved 01 April 2022, from https://epirhandbook.com/en/gis-basics.html
Lovelace, R., Nowosad, J., & Muenchow, J. Geocomputation with R. Chapter 2: Geographic data in R. (2019). Retrieved 01 April 2022, from https://geocompr.robinlovelace.net/spatial-class.html
Moraga, Paula. Geospatial Health Data: Modeling and Visualization with R-INLA and Shiny. Chapter 2: Spatial data and R packages for mapping. (2019). Retrieved 01 April 2022, from https://www.paulamoraga.com/book-geospatial/sec-spatialdataandCRS.html
Baumer, Benjamin S., Kaplan, Daniel T., and Horton, Nicholas J. Modern Data Science with R. Chapter 17: Working with geospatial data. (2021). Retrieved 05 June 2022, from https://mdsr-book.github.io/mdsr2e/ch-spatial.html
This work is licensed under the Creative Commons Attribution Share Alike license. 
q1 <- ggplot(data = world) +
geom_sf(aes(fill = pop))
q1
q2 <- ggplot(data = afriairports) +
geom_sf(aes(fill = type))
q2
Which of the following options of Thematic map types corresponds to each of these Epidemic map figures? Your answer should be either “choropleth_map” or “dot_distribution_map”.