5 Control flow

5.1 Introduction

There are two primary tools of control flow: choices and loops. Choices, like if statements and switch() calls, allow you to run different code depending on the input. Loops, like for or while, allow you to repeatedly run code, typically with changing options. I expect that you’re already familiar with the basics of these functions so I’ll briefly cover some technical details and introduce you to some useful, but lesser known, features.

The condition system (messages, warnings, and errors) also provides “non-local” control flow. You’ll learn about them in Chapter 8.


Want to skip this chapter? Go for it, if you can answer the questions below. Find the answers at the end of the chapter in Section 5.4.

  • What is the different between if and ifelse()?

  • In the following code, what will the value of y be if x is TRUE? What if x is FALSE? What if x is NA?

    y <- if (x) 3
  • What does switch("x", x = , y = 2, z = 3) return?


  • Section 5.2 dives into the details of if, then discusses the close relatives ifelse() and switch().

  • Section 5.3 starts off by reminding you of the basic structure of the for loop in R, discusses some common pitfalls, and then talks about the related while and repeat statements.

5.2 Choices

The basic form of R’s if statement is as follows:

if (condition) true_action
if (condition) true_action else false_action

If condition is TRUE, true_action will be evaluated; if condition is FALSE, the optional else statement will be evaluated.

Typically the actions are compound statements contained within {:

grade <- function(x) {
  if (x > 90) {
  } else if (x > 80) {
  } else if (x > 50) {
  } else {

if returns a value so that you can assign the results:

x1 <- if (TRUE) 1 else 2
x2 <- if (FALSE) 1 else 2

c(x1, x2)
#> [1] 1 2

(I recommend using assigning the results of an entire if statement only when the whole thing fits one line; otherwise it tends to be hard to read.)

When you use the single argument form without an else statement, if invisibly (Section 6.7.2) returns NULL if the condtion is FALSE. Since functions like c() and paste() drop NULL inputs, this allows for a compact expression of certain idioms:

greet <- function(name, birthday = FALSE) {
    "Hi ", name,
    if (birthday) " and HAPPY BIRTHDAY"
greet("Maria", FALSE)
#> [1] "Hi Maria"
greet("Jaime", TRUE)
#> [1] "Hi Jaime and HAPPY BIRTHDAY"

5.2.1 Invalid inputs

The condition should evaluate to a single TRUE or FALSE. Most other inputs generate an error:

if ("x") 1
#> Error in if ("x") 1:
#>   argument is not interpretable as logical
if (logical()) 1
#> Error in if (logical()) 1:
#>   argument is of length zero
if (NA) 1
#> Error in if (NA) 1:
#>   missing value where TRUE/FALSE needed

The exception is a logical vector of length greater than 1, which only generates a warning:

if (c(TRUE, FALSE)) 1
#> Warning in if (c(TRUE, FALSE)) 1: the condition has length > 1 and only the
#> first element will be used
#> [1] 1

In R 3.5.0 and greater, thanks to Henrik Bengtsson, you can turn this an error by setting an environment variable:

Sys.setenv("_R_CHECK_LENGTH_1_CONDITION_" = "true")
if (c(TRUE, FALSE)) 1
#> Error in if (c(TRUE, FALSE)) 1:
#>   the condition has length > 1

I think it’s good practice to set this environment variable, as it reveals a clear mistake that you might miss if only shown as a warning.

5.2.2 Vectorised if

Given that if only works with a single TRUE or FALSE, you might wonder what to do if you have a vector of logical values. Handling vectors of values is the job of ifelse(): a vectorised function with test, yes, and no vectors (that will be recycled to the same length):

x <- 1:10
ifelse(x %% 5 == 0, "XXX", as.character(x))
#>  [1] "1"   "2"   "3"   "4"   "XXX" "6"   "7"   "8"   "9"   "XXX"

ifelse(x %% 2 == 0, "even", "odd")
#>  [1] "odd"  "even" "odd"  "even" "odd"  "even" "odd"  "even" "odd"  "even"

Missing values will be propagated into the output.

I recommend only using ifelse() when the yes and no vectors are the same type, as it is otherwise hard to predict the output type. See about https://vctrs.r-lib.org/articles/stability.html#ifelse for additional discussion.

Another vectorised equivalent is the more general dplyr::case_when(). It uses a special syntax to allow any number of condition-vector pairs:

  x %% 35 == 0 ~ "fizz buzz",
  x %% 5 == 0 ~ "fizz",
  x %% 7 == 0 ~ "buzz",
  is.na(x) ~ "???",
  TRUE ~ as.character(x)
#>  [1] "1"    "2"    "3"    "4"    "fizz" "6"    "buzz" "8"    "9"    "fizz"

5.2.3 switch() statement

Closely related to if is the switch()-statement. It’s a compact, special purpose equivalent that lets you replace code like:

x_option <- function(x) {
  if (x == "a") {
    "option 1"
  } else if (x == "b") {
    "option 2" 
  } else if (x == "c") {
    "option 3"
  } else {
    stop("Invalid `x` value")

with the more succinct:

x_option <- function(x) {
    a = "option 1",
    b = "option 2",
    c = "option 3",
    stop("Invalid `x` value")

The last component of a switch() should always throw an error, otherwise unmatched inputs will invisibly return NULL:

(switch("c", a = 1, b = 2))

If multiple inputs have the same output, you can leave the right hand side of = empty and the input will “fall through” to the next value. This mimics the behaviour of C’s switch statement:

legs <- function(x) {
    cow = ,
    horse = ,
    dog = 4,
    human = ,
    chicken = 2,
    plant = 0,
    stop("Unknown input")
#> [1] 4
#> [1] 4

It is also possible to use switch() with numeric x, but is harder to read, and has undesirable failure modes if x is a not a whole number. I recommend using switch() only with character inputs.

5.2.4 Exercises

  1. What type of vector does each of the following calls to ifelse() return?

    ifelse(TRUE, 1, "no")
    ifelse(FALSE, 1, "no")
    ifelse(NA, 1, "no")

    Read the documentation and write down the rules in your own words.

  2. Why does the following code work?

    x <- 1:10
    if (length(x)) "not empty" else "empty"
    #> [1] "not empty"
    x <- numeric()
    if (length(x)) "not empty" else "empty"
    #> [1] "empty"

5.3 Loops

For loops are used to iterate over items in a vector. They have the following basic form:

for (item in vector) perform_action

perform_action is called once for each item in vector, rebinding the value of item each time.

for (i in 1:3) {
#> [1] 1
#> [1] 2
#> [1] 3

(When iterating over a vector of indices, it’s conventional to use very short variable names like i, j, or k.)

NB: for assigns the item in the current environment, so that an existing variable with the same name will be overwritten:

i <- 100
for (i in 1:3) {}
#> [1] 3

There are two ways to terminate a for loop early:

  • next terminates the current iteration and advances to the next.
  • break exits the entire for loop.
for (i in 1:10) {
  if (i < 3) 

  if (i >= 5)
#> [1] 3
#> [1] 4
#> [1] 5

5.3.1 Common pitfalls

There are three common problems that you might encounter when using for. First, if you’re generating data, make sure to preallocate the output. Otherwise the loop will be very slow; see Sections 23.2.2 and 24.6 for more details. The vector() function is helpful here.

means <- c(1, 50, 20)
out <- vector("list", length(means))
for (i in 1:length(means)) {
  out[[i]] <- rnorm(10, means[[i]])

Next, beware iterating over 1:length(x). This will fail in unhelpful ways if x has length 0:

means <- c()
out <- vector("list", length(means))
for (i in 1:length(means)) {
  out[[i]] <- rnorm(10, means[[i]])
#> Error in rnorm(10, means[[i]]):
#>   invalid arguments

That’s because : works with both positive and increasing sequences:

#> [1] 1 0

Instead switch to seq_along(x) which always returns a value the same length as x:

#> integer(0)

out <- vector("list", length(means))
for (i in seq_along(means)) {
  out[[i]] <- rnorm(10, means[[i]])

Finally, you might encounter problems when iterating over S3 vectors, as loops typically strip the attributes:

xs <- as.Date(c("2020-01-01", "2010-01-01"))
for (x in xs) {
#> [1] 18262
#> [1] 14610

Work around this by calling [[ yourself:

for (i in seq_along(xs)) {
#> [1] "2020-01-01"
#> [1] "2010-01-01"

5.3.2 Related tools

for loops are useful if you know in advance the set of values that you want to iterate over. There are two related tools with more flexible specifications:

  • while(condition) action: performs action while condition is TRUE.

  • repeat(action): repeats action forever (i.e. until it encounters break).

R does not have an equivalent to the do {action} while (condition) syntax found in other languages.

You can rewrite any for loop to use while instead, and you can rewrite any while loop to use repeat, but the converses are not true. That means while is more flexible than for, and repeat is more flexible than while. It’s good practice to use the least-flexible solution to a problem, which means that you should use for where possible.

Generally speaking you should not need to use for loops for data analysis tasks, as map()/apply() functions provide an even less flexible solution to most problems. You’ll learn more in Chapter 9.

5.3.3 Exercises

  1. Why does this code succeed without errors or warnings?

    x <- numeric()
    out <- vector("list", length(x))
    for (i in 1:length(x)) {
      out[i] <- x[i] ^ 2
  2. What does the following code tell you about when the vector being iterated over is evaluated?

    xs <- c(1, 2, 3)
    for (x in xs) {
      xs <- c(xs, x * 2)
    #> [1] 1 2 3 2 4 6
  3. What does the following code tell you about when the index is updated?

    for (i in 1:3) {
      i <- i * 2
    #> [1] 2
    #> [1] 4
    #> [1] 6

5.4 Answers

  • if works with scalars; ifelse() works with vectors.

  • When x is TRUE, y will be 3, when FALSE y will be NULL, when NA the if statement will throw an error.

  • This switch() statement makes use of fall-through so it will return 2. See details in Section 5.2.3.