R Foundations

Getting started with R

R is a programming language and statistical computing environment. RStudio is an integrated development environment (IDE) for editing scripts, running R, inspecting objects, and viewing plots.

Install R and RStudio

Runnable examples on this page require no local installation. File-based analyses and saved outputs use a local installation of R and RStudio.

1. Install the current version of R from CRAN. Choose the download for your operating system and follow its installer.
2. Install RStudio Desktop, then open it. R must be installed first.
3. In RStudio, choose File → New Project → New Directory → New Project, choose a name and location, then select Create Project.
4. Choose File → New File → R Script. Type 2 + 2, leave the cursor on that line, then click Run or press Cmd+Enter on macOS / Ctrl+Enter on Windows or Linux. You should see [1] 4 in the Console.
5. Save the script as analysis.R. In the Files pane, use New Folder three times to create data, figures, and results inside the project.

Your first commands

# Anything after # is a comment.
2 + 2

minutes <- 90
hours <- minutes / 60
hours

mean(c(3, 6, 9))
round(10 / 3, digits = 2)

The assignment arrow <- stores a value in an object. R usually prints nothing when an object is created. Type the object name to see its value. The [1] in printed output marks the position of the first displayed value; it is not part of the answer.

How to read a function call

• round is the function: the action R will perform.
• The values inside parentheses are inputs, called arguments.
• digits = 2 is a named argument that changes how the function behaves.
• The returned value is the result.

Objects and data types

An object is a name that points to a value in memory. The value has a type, which tells R which operations make sense. Adding numbers makes sense; adding two student IDs does not.

student_id <- "S001"
quiz_score <- 84
completed <- TRUE
lesson_date <- as.Date("2026-09-08")

class(student_id)
class(quiz_score)
class(completed)
class(lesson_date)

typeof(quiz_score)

class() reports the class used by most analysis methods; typeof() reports the underlying storage type.

Explicit type conversion

text_number <- "42"
as.numeric(text_number)

condition <- factor(
  c("Control", "Treatment", "Control"),
  levels = c("Control", "Treatment")
)

levels(condition)

Vectors and missing values

A vector is an ordered, one-dimensional collection. Ordinary atomic vectors contain one type. Many R operations are vectorized: one instruction is applied to every element.

scores <- c(68, 84, NA, 93, 76)

length(scores)
scores[2]
scores[c(1, 4)]
scores[-3]

scores + 5
scores >= 80

R indexing starts at 1. A positive index keeps a position; a negative index removes it. The missing element stays missing because R cannot determine the result of arithmetic or comparison for an unknown value.

Zero and missing are different

0

A known numeric value equal to zero.

NA

The value is unavailable or unknown.

"0"

A character string containing the symbol zero.

""

An empty character string; it is not automatically missing.

mean(scores)
mean(scores, na.rm = TRUE)

is.na(scores)
sum(is.na(scores))
sum(scores >= 80, na.rm = TRUE)

Vector constructors

1:5
seq(from = 0, to = 1, by = 0.25)
rep(c("A", "B"), times = 3)

height_cm <- c(160, 172, 168)
height_m <- height_cm / 100
height_m
# 1.60 1.72 1.68

Datasets and importing data

In R, rectangular data are typically stored as data frames or tibbles. Rows represent observations, columns represent variables, and cells contain individual values.

Download and import the practice file

Download r_student_scores.csv, place it inside your project’s data folder, and run the following code.

# Run this once on your computer:
install.packages("tidyverse")

# Run this at the beginning of each new session:
library(tidyverse)

students <- read_csv(
  "data/r_student_scores.csv",
  show_col_types = FALSE
)

Inspect before analyzing

students
glimpse(students)
dim(students)
names(students)
head(students, 3)
summary(students)

colSums(is.na(students))
anyDuplicated(students$student_id)

glimpse() shows column types and sample values. dim() should return 12 rows and 6 columns. anyDuplicated() returns 0 when no repeated ID is found. These checks should precede transformation and analysis.

Select values by name or position

students$quiz_score
students[["quiz_score"]]
students$quiz_score[4]
students[1:3, c("student_id", "quiz_score")]

The first two forms return the same score vector. The third returns the fourth score, 93. The final command returns a small two-column table. Prefer column names over numeric column positions in an analysis script.

Transform and summarize a dataset

The native pipe |> passes the result on its left to the next function call. Pipelines express transformations in execution order.

# Keep three columns.
students |>
  select(student_id, major, quiz_score)

# Keep rows where completed is TRUE and score is at least 80.
students |>
  filter(completed == TRUE, quiz_score >= 80)

# Highest quiz score first.
students |>
  arrange(desc(quiz_score))

Inside filter(), commas mean AND. Use | for OR and %in% to match any value in a set. select() changes columns; filter() changes rows.

• The first result has 12 rows and only 3 columns.
• The filter returns S002, S004, S007, S009, and S011.
• The descending arrangement begins with S004, whose quiz score is 93.

students_prepared <- students |>
  mutate(
    study_minutes = study_hours * 60,
    major = factor(
      major,
      levels = c("Mathematics", "Biology", "Public Health")
    )
  )

major_summary <- students_prepared |>
  group_by(major) |>
  summarise(
    students = n(),
    completed = sum(completed == TRUE),
    mean_hours = mean(study_hours),
    mean_score = mean(quiz_score),
    .groups = "drop"
  )

major_summary

Assignment creates a new table named students_prepared; the original students table is unchanged. The expression completed == TRUE produces TRUE/FALSE values, and sum() counts the TRUE values when no completion values are missing.

students |>
  filter(
    major %in% c("Biology", "Public Health"),
    attendance_pct >= 80
  ) |>
  select(student_id, major, attendance_pct)

Visualize with ggplot2

A ggplot is built in layers. You provide a dataset, map variables to visual properties inside aes(), choose a geometric layer, then add labels and a theme.

ggplot(DATA, aes(x = X_VARIABLE, y = Y_VARIABLE)) +
  geom_...() +
  labs(...) +
  theme_minimal()

Uppercase tokens and ellipses are placeholders; replace them with data, variables, layers, and labels.

Scatterplot: two numeric variables

score_plot <- ggplot(
  students,
  aes(
    x = study_hours,
    y = quiz_score,
    color = major,
    shape = completed
  )
) +
  geom_point(size = 3, alpha = 0.85) +
  labs(
    title = "Study time and quiz score",
    subtitle = "Synthetic data for code practice",
    x = "Study time (hours)",
    y = "Quiz score",
    color = "Major",
    shape = "Completed"
  ) +
  theme_minimal(base_size = 12)

score_plot

Position encodes study hours and score; color and shape encode major and completion status. Association in these synthetic data does not establish causation.

Bar chart: count rows in categories

ggplot(students, aes(x = major)) +
  geom_bar(fill = "#315e7d", width = 0.7) +
  labs(
    title = "Students in each major",
    x = NULL,
    y = "Number of students"
  ) +
  theme_minimal(base_size = 12)

Each bar has height 4 because each major appears in four rows. If you already have a summarized count column, use geom_col() instead.

Histogram and boxplot

ggplot(students, aes(x = quiz_score)) +
  geom_histogram(binwidth = 5, boundary = 0) +
  labs(x = "Quiz score", y = "Number of students") +
  theme_minimal()

ggplot(students, aes(x = major, y = quiz_score)) +
  geom_boxplot(outlier.shape = NA) +
  geom_jitter(width = 0.08, height = 0, alpha = 0.8) +
  labs(x = NULL, y = "Quiz score") +
  theme_minimal()

A histogram groups numeric values into intervals; changing binwidth changes the display. The boxplot summarizes a distribution, while jittered points keep the twelve observations visible. With such a small dataset, focus on the points and avoid strong conclusions.

Export a plot

dir.create("figures", showWarnings = FALSE)

ggsave(
  filename = "figures/study-time-vs-score.png",
  plot = score_plot,
  width = 7,
  height = 4.5,
  dpi = 300
)

ggplot(
  students,
  aes(x = attendance_pct, y = quiz_score, color = completed)
) +
  geom_point(size = 3) +
  facet_wrap(~ major) +
  labs(
    x = "Attendance (%)",
    y = "Quiz score",
    color = "Completed"
  ) +
  theme_minimal()

Reproducible analysis workflow

A reproducible analysis regenerates its outputs from the same inputs and script in a fresh R session. Manual or undocumented steps prevent exact regeneration.

Recommended project structure

library(tidyverse)

# 1. Import
students <- read_csv(
  "data/r_student_scores.csv",
  show_col_types = FALSE
)

# 2. Inspect and validate
glimpse(students)
colSums(is.na(students))

stopifnot(nrow(students) > 0)
stopifnot(anyDuplicated(students$student_id) == 0)
stopifnot(!anyNA(students$quiz_score))
stopifnot(all(students$quiz_score >= 0))
stopifnot(all(students$quiz_score <= 100))

# 3. Prepare
students_clean <- students |>
  mutate(
    major = factor(
      major,
      levels = c("Mathematics", "Biology", "Public Health")
    )
  )

# 4. Summarize
major_summary <- students_clean |>
  group_by(major) |>
  summarise(
    students = n(),
    completed = sum(completed == TRUE),
    mean_score = mean(quiz_score),
    .groups = "drop"
  )

# 5. Visualize
p <- ggplot(
  students_clean,
  aes(x = study_hours, y = quiz_score, color = major)
) +
  geom_point(size = 3) +
  labs(
    title = "Study time and quiz score",
    subtitle = "Synthetic tutorial data",
    x = "Study time (hours)",
    y = "Quiz score",
    color = "Major"
  ) +
  theme_minimal(base_size = 12)

# 6. Save outputs
dir.create("results", showWarnings = FALSE)
dir.create("figures", showWarnings = FALSE)

write_csv(major_summary, "results/major-summary.csv")
ggsave(
  "figures/study-time-vs-score.png",
  plot = p,
  width = 7,
  height = 4.5,
  dpi = 300
)

# 7. Save package and R versions
capture.output(
  sessionInfo(),
  file = "results/session-info.txt"
)

Exercises and reference

Complete each exercise before opening the supplied solution, and interpret each result in words.

students |>
  filter(quiz_score > 80) |>
  arrange(desc(quiz_score)) |>
  select(student_id, major, quiz_score)

students |>
  group_by(completed) |>
  summarise(
    students = n(),
    mean_attendance = mean(attendance_pct),
    .groups = "drop"
  )

attendance_plot <- ggplot(
  students,
  aes(x = attendance_pct, y = quiz_score, shape = completed)
) +
  geom_point(size = 3, color = "#315e7d") +
  labs(
    title = "Attendance and quiz score",
    subtitle = "Synthetic tutorial data",
    x = "Attendance (%)",
    y = "Quiz score",
    shape = "Completed"
  ) +
  theme_minimal(base_size = 12)

dir.create("figures", showWarnings = FALSE)

ggsave(
  "figures/attendance-vs-score.png",
  plot = attendance_plot,
  width = 7,
  height = 4.5,
  dpi = 300
)

References and further reading

• An Introduction to R — the official R manual.
• Introduction to dplyr — official documentation for the data verbs used here.
• Introduction to ggplot2 — official explanation of the grammar of graphics.
• RStudio Projects — Posit’s guide to project-based work.
• webR — the browser-based R runtime used by the interactive examples.