Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
Makeovermonday is a weekly social data project with the intention to rework some random chosen chart. A new challenge is posted every week on the data set page. Although it is more focused to the Tableau community, I took the challenge to rework the chart with R. This is the challenge from last week which holds a dataset with a ratio of internet users per 100 people for most of the Countries.
This is a step-by-step explanation of my own process, so let’s start:
Startingpoint
First, look a little bit at the data and make a simple line/dot plot.
# Load library
library(readxl)
library(ggplot2)
library(tidyverse)
library(countrycode)
library(rworldmap)
library(leaflet)
library(viridis)
library(ggmap)
# ingest data and add usefull labels
internet_set <- read_excel("datasets/Internet_users_per_100_ppl.xlsx")
names(internet_set) <- c("Country", "Year", "Users")
# look at first rows
head(internet_set)
## # A tibble: 6 x 3
## Country Year Users
## <chr> <dbl> <dbl>
## 1 Iceland 1990 0.00000000
## 2 Luxembourg 1990 0.00000000
## 3 Andorra 1990 0.00000000
## 4 Norway 1990 0.70729945
## 5 Liechtenstein 1990 0.00000000
## 6 Denmark 1990 0.09727727
# Plot basic chart
ggplot(internet_set, aes(x = factor(Year), y = Users, group = Country)) + geom_line() + geom_point()
The dataset includes mostly 0 or very low values for the year 1990 so lets remove this year. Choosing year 2000 is a better starting point for the visualization.
# remove irrelevant year data (1990) # also round the ration value to 2 decimals (%) internet_set <- filter(internet_set, Year != 1990) %>% mutate(Users = round(Users, 2)) #plot ggplot(internet_set, aes(x = factor(Year), y = Users, group = Country)) + geom_line() + geom_point()
OK, this seems to be a better starting point that actually make sense. Next, it should be nice to a distinction between continents of the countries. I assume that more wealthy countries are also noticeable at continent level. Wealthy countries should be visual at the top of the chart and less wealthy countries more at the bottom.
# add continent to countries internet_set$continent <- countrycode(sourcevar = internet_set$Country, origin = "country.name", destination = "continent" , origin_regex = TRUE) # plot ggplot(internet_set, aes(x = factor(Year), y = Users, group = Country, colour = continent)) + geom_line() + geom_point() + theme(legend.position = "bottom")
We can already deduce something from this, but it is still a bunch of (overlapping) lines. I deciced to split up the plot into facets for each continent.
# plot basic chart ggplot(internet_set, aes(x = factor(Year), y = Users, group = Country, colour = continent)) + geom_line() + geom_point() + facet_wrap(~continent, nrow = 1)
We can see that our dataset hold NA’s for some countries. Just erase all NA’S in the set should be fine at this point. It’s getting better, but there is still to much detail.
# remove NA's in the dataset internet_set <- na.omit(internet_set) # plot basic chart ggplot(internet_set, aes(x = factor(Year), y = Users, group = Country, colour = continent)) + geom_line() + geom_point() + facet_wrap(~continent, nrow = 1)
filter(internet_set, Year %in% c(2000, 2015), continent == "Europe") %>% # plot ggplot(., aes(x = factor(Year), y = Users, group = Country, colour = continent)) + geom_line() + geom_point() + theme(legend.position = "bottom")
Little parrallel track on the plot
Instead of using lines/dot plot, why not scatter it per year and again color it per continent. But this plot doesn’t really add support to understand the information.
# ggplot(internet_set, aes(x = factor(Year), y = Users, group = Country, colour = continent)) + geom_jitter(aes(size = Users), alpha = 0.4, width = 0.3) + theme_minimal() + theme(legend.position = "top")
Final plot
To continue with the intial setup.
- lets minimize the included year: only take 2000, 2010, 2015
- add some nice colors
- remove / add usefull axis titles
- add titles + subtitles
ggplot_fun <- function(dataset) {
dataset %>%
ggplot(., aes(x = factor(Year), y = Users, group = Country, colour = continent)) + geom_blank() +
annotate("rect", xmin=1.5, xmax=2.5, ymin=0, ymax=Inf, alpha=0.2, fill="#DADFE1") +
geom_line(colour = "gray95", position = position_dodge(0.8)) +
geom_point(alpha = 0.7, position = position_dodge(0.8)) +
geom_vline(xintercept = c(1.5,2.5), colour = "white", size = 0.3)+
theme_minimal() +
theme(legend.position = "none",
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
panel.grid.major.y = element_line(colour = "gray95"),
panel.spacing.x = unit(1.5, "lines")
) +
scale_y_continuous(breaks = c(0,10,20,30,40,50,60,70,80,90,100)) +
scale_color_viridis(discrete=TRUE) +
facet_wrap(~continent, nrow = 1) +
ylab("Internetuser per 100 citizens") + xlab("") +
ggtitle(label = "Internet users per 100 people", subtitle = "Internet users are people with access to the worldwide network.")
}
ggplot_fun(filter(na.omit(internet_set), Year %in% c(2000, 2010, 2015)))
Lets focus more on the overall evalution and only includes the first year 2000 and the last 2015
ggplot_fun(filter(na.omit(internet_set), Year %in% c(2000, 2015)))
Another appraoch
Only focus on one continent, for example Europe and see the country’s progress from 2000 to 2015.
ggplot_fun2 <- function(dataset) {
dataset %>%
ggplot(., aes(x = factor(Year), y = Users, group = Country, colour = continent)) + geom_blank() +
annotate("rect", xmin=1.5, xmax=2.5, ymin=0, ymax=Inf, alpha=0.2, fill="#DADFE1") +
geom_line(colour = "gray95") +
geom_point(alpha = 0.7, size = 3) +
geom_vline(xintercept = c(1.5,2.5), colour = "white", size = 0.3)+
theme_minimal() +
theme(legend.position = "none",
panel.grid.minor = element_blank(),
panel.grid.major = element_blank(),
panel.grid.major.y = element_line(colour = "gray95"),
panel.spacing.x = unit(1.5, "lines")
) +
scale_y_continuous(breaks = c(0,10,20,30,40,50,60,70,80,90,100)) +
scale_color_viridis(discrete=TRUE) +
facet_wrap(~continent, nrow = 1) +
ylab("Internetuser per 100 citizens") + xlab("") +
ggtitle(label = "Internet users per 100 people", subtitle = "Internet users are people with access to the worldwide network.")
}
ggplot_fun2(filter(na.omit(internet_set), Year %in% c(2000, 2015), continent == "Europe"))
I should be nice to explore this graph interactively. Lets make a dynamic or intertive plot with htmlwidget. Lets pick library(highcharter) for this. The plot is definitely not perfect. I want to highlight one indidivual country, whereas the rest should color gray.
library(highcharter)
library(htmlwidgets)
a <- filter(na.omit(internet_set), Year %in% c(2000, 2015), continent == "Europe") %>% droplevels() %>%
hchart(., "scatter", hcaes(x = factor(Year), y = Users, group = Country)) %>%
hc_title(text = "Internet users per 100 people",
margin = 20, align = "left",
style = list(color = "#2b908f", useHTML = TRUE, Weight="bold")) %>%
hc_subtitle(text = "Internet users are people with access to the worldwide network.",
align = "left",
style = list(color = "#90ed7d")) %>%
hc_yAxis( lineColor = 'transparent',
tickLength = 0) %>%
hc_xAxis( lineColor ='transparent',
tickLength= 0,
title = list(text = "Year")) %>%
hc_legend(enabled = FALSE) %>%
hc_exporting(enabled = TRUE) %>%
hc_plotOptions(scatter = list(lineWidth = 2))
a
Back to the original
R has also some nice packages to make interactive geo-spatial plots such as the original chart, such as the leaflet package or highcharter. First, a SpatialPolygonsDataFrame must be created that hold the countries shapes and the data that we want to visualize.
# get iso3 codes
# this will match with reg expression. a SpatialPolygonsDataFrame will be created by merging with this iso3 column.
internet_set$iso3c <- countrycode(sourcevar = internet_set$Country, origin = "country.name", destination = "iso3c" , origin_regex = TRUE)
pal <- colorQuantile(palette = "viridis", domain = internet_set$Users, 5)
# plot function
leaflet_plot <- function(dataset, year) {
bins <- c(0, 20, 40, 60, 80, 100)
pal <- colorBin(palette = "viridis", domain = dataset$Users, bins = bins)
user_values <- dataset$Users
# apply filter
internet_set_filtered <- filter(internet_set, Year == year) %>% na.omit()
# get SpatialPolygonsDataFrame
sp_internet <- joinCountryData2Map(dF = internet_set_filtered, joinCode = "ISO3", nameJoinColumn = "iso3c", nameCountryColumn = "Country2")
# Make colorpalette for leaflet
leaflet(sp_internet, width = "100%") %>%
setView(lat = 50.85045, lng = 4.3487800, zoom = 3) %>%
addPolygons(
fillColor = ~pal(Users),
label = ~paste(Country, Users),
smoothFactor = 0.5,
weight = 1,
opacity = 1,
color = "white",
dashArray = "3",
fillOpacity = 0.7,
highlight = highlightOptions(
weight = 3,
color = "#666",
dashArray = "",
fillOpacity = 0.7,
bringToFront = TRUE),
labelOptions = labelOptions(direction = "auto")) %>%
addLegend("topright", pal = pal, values = ~user_values, position = "bottomleft",
title = "Internetusers per 100 citizens",
# labFormat = labelFormat(prefix = "%"),
opacity = 1)
}
leaflet_plot(internet_set, 2015)
## 191 codes from your data successfully matched countries in the map
## 0 codes from your data failed to match with a country code in the map
## 52 codes from the map weren't represented in your data
R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.