Functions and Iteration

Lecture 34

Dr. Eric Friedlander

College of Idaho
CSCI 2025 - Winter 2026

Introduction

The Motivation

• DRY Principle: Don’t Repeat Yourself.
• Copy-pasting is dangerous:
  • It’s easy to make mistakes (copying the wrong variable).
  • It’s hard to update (you have to change code in multiple places).
• Solution:
  1. Functions: Automate common tasks.
  2. Iteration: Apply functions to multiple inputs.

Functions

When to write a function?

• Rule of Three:
  • If you’ve copied and pasted a block of code more than twice, write a function.
• Advantages:
  • Given a name that makes it easier to understand.
  • As requirements change, you only need to update code in one place.
  • Eliminates the chance of making incidental mistakes when copying and pasting.

Anatomy of a Function

my_function <- function(arg1, arg2) {
  # Body of the function
  result <- arg1 + arg2
  return(result)
}

• Name: Should be descriptive (use verbs).
• Arguments: Inputs to the function (can have defaults).
• Body: The code that executes.
• Return: The output of the function (implicit or explicit).

Types of Functions

1. Vector Functions:
   • Input: Vector.
   • Output: Vector.
   • Example: rescale01 <- function(x) { (x - min(x)) / (max(x) - min(x)) }
2. Data Frame Functions:
   • Input: Data Frame.
   • Output: Data Frame (usually).
   • Example: Wrapping a complex dplyr pipeline.
3. Plot Functions:
   • Input: Data Frame + Parameters.
   • Output: ggplot object.

Iteration

The “R” Way: Vectorization

R is designed for iteration. Many functions work on vectors automatically.

x <- 1:5
x * 2  # No loop needed!

[1]  2  4  6  8 10

Implicit Iteration should be your first choice.

Column Iteration: across()

Apply a function to multiple columns in a data frame.

library(dplyr)

df %>%
  mutate(
    across(c(x, y, z), ~ .x * 2)
  )

• Selection: Select columns by name, type (where(is.numeric)), etc.
• Function: Pass the function to apply.

List Iteration: purrr::map()

When you need to iterate over a list or a vector and implicit iteration isn’t enough.

• map(): Returns a list.
• map_lgl(), map_int(), map_dbl(), map_chr(): Return an atomic vector.

library(purrr)

path_list <- list("data/file1.csv", "data/file2.csv")
data_list <- map(path_list, read_csv)

Reading Multiple Files

A common data science task:

1. List files in a directory (list.files()).
2. Read each file (map()).
3. Combine into one data frame (list_rbind()).

files <- list.files("data/", pattern = "*.csv", full.names = TRUE)

combined_data <- files %>%
  map(read_csv) %>%
  list_rbind()

Wrap Up

Summary

• Functions allow you to reuse code logic and avoid copy-paste errors.
• Iteration allows you to apply that logic to many inputs efficiently.
• Toolkit:
  • function() for creating tools.
  • dplyr::across() for data frame columns.
  • purrr::map() for lists and general iteration.

Do Next

1. Read Chapter 25 (Functions) and Chapter 26 (Iteration) in r4ds..
2. There’s NO recitation Gem for this textbook but I recommend creating your own and adding the textbook chapter and these slides.
3. That’s it for tonight!