bangladesh: introduction

The bangladesh package provides ready-to-use shapefiles for different administrative regions of Bangladesh (e.g., Division, District, Upazila, and Union). Usually, it is difficult to plot choropleth maps for Bangladesh in R. This package will help users to draw thematic maps of administrative regions of Bangladesh easily as it comes with the sf objects for the boundaries and regions’ names in English. It also provides functions allowing users to efficiently get specific area maps and center coordinates for regions. Users can also search for a specific area and calculate the centroids of those areas.

Getting Started

This packages comes with sf objects for administrative levels 0-4 (Country, Division, District, Upazila, Union). The easiest way to get the shapefile for a level is to is to use get_map() function.

# remotes::install_github("ovirahman/bangladesh")
library(bangladesh)

country <- get_map("country")
division <- get_map("division")
district <- get_map("district")
upazila <- get_map("upazila")
union <- get_map("union")

Plotting Map

To start with we can check the sample function bd_plot() to draw the map of different administrative levels of Bangladesh, which uses tmap, a very flexible and cool package to visualize thematic maps.

bd_plot("country")

bd_plot("division")

bd_plot("district")

We can also plot beautiful interactive maps with this.

Choropleth map with data

Using the tmap package (my favorite for creating thematic maps), we can make cool choropleths, both static and interactive. When plotting mode is chosen as static (plot) it returns a ggplot object, when interactive (view) it returns a leaflet object.

library(tmap)
population <- bangladesh::pop_district_2011[, c("district", "population")]
district <- get_map("district")

map_data <- dplyr::left_join(district, population, by = c("District" = "district"))

map <- tm_shape(map_data) + 
  tm_polygons("population",id = "District",palette = "Reds", title = "Population") +
  tm_style("cobalt")+
  tm_layout(
    "Bangladesh District Wise Population Map\nSource: BBS",
    title.position = c("left", "bottom"),
    legend.position = c("right", "top")
    )

tmap::tmap_mode("plot")
map

To make it interactive we can do the following:

Using ggplot2 and leaflet

We can also use ggplot2 and leaflet to draw customized choropleths with the sf objects provided in bangladesh package.

library(ggplot2)
ggplot(data = map_data) +
  geom_sf(aes(fill = population))+
  theme_void()+
  viridis::scale_fill_viridis(trans = "log", name="Population", labels = scales::unit_format(unit = "M", scale = 1e-6)) +
  labs(
    title = "Bangladesh Population Map",
    subtitle = "Population & Housing Census 2011",
    caption = "Data Source: BBS"
  )

Other useful functions

It is also possible to get the approximate center points (centroids) of administrative regions easily

by using get_coordinates() function in bangladesh package.

division_map <- get_map("division")
division_centroids <- bangladesh::get_coordinates(level = "division")
knitr::kable(division_centroids, format = "html")
Division lat lon
Barisal 22.41889 90.34684
Chittagong 22.70692 91.73546
Dhaka 23.83870 90.24064
Khulna 22.91367 89.29437
Mymensingh 24.84675 90.38088
Rajshahi 24.58846 89.04540
Rangpur 25.77920 89.05685
Sylhet 24.71515 91.66400
ggplot(data = division_map) +
  geom_sf() +
  geom_sf_label(aes(label = Division)) +
  geom_point(data = division_centroids, x = division_centroids$lon, y = division_centroids$lat, col = "red", size = 3) +
  xlab("")+ ylab("")+
  theme_minimal()

Suppose someone needs to plot partially a single or selected number of divisions instead of whole country map, in that case the function get_divisions() might be beneficial.


sylhet <- get_divisions(divisions = "Sylhet",level =  "upazila")
# single division
ggplot(data = sylhet) +
  geom_sf() +
  xlab("")+ ylab("")+
  theme_minimal()


#multiple division
sylhet_chittagong_dhaka <- get_divisions(divisions = c("Sylhet", "Chittagong", "Dhaka"),level =  "upazila")
ggplot(data = sylhet_chittagong_dhaka) +
  geom_sf() +
  xlab("")+ ylab("")+
  theme_minimal()

To search for an area within the provided names for administrative regions we can apply the bd_search() function. The result can also include centroids for those areas.

amtali <- bd_search("amtali", level = "union", as.is = TRUE, coordinates = TRUE)
knitr::kable(amtali, format = "html")
Division District Upazila Union lat lon
Barisal Barguna Amtali Amtali 22.07556 90.24699
Chittagong Rangamati Baghai Chhari Amtali 23.10559 92.18832
Dhaka Gopalganj Kotali Para Amtali 22.98264 90.03070
Chittagong Khagrachhari Matiranga Amtali 23.11701 91.88545
ggplot(bangladesh::map_union) +
  geom_sf() +
  geom_point(data = amtali, x = amtali$lon, y = amtali$lat, col = "red", size = 3)