Data types and classes

Lecture 9

Dr. Mine Çetinkaya-Rundel

Duke University
STA 199 - Fall 2026

September 23, 2026

Warm-up

While you wait: Participate 📱💻

Fill in the blank:

On Wednesdays, I have _____ class(es), including STA 199.

Go to wooclap.com and use the code TJEJAXK.

Participate 📱💻

Which of the following best describes you?

First-year
Sophomore
Junior
Senior

Go to wooclap.com and use the code TJEJAXK.

Announcements

Team preferences:
- Exactly 4 students per team
- Let me know in my office hours today or at the end of lab to your TA tomorrow
- All 4 members must be in attendance when submitting the preference
No AI use in lab – work with your teammates and ask your TAs for help

Recap: The tidyverse package

When you load the tidyverse package, you get access to a suite of packages that work well together for data manipulation and visualization:
```
library(tidyverse)
```
You never need to load one of these packages individually after you load the tidyverse, e.g., library(dplyr).

❌ What not to do:

```{r}
#| label: load-packages
#| message: false
library(tidyverse)
library(dplyr)
library(tidyr)
```

✅ What to do:

```{r}
#| label: load-packages
#| message: false
library(tidyverse)
```

Recap: Loading packages

Only need to load a package once per R session or Quarto document.
Good practice: Load all packages you need at the start of your document, that’s why the templates I give you usually have a load-packages code cell at the top.
Don’t load these packages again in the same document.
If you discover that you need a new package, go back and add it to the load-packages cell.

Recap: Loading packages

❌ What not to do:

```{r}
#| label: long-running-near-peak
library(dplyr)
spotify |>
    filter(weeks_on_chart > 275, positions_below_peak < 30) |>
    # and more code...
```

some content...

```{r}
#| label: most-streamed-song
library(dplyr)
spotify |>
    group_by(genre) |>
    filter(n() >= 10) |>
    # and more code...
```

✅ What to do:

```{r}
#| label: long-running-near-peak
library(tidyverse)
spotify |>
    filter(weeks_on_chart > 275, positions_below_peak < 30) |>
    # and more code...
```

some content...

```{r}
#| label: most-streamed-song
spotify |>
    group_by(genre) |>
    filter(n() >= 10) |>
    # and more code...
```

Recap: Pipes

This is not a pipe.

Recap: Pipes

This is not our pipe [operator].

Recap: Pipes

This is our a pipe [operator].

Recap: Pipes

❌ What not to do:

```{r}
#| label: family-sustaining-wage
nc_county %>%
    filter(p_family_sustaining_wage < 0.5) %>%
    count(county_type, name = "n_counties") %>%
    arrange(desc(n_counties))
```

some content...

```{r}
#| label: edu-he-create
nc_county <- nc_county |>
    mutate(
        p_edu_he = p_edu_assoc + p_edu_ba + p_edu_mapl
    )
```

✅ What to do:

```{r}
#| label: family-sustaining-wage
nc_county |>
    filter(p_family_sustaining_wage < 0.5) |>
    count(county_type, name = "n_counties") |>
    arrange(desc(n_counties))
```

some content...

```{r}
#| label: edu-he-create
nc_county <- nc_county |>
    mutate(
        p_edu_he = p_edu_assoc + p_edu_ba + p_edu_mapl
    )
```

Aside: The Itsy Bitsy Spider

🕷️ Any volunteers to sing (or just recite) a bit of the Itsy Bitsy Spider for us?

Recap: Data pipelines

Now that we all remember how The Itsy Bitsy Spider goes…

Piped: Read from top to bottom.

spider |>
  climb_spout() |>
  wash_down_in_rain() |>
  dry_in_sun() |>
  climb_again()

Nested: Read from the inside out.

climb_again(dry_in_sun(wash_down_in_rain(climb_spout(spider))))

The pipe |> passes the result of each step to the next function as its first argument.

Recap: Data pipelines

❌ What not to do:

```{r}
#| label: q2-a
distinct(df_2, member, shift)
```

✅ What to do:

```{r}
#| label: q2-a
df_2 |>
  distinct(member, shift)
```

Data types

Data types in R

character: "First-year" – Character strings
double: 2.5 – Floating point numerical values (default numerical type)
integer: 3L – Integer numerical values (indicated with an L)
logical: TRUE or FALSE – Boolean values
and some more, but we won’t be focusing on those

Vectors

Vectors can be constructed using the c() function.

Numeric vector:

c(1, 2, 3)

[1] 1 2 3

Character vector:

c("Hello", "World!")

[1] "Hello"  "World!"

Vector made of vectors:

c(c("hi", "hello"), c("bye", "jello"))

[1] "hi"    "hello" "bye"   "jello"

Vectors and their types

Why do we care about vectors and their types, especially when all we’ve worked with so far are data frames and their columns…

Each column of a data frame is (generally) a vector, and the type of that vector determines what operations can be performed on it or how it will behave when certain functions are applied to it. For example, you can’t take the mean of a character vector, but you can take the mean of a numeric vector.

Making vectors from vectors

We have a vector, x, of type character:
x <- c("a", "b", "c")

And another vector, y, of type double:
y <- c(1, 2, 3)

What happens if we combine them into a new vector, z <- c(x, y)? What type will z be?

character
double
integer
logical

Go to wooclap.com and use the code TJEJAXK.

Making vectors from vectors

x <- c("a", "b", "c")
typeof(x)

[1] "character"

y <- c(1, 2, 3)
typeof(y)

[1] "double"

z <- c(x, y)
z

[1] "a" "b" "c" "1" "2" "3"

typeof(z)

[1] "character"

Converting between types

with intention…

x <- 1:3
x

[1] 1 2 3

typeof(x)

[1] "integer"

y <- as.character(x)
y

[1] "1" "2" "3"

typeof(y)

[1] "character"

Converting between types

with intention…

x <- c(TRUE, FALSE)
x

[1]  TRUE FALSE

typeof(x)

[1] "logical"

y <- as.numeric(x)
y

[1] 1 0

typeof(y)

[1] "double"

Converting between types

without intention…

c(2, "Just this one!")

[1] "2"              "Just this one!"

R will happily convert between various types without complaint when different types of data are concatenated in a vector, and that’s not always a great thing!

Converting between types

without intention…

c(FALSE, 3L)

[1] 0 3

c(FALSE, 1.2)

[1] 0.0 1.2

c(2L, "two")

[1] "2"   "two"

c(TRUE, "two")

[1] "TRUE" "two"

Participate 📱💻

What is the output of typeof(c(1.2, 3L))?

"character"
"double"
"integer"
"logical"

Go to wooclap.com and use the code TJEJAXK.

Explicit vs. implicit coercion

Explicit coercion:

When you call a function like as.logical(), as.numeric(), as.integer(), as.double(), or as.character().

Implicit coercion:

Happens when you use a vector in a specific context that expects a certain type of vector.

Data classes

Vectors are like Lego building blocks
We stick them together to build more complicated constructs, e.g. representations of data
The class attribute relates to the S3 class of an object which determines its behaviour
- You don’t need to worry about what S3 classes really mean, but you can read more about it here if you’re curious
Examples: factors, dates, and data frames

Factors

R uses factors to handle categorical variables, variables that have a fixed and known set of possible values

class_years <- factor(
  c(
    "First-year",
    "Sophomore",
    "Sophomore",
    "Senior",
    "Junior"
  )
)
class_years

[1] First-year Sophomore  Sophomore  Senior     Junior    
Levels: First-year Junior Senior Sophomore

typeof(class_years)

[1] "integer"

class(class_years)

[1] "factor"

More on factors

We can think of factors like character (level labels) and an integer (level numbers) glued together

glimpse(class_years)

 Factor w/ 4 levels "First-year","Junior",..: 1 4 4 3 2

as.integer(class_years)

[1] 1 4 4 3 2

Dates

today <- as.Date("2026-09-23")
today

[1] "2026-09-23"

typeof(today)

[1] "double"

class(today)

[1] "Date"

More on dates

We can think of dates like an integer (the number of days since the origin, 1 Jan 1970) and an integer (the origin) glued together

as.integer(today)

[1] 20719

as.integer(today) / 365 # roughly 56.8 yrs

[1] 56.76438

Data frames

We can think of data frames like like vectors of equal length glued together

df <- data.frame(x = 1:2, y = 3:4)
df

  x y
1 1 3
2 2 4

typeof(df)

[1] "list"

class(df)

[1] "data.frame"

Lists

Lists are a generic vector container; vectors of any type can go in them

l <- list(
  x = 1:4,
  y = c("hi", "hello", "jello"),
  z = c(TRUE, FALSE)
)
l

$x
[1] 1 2 3 4

$y
[1] "hi"    "hello" "jello"

$z
[1]  TRUE FALSE

Lists and data frames

A data frame is a special list containing vectors of equal length

df

  x y
1 1 3
2 2 4

When we use the pull() function, we extract a vector from the data frame

df |>
  pull(y)

[1] 3 4

Working with factors

Read data in as character strings

fake_survey

# A tibble: 100 × 2
   class_year n_wed_classes
   <chr>              <dbl>
 1 Sophomore              2
 2 Senior                 4
 3 Sophomore              2
 4 Sophomore              1
 5 Sophomore              1
 6 Junior                 4
 7 Senior                 2
 8 Senior                 4
 9 Sophomore              3
10 First-year             4
# ℹ 90 more rows

But coerce when plotting

ggplot(fake_survey, mapping = aes(x = class_year)) +
  geom_bar()

Use forcats to reorder levels

fake_survey |>
  mutate(
    class_year = fct_relevel(
      class_year,
      "First-year",
      "Sophomore",
      "Junior",
      "Senior"
    )
  ) |>
  ggplot(mapping = aes(x = class_year)) +
  geom_bar()

A peek into forcats

Reordering levels by:

fct_relevel(): hand
fct_infreq(): frequency
fct_reorder(): sorting along another variable
fct_rev(): reversing

…

Changing level values by:

fct_lump(): lumping uncommon levels together into “other”
fct_other(): manually replacing some levels with “other”

…

Application exercise

`ae-05-survey-types`

Go to your ae project in Positron.
If you haven’t yet done so, make sure all of your changes up to this point are committed and pushed, i.e., there’s nothing left in your source control pane.
If you haven’t yet done so, pull to get today’s application exercise file: ae-05-survey-types.qmd.
Work through the application exercise in class, and render, commit, and push your edits by the end of class.