Welcome to the tidyverse

An introduction to the tidyverse package
Author

BIOL33031/BIOL65161

What is the tidyverse?

The tidyverse is an R package that makes working with data easier than in ‘base R’, especially for life scientists:

  • Import your lab or field data quickly (e.g. read_csv() brings in spreadsheets).
  • Clean and organise data with simple commands so your tables look exactly how you need them.
  • Summarise measurements by groups (e.g. calculate mean height per treatment).
  • Reshape data between “wide” and “long” formats for different analyses. (More on this later…)
  • Plot effectively with ggplot2 for dissertation or publication-ready figures.

Introducing the tidyverse

The tidyverse is a collection of R packages designed to work together. Examples include:

  • dplyr for data manipulation (filter(), mutate(), summarise()).
  • readr for reading files.
  • tidyr for reshaping data.
  • ggplot2 for plotting.

You can install and load tidyverse with these functions.

install.packages("tidyverse")  # Install first-time only
library(tidyverse)             # Load after installing

Throughout this course, you will start every script that you make by loading tidyverse with library(tidyverse).

Common tidyverse functions

Let’s look at some of the common functions in tidyverse

  • tibble(): creates a ‘data frame’, an object with columns and rows, similar to an Excel file
  • filter(): keeps rows matching certain conditions
  • select(): chooses specific columns from a dataset
  • mutate(): creates or transforms columns
  • group_by(): defines groups of rows to apply operations within each group
  • summarise(): collapses values into summary statistics (often used with group_by)
  • arrange(): sorts rows in ascending or descending order
install.packages("tidyverse")  # Install first-time only
library(tidyverse)             # Load after installing

What is a tibble?

  • A tibble object is a type of data frame
  • It stores data in rows and columns, like a spreadsheet
  • Each column has a “type”, e.g.
    • text <chr>, numbers <dbl>, logical (TRUE/FALSE) <lgl>—more on types later!
# Example of a tibble containing names, heights, and registration status of students
students_tbl <- tibble(
  name = c("Rosie", "Ben", "Chloe", "Dinesh", "Sara", "Zahraa", "Rana", "Alejandro"),
  height_m = c(1.65, 1.80, 1.72, 1.59, 1.54, 1.70, 1.51, 1.67),
  registered = c(TRUE, FALSE, FALSE, FALSE, FALSE, TRUE, TRUE, TRUE)
)

# Print students_tbl to the screen
students_tbl
# A tibble: 8 × 3
  name      height_m registered
  <chr>        <dbl> <lgl>     
1 Rosie         1.65 TRUE      
2 Ben           1.8  FALSE     
3 Chloe         1.72 FALSE     
4 Dinesh        1.59 FALSE     
5 Sara          1.54 FALSE     
6 Zahraa        1.7  TRUE      
7 Rana          1.51 TRUE      
8 Alejandro     1.67 TRUE

Note by default, only the first 10 rows are printed to the screen as a preview.

Revisiting the pipe: %>% or |>

Recall that:

  • The pipe %>% or |> lets you chain operations in a left-to-right flow.
  • The pipe ‘sends’ objects or function outputs to other functions.
  • %>% used to be available only through one of the packages of the tidyverse (dplyr), but recently base R has added |>. These are (currently) interchangable.

The pipe is especially useful in tidyverse workflows.

# Without pipe; confusing and hard to read.
result <- summarise(group_by(data, group), mean_height = mean(height))

# With pipe: read each function line-by-line, much clearer to read
result <- data %>%
  group_by(group) %>%
  summarise(mean_height = mean(height))

# Equivalently with the native pipe in R 4.1+
result <- data |>
  group_by(group) |>
  summarise(mean_height = mean(height))

Using the pipe is the tidy way

Let’s revisit our heights example from the last section to see how the pipe is used in a tidyverse workflow.

# Load the tidyverse package
library(tidyverse)

# Sample heights of students in meters (m)
height_m <- c(
  1.56, 1.63, 1.81, 1.69, 1.77,
  1.73, 1.59, 1.73, 1.65, 1.63,
  1.68, 1.50, 1.80, 1.60, 1.78,
  1.68, 1.84, 1.60, 1.64, 1.71,
  1.43, 1.76, 1.84, 1.79, 1.75,
  1.65, 1.81, 1.78, 1.72, 1.68
)

# Now create a tibble with heights as a column
heights <- tibble(height_m)

# Compute mean and standard deviation of height
heights_summary <- heights |>
  summarise(
    mean_height = mean(height_m),
    sd_height   = sd(height_m)
  )

The pipe helps with reading code because each line does one function.

Here is how you would create a plot using ggplot2 from tidyverse. This will look a bit complicated for now, but we’ll come back to plotting in a later session.

# Plot distribution using tidyverse
# 'Send' the heights tibble to ggplot()
histogram_fancy <- heights |>
  ggplot(aes(x = height_m)) +
  geom_histogram()

# Show the histogram on the screen
histogram_fancy # or print(histogram_fancy)
`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Getting help

Next steps

You’ve now set up R, learned the basic grammar of objects and functions, and run your first tidyverse commands. Make sure you complete the “Check your understanding” quiz at the end of the notes page.

In the next session, we’ll look at how to import and ‘wrangle’ data, explore summary statistics of data further, data visualisation and transformation.

You are now ready to attend the first hands-on practical.

❓Check your understanding

  1. Tibbles
    What does this code create?

    animals <- tibble(
  species = c("Frog", "Beetle", "Bird"),
  count   = c(12, 30, 9)
)

    A tibble (a modern data frame) with two columns: species and count.

  2. Filter
    If animals is the tibble above, what does

    filter(animals, count > 10)

    return?

    A tibble with the rows for Frog (12) and Beetle (30), since their counts are greater than 10.

  3. Mutate
    Add a column converting counts to proportions.
    How would you do this?

    animals |> 
  mutate(prop = count / sum(count))

  4. Group and summarise
    Suppose you have a tibble of plant heights with a column treatment.
    How would you calculate the mean height per treatment?

    plants |>
  group_by(treatment) |>
  summarise(mean_height = mean(height_m))

  5. Pipes
    Rewrite the following without the pipe:

    animals |> summarise(total = sum(count))
    summarise(animals, total = sum(count))