In jr-packages/jrTidyverse: Jumping Rivers: Getting to Grips with the Tidyverse

We've already covered a fair bit of tidyverse stuff in the Intro to R course. Namely tibbles, dplyr and ggplot2. This chapter just serves as a general recap into what we touched upon.

tibbles

A tibble (or data frame) is how we store a sheet of data. A standard tibble looks like this

data(example, package = "jrTidyverse")
example

dplyr

dplyr is a package for manipulating tibbles. We covered several functions, such as filter() and summarise()

library("dplyr")
## Give me all the rows where gender is "Male"
filter(example, gender == "Male")
## What is the average of the age variable
summarise(example, av_age = mean(age))

Filter the example data set so that the rows left have age > 24.

filter(example, age > 24)

Filter the example data set so that the rows left have respond = TRUE.

filter(example, respond)

We can pass outputs to the first argument of the next function using the piping operator, %>%

# Give me the average age of males
example %>%
  filter(gender == "Male") %>%
  summarise(av_age = mean(age))

Grab the average age of the people who didn't respond

example %>%
  filter(!respond) %>%
  summarise(av_age = mean(age))

The piping operator can be used in any functions, not just dplyr

# Pass 1:5 on the left as the first argument to mean
1:5 %>%
  mean(na.rm = TRUE)
# Explicitly pass 1:5 into the function
mean(1:5, na.rm = TRUE)

We can apply functions to groups within variables using group_by()

# Give me the average age of each group within gender
example %>%
  group_by(gender) %>%
  summarise(av_age = mean(age))

Grab the average age of each grouping within respond

example %>%
  group_by(respond) %>%
  summarise(av_age = mean(age))

ggplot2

ggplot2 is a fantastic package for graphics. The ggplot() function creates a ggplot2 object.

library("ggplot2")
data(movies, package = "jrTidyverse")
ggplot(movies)

To add axes to this we add aesthetics

# visible in figure 2.1
g = ggplot(movies, aes(x = duration, y = rating))
g

Notice we can save plots as variables.

Then to add information onto the graph we use geoms

# figure 2.2
g +
  geom_point()
ggplot(movies, aes(x = rating)) +
  geom_histogram()
ggplot(movies, aes(x = classification)) +
  geom_bar()
ggplot(movies, aes(x = classification, y = rating)) +
  geom_boxplot()

a = g +
  geom_point()
b = ggplot(movies, aes(x = rating)) +
  geom_histogram()
c = ggplot(movies, aes(x = classification)) +
  geom_bar()
d = ggplot(movies, aes(x = classification, y = rating)) +
  geom_boxplot()
gridExtra::grid.arrange(a, b, c, d, ncol = 2, nrow = 2)

Using dplyr, work out the average rating for each year of movies in the data set.

movies %>%
  group_by(year) %>%
  summarise(av_rat = mean(rating))

Use ggplot2 to plot a line graph of the year against average rating. Hint: Use geom_line()

movies %>%
  group_by(year) %>%
  summarise(av_rat = mean(rating)) %>%
  ggplot(aes(x = year, y = av_rat)) +
  geom_line()

Change the axis labels and titles to something more sensible. Remember, to change labels and titles, we use labs()

movies %>%
  group_by(year) %>%
  summarise(av_rat = mean(rating)) %>%
  ggplot(aes(x = year, y = av_rat)) +
  geom_line() +
  labs(x = "Year",
       y = "Average Rating",
       title = "Average rating per year of movies")