Practice Session & Exercise

Introduction to R and Basic Programming Concepts

Author

Affiliation

Bioinformatics Core Facility CECAD

1 Intermediate R-Course 2026

2 Session 3 Practice: ggplot2

This one-hour practice session follows the slides and the Kahoot quiz. Try each task before you open the solution. The exercises use iris, the built-in plant-physiology dataset CO2, and ggplot2.

Not everyone works through exercises at the same pace, so each one below is flagged — but the exercises build on each other, so everyone still works through them in order, together as a group. The flags aren’t an invitation to skip ahead; they tell you what to expect, and what to do if you’re falling behind:

• 🟢 Easy — everyone should get through these by writing the code themselves.
• 🟡 Medium — a bit harder, or an optional add-on; keep typing if you’re keeping pace, but copying from the solution is fine too.
• 🔴 Challenge — the most demanding exercises, including today’s only new programming idea (writing your own function). If you’re falling behind, copy the code from the solution instead of typing it yourself, then move on together with the group.

2.1 Session 3

2.1.1 Exercise 1: Fixed versus mapped colour 🟢 Core

Start with this scatter plot:

p <- ggplot(iris, aes(Sepal.Length, Sepal.Width)) +
  geom_point()
print(p)

Create two named versions:

• p_grouped: map Species to colour so a legend appears.
• p_fixed: make every point one fixed blue colour.

Print both versions. Which one has a legend, and why?

Solution 1.

p_grouped <- p + aes(colour = Species)

p_fixed <- ggplot(iris, aes(Sepal.Length, Sepal.Width)) +
  geom_point(colour = "#0072B2")

print(p_grouped)

print(p_fixed)

2.1.2 Exercise 2: Extend a plot object without overwriting it 🟢 Core

Using p_grouped from Exercise 1, create three named versions:

• p_trend: add one linear trend per species with geom_smooth(method = "lm", se = FALSE).
• p_panels: split the plot into species panels with facet_wrap(~ Species).
• p_report: start from p_panels, then add labels and theme_classic().

Print p_report and then print p_grouped. Is the original still unchanged?

Solution 2.

p_trend <- p_grouped +
  geom_smooth(method = "lm", se = FALSE)

p_panels <- p_grouped +
  facet_wrap(~ Species)

p_report <- p_panels +
  labs(
    title = "Iris sepal measurements",
    x = "Sepal length (cm)",
    y = "Sepal width (cm)",
    colour = "Species"
  ) +
  theme_classic()

print(p_report)

print(p_grouped)

2.1.3 Exercise 3: Put plot objects in lists and functions 🔴 Challenge

Keep the versions from Exercise 2 in a list called plots, with elements named grouped, trend, panels, and report.

Then write render_plot(plot, file) using ggsave(). Use it to save the report version as iris_report.png.

Hint: A named list works just like list(a = 1, b = 2). Here it’s list(grouped = p_grouped, trend = p_trend, panels = p_panels, report = p_report). Writing your own function just needs a name, the inputs in parentheses, and the steps in { }: render_plot <- function(plot, file) { ggsave(filename = file, plot = plot, ...) }. Once defined, you call it exactly like any built-in function, e.g. render_plot(plots$report, "iris_report.png"). So the rest is mostly copying the ggsave() line you already used earlier into the function body.

This should make the exercise a lot easier — the list and the function body are both short, familiar pieces, just assembled in a new shape.

Solution 3.

plots <- list(
  grouped = p_grouped,
  trend = p_trend,
  panels = p_panels,
  report = p_report
)

render_plot <- function(plot, file) {
  ggsave(
    filename = file,
    plot = plot,
    width = 8,
    height = 5,
    units = "in",
    dpi = 300
  )
  invisible(file)
}

dir.create("ggplot2_practice_output", showWarnings = FALSE)
render_plot(
  plots$report,
  "ggplot2_practice_output/iris_report.png"
)

2.1.4 Exercise 4: Choose the geom and add observations 🟡 Stretch

Using iris, make three complementary plots:

1. A count of flowers in each species with geom_bar().
2. A histogram of Petal.Length. Compare binwidth = 0.5 and binwidth = 1.
3. A boxplot of Petal.Length by species with jittered observations.

Optional: create a count table with dplyr::count() and reproduce the first plot with geom_col(). What changes when the y values are supplied?

Hint: The first three plots use exactly the geoms and columns named in the task, the same way the slides did. For the optional part, count(iris, Species) counts rows per species and returns a tibble with two columns already named Species and n, ready to map with aes(Species, n) and geom_col(). So the optional part is really just a small variation on p_bar, not a new plot from scratch.

This should make the optional part a lot easier — it’s only a renaming step before the geom_col() you already know.

Solution 4.

p_bar <- ggplot(iris, aes(Species)) +
  geom_bar() +
  labs(x = "Species", y = "Number of flowers")

sp_counts <- count(iris, Species)
p_col <- ggplot(sp_counts, aes(Species, n)) +
  geom_col() +
  labs(x = "Species", y = "Number of flowers")

p_hist <- ggplot(iris, aes(Petal.Length)) +
  geom_histogram(binwidth = 0.5, boundary = 0, colour = "white") +
  labs(x = "Petal length (cm)", y = "Number of flowers")

p_box <- ggplot(iris, aes(Species, Petal.Length)) +
  geom_boxplot(outlier.shape = NA) +
  geom_jitter(width = 0.12, height = 0, alpha = 0.7) +
  labs(x = "Species", y = "Petal length (cm)")

print(p_bar)

print(p_col)

print(p_hist)

print(p_box)

2.1.5 Exercise 5: Capstone figure and reproducible outputs 🔴 Challenge

iris only gives one grouping variable (Species), so for a figure that needs two we switch to another built-in biological dataset: CO2, CO2 uptake rates in chilled and non-chilled grass plants from Quebec and Mississippi (see ?CO2).

Build one figure in named stages. The final plot should:

• put conc (ambient CO2 concentration) on x and uptake on y;
• colour points by Type (Quebec vs. Mississippi);
• add one linear trend per Type;
• split into panels by Treatment (chilled vs. nonchilled);
• include readable labels, a title, and a clean theme.

Keep p_points, p_fits, and p_final as separate objects. Print all three and describe what each added layer contributes.

Finally:

• save p_final as co2_summary.png at 300 dpi;
• save the plot object as co2_plot.rds;
• restore it with readRDS() and add a presentation theme.

Hint: Build it in the same three stages the task lists, one bullet at a time: first p_points <- ggplot(CO2, aes(conc, uptake, colour = Type)) + geom_point(); then p_fits <- p_points + geom_smooth(method = "lm", se = FALSE); then p_final <- p_fits + facet_wrap(~ Treatment) + labs(...) + theme_minimal(). Each stage only adds one more requirement on top of the previous plot object, so you never have to write the whole figure in a single step. Saving and reloading uses the exact same ggsave()/saveRDS()/readRDS() calls as Exercise 3.

This should make the capstone a lot easier — build and print one stage at a time instead of aiming for the finished figure right away.

Solution 5.

p_points <- ggplot(CO2, aes(conc, uptake, colour = Type)) +
  geom_point(size = 2.5)

p_fits <- p_points +
  geom_smooth(method = "lm", se = FALSE)

p_final <- p_fits +
  facet_wrap(~ Treatment) +
  labs(
    title = "CO2 uptake versus ambient concentration",
    x = "Ambient CO2 concentration (mL/L)",
    y = "CO2 uptake rate (\u03bcmol/m\u00b2/s)",
    colour = "Origin"
  ) +
  theme_minimal(base_size = 12)

print(p_points)

print(p_fits)

print(p_final)

dir.create("ggplot2_practice_output", showWarnings = FALSE)
ggsave(
  "ggplot2_practice_output/co2_summary.png",
  plot = p_final,
  width = 8,
  height = 5,
  units = "in",
  dpi = 300
)

saveRDS(p_final, "ggplot2_practice_output/co2_plot.rds")

restored <- readRDS("ggplot2_practice_output/co2_plot.rds")
restored_large <- restored +
  theme(
    legend.position = "bottom",
    text = element_text(size = 16)
  )
print(restored_large)

2.2 Exit ticket

1. What belongs inside aes(), and what belongs outside it?
2. Which object is your reusable starting point?
3. What is the difference between the PNG and the RDS file?

Solution 6.

1. Values that come from a data column (e.g. colour = Species) go inside aes(). Fixed, one-off settings that don’t depend on the data (e.g. colour = "blue") go outside it, as plain arguments to the geom.
2. p_grouped, the first object you built and then extended with + in later exercises, without ever needing to rewrite it from scratch.
3. The PNG is a flattened image (pixels only), good for reports and slides, but you can’t edit it afterward. The RDS stores the actual ggplot object, so readRDS() gives you back something you can still change (themes, labels, layers) and re-plot.

Sources: ggplot2 reference, ggplot2 book, and ggsave() documentation.