Working with vectors
The atomic vector class in R allows for storing multiple values in a single variable. Users then have a variety of options for what they can do with this vector, from extracting a subset of values from the vector, performing mathematical operations on the vector, or summarizing the contents of the vector in different ways. I provide a variety of examples of operations a user can perform on a simple atomic vector in R in this section.
Note that indexing in R is very flexible, but sometimes unintuitive, and can produce unintended results without any error messages or warnings.
Table of contents
1. Extracting a subset of values
R provides a special operator function [ (i.e., square brackets) to extract specific elements from a vector. This is also known as ‘indexing’ a vector:
# Define example vector
vec <- c( 1, 2, 5, 100 )
# Extract 1st element in the vector
vec[1]
#> [1] 1
# Extract 2nd element in the vector
vec[2]
#> [1] 2
The [ function can be combined with the c function and : operator to extract a range of values:
# Define example vector
vec <- c( 1, 2, 5, 10, 100 )
# Extract first 3 elements
vec[1:3]
#> [1] 1 2 5
# Extract 4th and 5th elements
vec[4:5]
#> [1] 10 100
# Extract 1st, 3rd, and 5th elements
vec[ c( 1, 3, 5 ) ]
#> [1] 1 5 100
# Define vector for indexing purposes
index <- c( 1, 3, 5 )
vec[ index ]
#> [1] 1 5 100
In a more complicated example, one can label each element in a vector, and then use the labels to extract specific elements:
# Create a vector with named/labeled elements
number_of_pets <- c(
'Dogs' = 2, 'Ferrets' = 1, 'Fish' = 2
)
number_of_pets['Dogs']
#> Dogs
#> 2
number_of_pets[ c( 'Dogs', 'Fish' ) ]
#> Dogs Fish
#> 2 2
2. Conditional indexing
Users can use conditional logic to isolate a subset of a vector that match desired conditions:
# Range of values
vec <- -2:2
# Only values above 0
vec[ vec > 0 ]
#> [1] 1 2
# Only values less than or equal to 0
vec[ vec <= 0 ]
#> [1] -2 -1 0
# Values greater than 1 OR less than 1
vec[ ( vec > 1 ) | ( vec < -1 ) ]
#> [1] -2 2
# For complex conditional statements,
# it can help to define a logical
# vector for indexing purposes
sel <-
vec > 1 |
vec < -1
vec[sel]
#> [1] -2 2
3. Incorrect indexing
Incorrectly indexing a vector will not necessarily result in an error message:
vec <- c( 'Dogs' = 2, 'Cats' = 2, 'Fish' = 0 )
# Does not produce an error, returns a
# missing value symbol instead
vec[4]
#> <NA>
#> NA
vec['Ducks']
#> <NA>
#> NA
# Similar result for non-labeled vector
vec <- 1:3
vec[4]
#> [1] NA
vec['Ducks']
#> [1] NA
# Passing a missing value as an index
# results in strange behavior
vec[NA]
#> [1] NA NA NA
Return to: Foundations; Sections; Home page