#### import INTERNET dataset
library(readxl)
internet <- read_excel("../state-data.xlsx", "internet")

# tidy the dataset - denote NUTS3 region id and gender as factors
internet$nuts3 <- as.factor(internet$nuts3)
internet$gndr <- as.factor(internet$gndr)

# attach the dataset
attach(internet)

### compute some descriptive statistics:
# mean, variance, standard deviation
mean(netustm)
var(netustm)
sd(netustm)

# similar statistics per regions 
# using dplyr package (and its "pipe" %>% operator)
library(dplyr)
internet %>% 
  group_by(nuts3) %>%
  summarize(
    count = n(),
    mean = mean(netustm),
    var = var(netustm),
    sd = var(netustm)
  )

# other usual descriptive statistics
min(netustm)
max(netustm)
median(netustm)
quantile(netustm, probs = c(0.25, 0.75))
IQR(netustm)        # inter-quartile range
range(netustm)
# or shortly (Tukey's statistics)
fivenum(netustm)
summary(netustm)

### graphical visualizations
boxplot(netustm)
hist(netustm)
plot(density(netustm))

# the same using ggplot2
library(ggplot2)
ggplot(data = internet, aes(y = netustm)) +
  geom_boxplot()
ggplot(data = internet, aes(x = netustm)) +
  geom_histogram(binwidth = 5, color = "white")
ggplot(data = internet, aes(x = netustm)) +
  geom_density(fill = "blue")

# some regional statistics
ggplot(data = internet, aes(x = nuts3, y = netustm)) +
  geom_boxplot() + 
  theme(axis.text.x = element_text(angle = 90, hjust = 1))
ggplot(data = internet, aes(x = netustm)) +
  geom_histogram(binwidth = 5, color = "white") + 
  facet_wrap(~ nuts3)