Wikipedia World Population Data Visualization with maptools and ggplot in R

In a previous R Tutorial, Web Scraping Wikipedia World Population with rvest() in R we were able to scrape global population from Wikipedia.

Since we were working with 10 countries in the previous tutorial, we will continue to do so moving forward.

Install and Load Packages

Below are the packages and libraries that we will need to load to complete this tutorial:

Download and Load the Global Population Wikipedia Dataset

Now that the packages installed and loaded, let’s take a look at the data by importing the data as read.csv(). The dataset is available for download on the dataset page or directly from here Global Population Wikipedia Export – global_population.csv

View the Global Population Wikipedia Dataset

head() function

Let’s take a look at the population data for the first 10 lines.

lapply() function

The data that we will mostly use are Country, Population, and Percentage. Let’s confirm that Population and Percentage are numeric and not character by using lapply() function.

Clean-up Population and Percentage

In order to convert Population and Percentage to numeric, we must remove the commas, percentage signs, and decimals.

gsub() function

The below gsub() will remove all commas(,) within the population column and then re-run the previous head() function.

Now that the population column is now confirmed, so removing the percent sign (%) and convert the percentage to decimal will be complete.

as.numeric() function

The below as.numeric(gsub()) function will remove the percentage sign (%) and convert the percentage to a decimal by dividing the number by 100.

Now that both the Population and Percentage column are now cleaned up, we can convert both of these columns to numeric. Once we convert the columns to numeric, we will re-verify the columns were converted by running the lapply() function.

Top 10 World Population Plot

Below is a basic ggplot() + geom_boxplot() to view the population total for the top 10 countries.

Also, the below function will stop scientific notation and provide the full population number.

And the below will return scientific notation.

reorder() the countries in descending order

In addition to disabling scientific notation, we should plot the countries by population in descending order (most populous to least populous). In a previous tutorial, we used a factor(levels)) functions to produce this result by manually inputting the countries in descending order. The below reorder() function can achieve the same output with a fraction of the time. Below we will create two plots, the first will be a bar chart without re-ordering the countries by population. The second plot will re-order the countries based on population size. Lastly, we will use grid.arrage() to output both plots into a single output. The grid package provides low-level functions to create graphical objects (grobs), and position them on a page in specific viewports.

We can also use reorder(Category, Count) to have the population in ascending order(least populous to most populous).

Now that we have created a plot by descending order of population, we can label the x-axis and y-axis, scale the y-axis, and angle the xlab so the countries are readable.

Our population plot looks presentable, but we could add one additional effect for the plot by adding fill for the countries.

Could we Plot the Countries on a Map?

We can plot the countries on a map with the wrld_simpl (Simplified World Country Polygons) by installing the library maptools. The object loaded is a SpatialPolygonsDataFrame object containing a slightly modified version of Bjoern Sandvik’s improved version of world\_borders.zip – TM\_WORLD\_BORDERS\_SIMPL-0.2.zip dataset from the Mapping Hacks geodata site.

Before moving forward, install and load the library as shown below.

Now once we load wrld_simpl object, the object will consist of a Large SpatialPolygonsDataFrame with 5 slots; the data clot contains a data.frame with 246 obs. of 11 variables. Now let’s load the object and view wrld_simpl as a plot.

Now let’s take a look at wrld_simpl and see how we can connect this object to our top 10 global data. We must see how to connect the countries within our data.frame pop__top into the wrld_simpl. The easiest way to view the variable is by using the summary() function.

With the above output, we can easily view the variables and choose the variable that will meet the needs of our country name. With the above output, let’s view our countries by name again.

As we can see from the above outputs, the NAME variable for wrld_simple matches our Country variable for data.frame pop_top. We will need to create a variable for the countries and match our variable Country from pop_top with the variable NAME. In addition, the plot will produce the map from above with our top 10 countries mapped in blue.

The above is color maps each of our countries into the blue areas that are mapped. We know that the above mapped correctly based on our matching of wrld_simpl@data$NAME to our variable countries. However, let’s take a look at the shades of blue by installing and loading the library package RColorBrewer. Once the package is loaded, run the below for 9 shades of blue.

 

Matching Color Population Against Country

We’re making progress on the map, but we must matching color population to our country by creating two additional variables.

What are we even doing below?

1. Creating color_population variable based on different shades of blue for each input of unique population
2. Creating color_country by matching pop_top$Country and wrld_simpl@data$NAME against color_population variable

Now let’s take a look at the map with the blue shading for each country.

Feel free to work with this data and create other plots to visualize the data in other packages. Another well-known package for mapping is rworldmap. Take a look at this package and see what you can create!