R Video Tutorial #6: Violin Plots: What are they good for?

The code:

rm(list=ls())
setwd("")

library(data.table)
library(tidyverse)

#reading in the data with fread from data.table
data <- fread("human-freedom-index-data-update.csv")

#looking at the column names
colnames(data)

#looking at the first few column names
colnames(data)[1:10]

#selecting columns we need
data <- select(data, year, countries, region, hf_score)

#rename certain column names
colnames(data)[2] <- "country"

#let's look at the data one more time
data

#combining regions
unique(data$region)

data[region %like% "Europe", region_new := "Europe"]
data[region %like% "Africa", region_new := "Africa & Middle East"]
data[region %like% "Asia"] %>% select(region) %>% unique()
data[region %like% "Asia", region_new := "Asia"]
data[region %like% "Ocean" | region %like% "North Amer", region_new := "North America & Oceania"]
data[region %like% "Latin", region_new := "Latin America"]

#check if we got every region
data[is.na(data$region_new)]

#let's look at our new regions
unique(data$region_new)

#transforming new region variable to factor
data$region_new <- factor(data$region_new)

#finally we filter for 2016
data <- data[year == 2016]

#lets start nice and easy
plot <- ggplot(data, aes(x = region_new, y = hf_score, fill = region_new)) + 
  geom_violin(trim = FALSE) +
  #Let's add the mean and the standard deviation
  #now let's add the actual observations 
  #geom_point(aes(fill = region_new), position = "jitter")
  geom_jitter(aes(fill = region_new), width = 0.2, shape = 21) +
  stat_summary(fun.data=mean_sdl, fun.args = list(mult=1),
               geom="pointrange", shape = 17, color = "white") +
  #Let's add some titles
  ggtitle("Human Freedom Index") +
  theme(plot.title = element_text(hjust = 0.5)) +
  labs(x = "Region",
       y = "Human Freedom Index",
       fill = "Region")

ggsave("freedom.png", plot, width = 10, height = 7)