Using R

Tutorials and examples

Conditional statements

Conditional statements allow you to run code segments based on whether certain conditions are true or false. They allow you to choose what code you want (or need) to run at a given point, and are part of what R calls ‘basic control-flow constructs’. Conditional statements involve the evaluation of logical values, so it is useful to also review operations with logical values.

Working with logical values
Control-flow constructs

🡻

1. Working with logical values

To make use of conditional statements, we first need to indicate when conditions are met or not. We can do this using the logical values TRUE or FALSE. R provides a standard suite of tools to generate logical values and conduct boolean operations, both on univariate values and over vectors (and other data types). The following sections review the tools provided in base R.

Logical operators
Boolean operations
Functions any() and all()
The %in% operator

Logical operators

The ”==” operator tests whether the value on the left equals the value on the right:

X <- 1; Y <- 1; Z <- 2
X == Y
#> [1] TRUE
X == Z
#> [1] FALSE

# Each element in vectors of matching 
# length are compared to each other
v1 <- 1:3; v2 <- c( 1, 3, 5 )
v1 == v2
#> [1] TRUE FALSE FALSE

The ”!=” operator tests whether the value on the left does not equal the value on the right:

X <- 1; Y <- 1; Z <- 2
X != Z
#> [1] TRUE
X != Y
#> [1] FALSE

# Each element in vectors of matching 
# length are compared to each other
v1 <- 1:3; v2 <- c( 1, 3, 5 )
v1 != v2
#> [1] FALSE TRUE TRUE

The ”>” operator tests whether whether the value on the left is greater than the value on the right:

X <- 1; Y <- 1; Z <- 2
Z > X
#> [1] TRUE
X > Y
#> [1] FALSE

# Each element in vectors of matching 
# length are compared to each other
v1 <- 1:3; v2 <- c( 1, 3, 5 )
v2 > v1
#> [1] FALSE TRUE TRUE

The ”>=” operator tests whether whether the value on the left is greater than or equal to the value on the right:

X <- 1; Y <- 1; Z <- 2
Z >= X
#> [1] TRUE
X >= Y
#> [1] TRUE

# Each element in vectors of matching 
# length are compared to each other
v1 <- 1:3; v2 <- c( 1, 3, 5 )
v2 >= v1
#> [1] TRUE TRUE TRUE

The ”<” operator tests whether whether the value on the left is less than the value on the right:

X <- 1; Y <- 1; Z <- 2
X < Z
#> [1] TRUE
X < Y
#> [1] FALSE

# Each element in vectors of matching 
# length are compared to each other
v1 <- 1:3; v2 <- c( 1, 3, 5 )
v1 < v2
#> [1] FALSE TRUE TRUE

The ”<=” operator tests whether whether the value on the left is less than or equal to the value on the right:

X <- 1; Y <- 1; Z <- 2
X < Z
#> [1] TRUE
X < Y
#> [1] FALSE

# Each element in vectors of matching 
# length are compared to each other
v1 <- 1:3; v2 <- c( 1, 3, 5 )
v1 <= v2
#> [1] TRUE TRUE TRUE

🡹 🡻

Boolean operations

The “&” operator identifies all cases that are true for x AND y:

# Define logical vectors
x <- c( T, T, F )
y <- c( F, T, F )
x & y
#> [1] FALSE TRUE FALSE

The ”|” operator identifies all cases that are true for x OR y:

# Define logical vectors
x <- c( T, T, F )
y <- c( F, T, F )
x | y
#> [1] TRUE TRUE FALSE

🡹 🡻

Functions any() and all()

The functions any and all allow a user to collapse a logical vector down to a single TRUE or FALSE value.

The function any(x) returns TRUE if any element of the logical vector x is TRUE:

x <- c( T, F, F )
any( x )
#> [1] TRUE

x <- c( F, F, F )
any( x )
#> [1] FALSE

The function all(x) returns TRUE only when all elements of the logical vector x are TRUE:

x <- c( T, T, T )
all( x )
#> [1] TRUE

x <- c( T, F, F )
all( x )
#> [1] FALSE

🡹 🡻

The operator %in%

The operator %in% allow a user to compare vectors of non-equal length and identify elements present in both vectors.

# Define vectors of unequal length
v1 <- c( 'A', 'B', 'C' )
v2 <- c( 'A', 'A', 'D', 'E', 'C' )

# Identify elements in v1 that are also in v2, 
# even though length(v1) != length(v2)
v1 %in% v2
#> [1] TRUE FALSE TRUE

# Identify elements in v2 that are also in v1
v2 %in% v1
#> [1] TRUE TRUE FALSE FALSE TRUE

🡹 🡻

2. Control-flow constructs

Conditional statements are one of two primary types of control-flow constructs. A conditional statement checks if a condition is met, and if it evaluates to TRUE, runs a code segment. There are four methods of implementing a conditional statement:

if …
if … else …
ifelse()
switch()

if …

The general template for the simplest conditional statement ‘if’ is:

if (<condition is TRUE>) {
  <Do something>
}

Specific examples:

# If condition is true, outputs text to console
cond <- TRUE
if (cond) {
  print( 'Hello' )
  print( 'World' )
}
#> [1] "Hello"
#> [1] "World"

# For a single call, brackets can be excluded
if (cond)
  print( 'Hello world' )
#> [1] "Hello world"

# If evaluates to FALSE, no output
if (!cond)
  print( 'Hello world' )

# A logical vector of length greater 
# than 1 will return an error
vec <- c( TRUE, FALSE, FALSE ) 
if (vec) # Returns error
  print( 'Hello world' )
#> Error in if (vec) print("Hello world") : 
#>   the condition has length > 1

# Use 'any' or 'all' for vectors
if ( any( vec ) ) # TRUE
  print( 'Hello world' )
#> [1] "Hello world"
if ( all( vec ) ) # FALSE
  print( 'Hello world' )

🡹 🡻

if … else …

The general template for extending an if statement with an additional else statement is:

if (<condition is TRUE>) {
  <Do something>
} else {
  <Do something else>
}

Specific examples:

# If condition is TRUE, outputs first statement
cond <- TRUE
if (cond) {
  print('1st')
  print('  statement')
} else {
  print('2nd')
  print('  statement')
}
#> [1] "1st"
#> [1] "  statement"

# If condition is FALSE, outputs second statement
if (!cond) {
  print('1st')
  print('  statement')
} else {
  print('2nd')
  print('  statement')
}
#> [1] "2nd"
#> [1] "  statement"

# Use 'any' or 'all' for vectors
vec <- c( TRUE, FALSE, FALSE )
if ( any( vec ) ) # TRUE
  print( 'A' ) else print( 'B' )
#> [1] "A"
if ( all( vec ) ) # FALSE
  print( 'A' ) else print( 'B' )
#> [1] "B"

🡹 🡻

ifelse()

The ifelse function adapts the if and else combination to vectors. The general template for the function call is:

ifelse( <logical vector>, <output if TRUE>, <output if FALSE> )

Specific examples:

# Define logical vector
vec <- c( TRUE, FALSE, FALSE )
# Return a character vector with elements equal to:
# 'A' when 'vec' is TRUE and
# 'B' when 'vec' is FALSE
ifelse( vec, 'A', 'B' )
#> [1] "A" "B" "B"

# Type of output depends
# on data type of options 
# and what is returned

# If options are same data 
# type then returns vector 
# of same data type
ifelse( vec, 1, 2 ) # Numeric vector
#> [1] 1 2 2

# If data types are different 
# for options then returns most 
# flexible data type
ifelse( vec, 1, 'A' ) # Character vector
#> [1] "1" "A" "A"

# However, if only one option 
# is returned, then depends 
# what is selected
ifelse( c( T, T, T ), 1, 'A' ) # Numeric vector
#> [1] 1 1 1

# If one option is a vector of 
# multiple elements, only first 
# element is returned
ifelse( vec, 1:3, 4 ) # Only 1 is returned
#> [1] 1 4 4

# Multiple elements can be returned 
# by using lists
ifelse( vec, list( 1:3 ), 4 ) # 1:3 is returned
#> [[1]]
#> [1] 1 2 3
#> 
#> [[2]]
#> [1] 4
#> 
#> [[3]]
#> [1] 4

Note: The ifelse function strips attributes of an object, and this can result in unexpected output - for example, using ifelse with dates will convert the output to integers, which is unlikely to be useful for the standard user.

🡹 🡻

switch()

The function switch compares an input value against an internal set of labels - if a match is found, switch returns the output the user has associated with the internal label. Labels can be either a) numbers starting from 1 onwards, or b) character strings. The general template is:

# Using numbers to index positions
switch(
  <number from 1 onwards>,
  <output when number == 1>,
  <output when number == 2>,
  <etc.>
)

# Using character strings as labels
switch(
  <Character string that matches labels exactly>,
  <Label 1> = <output>,
  <Label 2> = <other output>,
  <etc.>
)

Specific examples:

input <- 1 # When input is an integer
switch(
  input,
  'A', # Picked if input == 1
  'B', # Picked if input == 2
  'C'  # Picked if input == 3
)
#> "A"

# A NULL value is returned if integer 
# is larger than number of outputs
input <- 4
x <- switch(
  input,
  'A', 
  'B', 
  'C'  
)
print( x )
#> NULL

input <- 'Label_2' # When input is a character string
switch(
  input,
  Label_1 = 'A',
  Label_2 = 'B',
  Label_3 = 'C'
)
#> "B"

# Again, a NULL value is returned if 
# input does not match any labels
input <- 'Label_4'
x <- switch(
  input,
  Label_1 = 'A',
  Label_2 = 'B',
  Label_3 = 'C'
)
print( x )
#> NULL

# One can override the NULL output 
# by including an unlabeled option
input <- 'Label_4'
x <- switch(
  input,
  Label_1 = 'A',
  Label_2 = 'B',
  Label_3 = 'C',
  'No match'
)
print( x )
#> [1] 'No match'

# Another special aspect of the 'switch' 
# function is the ability to leave the 
# output blank, in which case the function 
# skips to the next option
input <- 'B'
switch(
  input,
  'A' = 1, 
  'B' = , # Skips to 'C'
  'C' = 3,
  'No match'
)
#> [1] 3

🡹

Return to: Foundations; Sections; Home page

Conditional statements

Table of contents

1. Working with logical values

Logical operators

Boolean operations

Functions any() and all()

The operator %in%

2. Control-flow constructs

if …

if … else …

ifelse()

switch()