Egypt Choropleth map in R using GGPLOT
Our youtube channel
Video link https://youtu.be/MZ6Ki3eHQOg
Our youtube channel has lots of videos on data visualisation in r.
Visit our youtube channel https://www.youtube.com/c/TechAnswers88
Simple way to create a chropleth map of Egypt.
Download the shape files for the boundaries
We would need the shape files which will be used to draw the boundaries which are needed to make a choropleth maps. The following website has various boundary files for Egypt
https://data.humdata.org/dataset/cod-ab-egy?
Here is a link of the first adm1 level data files
Here is a link of the first adm2 level data files Which has suburb level details under each area
Extract the data from each zip file and we will be mainly interested in the .shp file. Do not delete any files in the zip folder.
R packages needed
If you do not have these packages then please install them using the install.packages command or follow the install option in your RStudio GUI.
Here are the install.packages commands foe each package.
install.packages(‘ggplot2’)
install.packages(‘sf’)
install.packages(‘rvest’)
install.packages(‘dplyr’)
install.packages(‘viridis’)
install.packages(‘ggrepel’)
install.packages(‘ggthemes’)
Step 1 Import the packages
library(ggplot2)
library(sf)
library(rvest)
library(dplyr)
library(viridis)
library(ggrepel)
library(ggthemes)
shp1 <- read_sf("d:\\tmp\\egy_admbnda_adm1_capmas_20170421\\egy_admbnda_adm1_capmas_20170421.shp")
shp2 <- read_sf("d:\\tmp\\egy_admbnda_adm2_capmas_20170421\\egy_admbnda_adm2_capmas_20170421.shp")
Our first chart
very basic chart to see how the plot works
pl <- ggplot(shp1)
pl <- pl+ geom_sf()
pl
The shapefile shows 27 governorates areas
See the wikipedia for more information on the Provincial divisions
List the areas in our shape file as below
str(shp1)
FALSE sf [27 x 17] (S3: sf/tbl_df/tbl/data.frame)
FALSE $ ADM1_EN : chr [1:27] "Alexandria" "Assiut" "Aswan" "Behera" ...
FALSE $ ADM1_AR : chr [1:27] "<U+0627><U+0644><U+0627><U+0633><U+0643><U+0646><U+062F><U+0631><U+064A><U+0629>\n" "<U+0623><U+0633><U+064A><U+0648><U+0637>" "<U+0623><U+0633><U+0648><U+0627><U+0646>" "<U+0627><U+0644><U+0628><U+062D><U+064A><U+0631><U+0629>\n" ...
FALSE $ ADM1_PCODE: chr [1:27] "EG02" "EG25" "EG28" "EG18" ...
FALSE $ ADM1_REF : chr [1:27] "Alexandria" "Assiut" "Aswan" "Behera" ...
FALSE $ ADM1ALT1EN: chr [1:27] NA NA NA NA ...
FALSE $ ADM1ALT2EN: chr [1:27] NA NA NA NA ...
FALSE $ ADM1ALT1AR: chr [1:27] NA NA NA NA ...
FALSE $ ADM1ALT2AR: chr [1:27] NA NA NA NA ...
FALSE $ ADM0_EN : chr [1:27] "Egypt" "Egypt" "Egypt" "Egypt" ...
FALSE $ ADM0_AR : chr [1:27] "<U+0645><U+0650><U+0635><U+0631>" "<U+0645><U+0650><U+0635><U+0631>" "<U+0645><U+0650><U+0635><U+0631>" "<U+0645><U+0650><U+0635><U+0631>" ...
FALSE $ ADM0_PCODE: chr [1:27] "EG" "EG" "EG" "EG" ...
FALSE $ date : Date[1:27], format: "2006-01-01" "2006-01-01" ...
FALSE $ validOn : Date[1:27], format: "2017-04-21" "2017-04-21" ...
FALSE $ validTo : Date[1:27], format: NA NA ...
FALSE $ Shape_Leng: num [1:27] 4.27 8 38.13 6.51 7.44 ...
FALSE $ Shape_Area: num [1:27] 0.237 1.498 5.102 1.068 0.981 ...
FALSE $ geometry :sfc_MULTIPOLYGON of length 27; first list element: List of 1
FALSE ..$ :List of 2
FALSE .. ..$ : num [1:50689, 1:2] 30.1 30.1 30.1 30.1 30.1 ...
FALSE .. ..$ : num [1:99, 1:2] 29.9 29.9 29.9 29.9 29.9 ...
FALSE ..- attr(*, "class")= chr [1:3] "XY" "MULTIPOLYGON" "sfg"
FALSE - attr(*, "sf_column")= chr "geometry"
FALSE - attr(*, "agr")= Factor w/ 3 levels "constant","aggregate",..: NA NA NA NA NA NA NA NA NA NA ...
FALSE ..- attr(*, "names")= chr [1:16] "ADM1_EN" "ADM1_AR" "ADM1_PCODE" "ADM1_REF" ...Add some fill colour and the labels in the map
pl <- ggplot(shp1)
pl <- pl+ geom_sf(aes(fill= ADM1_REF) , color = "white")
pl <- pl + ggthemes::theme_map()
pl <- pl + theme(legend.position = "none")
pl <- pl + ggrepel::geom_label_repel(aes(label = ADM1_REF
, geometry = geometry)
, stat= "sf_coordinates"
, box.padding = 0.5
, max.overlaps = Inf
, size = 4)
pl
You can use text instead of labels
pl <- ggplot(shp1)
pl <- pl+ geom_sf(aes(fill= ADM1_REF) , color = "white")
pl <- pl + ggthemes::theme_map()
pl <- pl + theme(legend.position = "none")
pl <- pl + ggrepel::geom_text_repel(aes(label = ADM1_REF
, geometry = geometry)
, stat= "sf_coordinates"
, box.padding = 0.5
, max.overlaps = Inf
, size = 4)
pl
Add some data into the shp file
We will create a new column in our shp1 file as shown below so that we can generate a heatmp or Choropleth chart As there are 27 governorates area we add 27 values in our sales column.
shp1$sales <- c(7261 ,18100 ,31282 , 45900, 7170, 6110, 987000, 71700, 61110, 987000
, 143000 , 162000 ,8200 , 1990, 65700, 36876, 89000, 45000, 35000, 870000
, 19900 , 56000 ,3400 , 2800, 6500 , 5600 , 45000)
pl <- ggplot(shp1)
pl <- pl+ geom_sf(aes(fill= sales) , color = "white")
pl <- pl + ggthemes::theme_map()
pl <- pl + scale_fill_viridis_c(option = "D")
pl <- pl + ggrepel::geom_text_repel(aes(label = ADM1_REF
, geometry = geometry)
, stat= "sf_coordinates"
, box.padding = 0.5
, max.overlaps = Inf
, size = 3)
pl
Plotting the second shape file which has smaller areas
Too many labels, so we should avoid the labelling or do only selective labelling.
pl <- ggplot(shp2)
pl <- pl+ geom_sf(aes(fill= ADM2_REF) , color = "white")
pl <- pl + ggthemes::theme_map()
pl <- pl + theme(legend.position = "none")
pl <- pl + ggrepel::geom_label_repel(aes(label = ADM2_REF
, geometry = geometry)
, stat= "sf_coordinates"
, box.padding = 0.5
, max.overlaps = Inf
, size = 4)
pl <- pl + scale_fill_viridis_d(option = "C")
pl
Example of selective labelling
See that we have used a defined a filter in the label, so that we can only show the labels for the areas under Aswan.
ggrepel::geom_label_repel(data = shp2%>%dplyr::filter(ADM1_EN %in% c(‘Aswan’))
#str(shp2)
#(shp2$ADM1_EN)
pl <- ggplot(shp2)
pl <- pl+ geom_sf(aes(fill= ADM2_REF) , color = "white")
pl <- pl + ggthemes::theme_map()
pl <- pl + theme(legend.position = "none")
pl <- pl + ggrepel::geom_label_repel(data = shp2%>%dplyr::filter(ADM1_EN %in% c('Aswan'))
, aes(label = ADM2_REF
, geometry = geometry)
, stat= "sf_coordinates"
, box.padding = 0.5
, max.overlaps = Inf
, size = 4)
pl <- pl + scale_fill_viridis_d(option = "C")
pl
Subscribe to our channel for various data visualisation and statistics videos in R.