Exercise 03: R Functions – Jihong Zhang, Ph.D.

1 Creating Custom Functions

Write a function called fahrenheit_to_celsius that converts a temperature from Fahrenheit to Celsius.

Formula: $C = \frac{(F - 32) \times 5}{9}$
Test your function with $F = 98.6$ .

⌘+C

fahrenheit_to_celsius <- function(temp_f) {
  return((temp_f - 32) * 5 / 9)
}
fahrenheit_to_celsius(temp_f = 98.6)

Create a function called circle_area that calculates the area of a circle given its radius.

Formula: $A = \pi \times r^2$
Use the built-in constant pi in R and test the function with $r = 5$ .

⌘+C

# r is the circle radius
circle_area <- function(r){
  return(pi * r^2) 
}
circle_area(r = 5)
circle_area(5) # 5 assign to r

2 Piping

Transform the following code into a pipe-based version:

⌘+C

result <- filter(mtcars, hp > 100)
result <- arrange(result, desc(mpg))
result <- select(result, hp, mpg)
result

⌘+C

mtcars |> 
  filter(hp > 100) |> 
  arrange(desc(mpg)) |> 
  select(hp, mpg)

3 Debugging Functions

Identify and correct the mistake in the following code:

⌘+C

df |> mutate(
  a = (a - mean(a, na.rm = TRUE)) / sd(a),
  b = (b - mean(a, na.rm = TRUE)) / sd(b) # a should be changed to b
)

Rewrite the code by creating a custom function standardized() to avoid code repetition.

⌘+C

standardized <- function(x){
  return((x - mean(x, na.rm = TRUE)) / sd(x))
}
df |> mutate(
  a = standardized(a),
  b = standardized(b)
)

4 Arguments in Functions

Create a function greet that takes two arguments, name (default = “World”) and punctuation (default = “!”). The function should return a greeting message.

Example: greet("Alice", "?") should return "Hello, Alice?".

⌘+C

greet <- function(name = "World", punctuation = "!") {
  paste0("Hello, ", name, punctuation)
}
greet(name = "Alice", punctuation = "?")

Test the sum() function by passing named arguments (x = 3, y = 5) and explain how the function handles them.

⌘+C

sum(x = 3, y = 5)
sum(3, 5)
sum(c(3, 5))
# mean(x = 3, 5)
# mean(c(3, 5))

5 Function Scope

Explain why the following code throws an error:

⌘+C

add <- function(x, y) {
  result <- x + y # beacuase result is in the local scope
  return(result)
}
add(5, 3)
result  # Error

Modify the code to make the variable result accessible outside the function.

⌘+C

add <- function(x, y) {
  result <<- x + y # beacuase result is in the local scope
  return(result)
}
add(6, 3)
result

6 Flexible Arguments

Write a function multiply_all() that accepts a variable number of arguments and multiplies all the numbers together. Test your function with multiply_all(2, 3, 4).

⌘+C

multiply_all <- function(x, y, z) { # flexible arguments
  x * y * z
}
multiply_all(2, 3, 4)

⌘+C

multiply_all_flexible <- function(...) {
  prod(c(...))
}
multiply_all_flexible(1, 2, 3)
multiply_all_flexible(1, 2, 3, 5)
multiply_all_flexible(1, 2, 3, 5, 8)
multiply_all_flexible(c(1, 2, 3, 5, 8))

Investigate the mean() function by using the help page. Explain why it can accept additional arguments like y in the code below:

⌘+C

mean(x = c(1, 2, 3), y = 3) # because y is flexible argument. It is not included into calculation of mean

7 Nested Functions

Write a function rect_prism_volume() that calculates the volume of a rectangular prism given its length, width, and height. Inside this function, create another function rect_area() to calculate the base area.
- The formula for the volume of a rectangular prism is:
  
  $V = A_b \times h$
  
  $A_b = l \times w$
  
  where:
  - V = Volume of the rectangular prism,
  - $A_b$ = base area
    - l = Length,
    - w = Width,
  - h = Height.

⌘+C

# the first function
rect_area <- function(l, w){
  A_base = l * w
  return(A_base)
}

# the second function that contains the first function
rect_prism_volume <- function(l, w, h){
  return(rect_area(l, w) * h)
}

Test your function with length = 5, width = 3, height = 4.

⌘+C

rect_prism_volume(l =5, w = 3, h = 4)

8 Rescale Function

Create a custom function rescale() that rescales a numeric vector to the range 0 to 1.

Formula: $x_{\text{scaled}} = \frac{x - \min(x)}{\max(x) - \min(x)}$

⌘+C

rescale <- function(x){
  res = (x - min(x)) / (max(x) - min(x))
  return(res)
}

Use your rescale() function to rescale the columns of the following data frame:

⌘+C

set.seed(1234)
df <- tibble(
 a = rnorm(5),
 b = rnorm(5),
 c = rnorm(5)
)
df

⌘+C

df |> 
  mutate(
    a = rescale(a),
    b = rescale(b),
    c = rescale(c)
  )

--- title: "Exercise 03: R Functions" subtitle: "" author: "Jihong Zhang*, Ph.D" institute: | Educational Statistics and Research Methods (ESRM) Program* University of Arkansas date: "2025-08-18" execute: warning: false message: false eval: false echo: true format: html: page-layout: full toc: true toc-depth: 2 lightbox: true code-fold: show --- ## Creating Custom Functions 1. Write a function called `fahrenheit_to_celsius` that converts a temperature from Fahrenheit to Celsius. - Formula: $C = \frac{(F - 32) \times 5}{9}$ - Test your function with $F = 98.6$. ```{r} fahrenheit_to_celsius <- function(temp_f) { return((temp_f - 32) * 5 / 9) } fahrenheit_to_celsius(temp_f = 98.6) ``` 2. Create a function called `circle_area` that calculates the area of a circle given its radius. - Formula: $A = \pi \times r^2$ - Use the built-in constant `pi` in R and test the function with $r = 5$. ```{r} # r is the circle radius circle_area <- function(r){ return(pi * r^2) } circle_area(r = 5) circle_area(5) # 5 assign to r ``` ------------------------------------------------------------------------ ## Piping 1. Transform the following code into a pipe-based version: ```{r} result <- filter(mtcars, hp > 100) result <- arrange(result, desc(mpg)) result <- select(result, hp, mpg) result ``` ```{r} mtcars |> filter(hp > 100) |> arrange(desc(mpg)) |> select(hp, mpg) ``` ------------------------------------------------------------------------ ## Debugging Functions 1. Identify and correct the mistake in the following code: ```{r} df |> mutate( a = (a - mean(a, na.rm = TRUE)) / sd(a), b = (b - mean(a, na.rm = TRUE)) / sd(b) # a should be changed to b ) ``` 2. Rewrite the code by creating a custom function `standardized()` to avoid code repetition. ```{r} standardized <- function(x){ return((x - mean(x, na.rm = TRUE)) / sd(x)) } df |> mutate( a = standardized(a), b = standardized(b) ) ``` ------------------------------------------------------------------------ ## Arguments in Functions 1. Create a function `greet` that takes two arguments, `name` (default = "World") and `punctuation` (default = "!"). The function should return a greeting message. - Example: `greet("Alice", "?")` should return `"Hello, Alice?"`. ```{r} greet <- function(name = "World", punctuation = "!") { paste0("Hello, ", name, punctuation) } greet(name = "Alice", punctuation = "?") ``` 2. Test the `sum()` function by passing named arguments (`x = 3, y = 5`) and explain how the function handles them. ```{r} sum(x = 3, y = 5) sum(3, 5) sum(c(3, 5)) # mean(x = 3, 5) # mean(c(3, 5)) ``` ------------------------------------------------------------------------ ## Function Scope 1. Explain why the following code throws an error: ```{r} add <- function(x, y) { result <- x + y # beacuase result is in the local scope return(result) } add(5, 3) result # Error ``` 2. Modify the code to make the variable `result` accessible outside the function. ```{r} add <- function(x, y) { result <<- x + y # beacuase result is in the local scope return(result) } add(6, 3) result ``` ------------------------------------------------------------------------ ## Flexible Arguments 1. Write a function `multiply_all()` that accepts a variable number of arguments and multiplies all the numbers together. Test your function with `multiply_all(2, 3, 4)`. ```{r} multiply_all <- function(x, y, z) { # flexible arguments x * y * z } multiply_all(2, 3, 4) ``` ```{r} multiply_all_flexible <- function(...) { prod(c(...)) } multiply_all_flexible(1, 2, 3) multiply_all_flexible(1, 2, 3, 5) multiply_all_flexible(1, 2, 3, 5, 8) multiply_all_flexible(c(1, 2, 3, 5, 8)) ``` 1. Investigate the `mean()` function by using the help page. Explain why it can accept additional arguments like `y` in the code below: ```{r} mean(x = c(1, 2, 3), y = 3) # because y is flexible argument. It is not included into calculation of mean ``` ------------------------------------------------------------------------ ## Nested Functions 1. Write a function `rect_prism_volume()` that calculates the volume of a rectangular prism given its length, width, and height. Inside this function, create another function `rect_area()` to calculate the base area. - The **formula for the volume of a rectangular prism** is: $V = A_b \times h$ $A_b = l \times w$ where: - V = Volume of the rectangular prism, - $A_b$ = base area - l = Length, - w = Width, - h = Height. ```{r} # the first function rect_area <- function(l, w){ A_base = l * w return(A_base) } # the second function that contains the first function rect_prism_volume <- function(l, w, h){ return(rect_area(l, w) * h) } ``` 1. Test your function with length = 5, width = 3, height = 4. ```{r} rect_prism_volume(l =5, w = 3, h = 4) ``` ------------------------------------------------------------------------ ## Rescale Function 1. Create a custom function `rescale()` that rescales a numeric vector to the range 0 to 1. - Formula: $x_{\text{scaled}} = \frac{x - \min(x)}{\max(x) - \min(x)}$ ```{r} rescale <- function(x){ res = (x - min(x)) / (max(x) - min(x)) return(res) } ``` 2. Use your `rescale()` function to rescale the columns of the following data frame: ```{r} set.seed(1234) df <- tibble( a = rnorm(5), b = rnorm(5), c = rnorm(5) ) df ``` ```{r} df |> mutate( a = rescale(a), b = rescale(b), c = rescale(c) ) ```