Learning Objectives: After this chapter you will be able to:
- Create a list that holds elements of different types.
- Name list elements with
names().- Access elements with
[[,[, and$.- Explain why
[[returns the element while[returns a list.- Convert a vector or matrix into a list with
as.list().- Convert a list back into a vector with
unlist().
Lists are very useful as they are heterogeneous i.e. they can contain different data types. If you remember, vectors and matrices are homogeneous i.e. they can contain only one type of data. If you include different data types, they will all be coerced to the same type. With lists, it is different. In this chapter, we will delve deeper into lists.
To create a list, we use the list() function. Let us create a simple list to demonstrate how they can contain different data types.
# numeric vector
vect1 <- 1:10
# character vector
vect2 <- c('Jack', 'John', 'Jill')
# logical vector
vect3 <- c(TRUE, FALSE)
# matrix
mat <- matrix(data = 1:9, nrow = 3)
# list
l <- list(vect1, vect2, vect3, mat)
l[[1]]
[1] 1 2 3 4 5 6 7 8 9 10
[[2]]
[1] "Jack" "John" "Jill"
[[3]]
[1] TRUE FALSE
[[4]]
[,1] [,2] [,3]
[1,] 1 4 7
[2,] 2 5 8
[3,] 3 6 9If you observe the output, all the elements of the list retain their data types. Now let us learn how to access the elements of the list.
You can access the elements of a list using the following operators:
[[[$Let us try them one by one.
# using [[
l[[1]] [1] 1 2 3 4 5 6 7 8 9 10# using [
l[1][[1]]
[1] 1 2 3 4 5 6 7 8 9 10[[ extracts the element itself (which may be a vector, matrix, data frame, or nested list), while [ returns a new list containing the selected element(s). The $ operator can be used only when we have named elements in the list. Let us add names to the elements. Use the names() function to add names to the list.
# named elements
names(l) <- c('vect1', 'vect2', 'vect3', 'mat')
l$vect1
[1] 1 2 3 4 5 6 7 8 9 10
$vect2
[1] "Jack" "John" "Jill"
$vect3
[1] TRUE FALSE
$mat
[,1] [,2] [,3]
[1,] 1 4 7
[2,] 2 5 8
[3,] 3 6 9# use $
l$vect1 [1] 1 2 3 4 5 6 7 8 9 10# use [[
l[['vect1']] [1] 1 2 3 4 5 6 7 8 9 10# use [
l['vect1']$vect1
[1] 1 2 3 4 5 6 7 8 9 10You can coerce other objects to list using as.list().
# vector
vect1 <- 1:10
as.list(vect1)[[1]]
[1] 1
[[2]]
[1] 2
[[3]]
[1] 3
[[4]]
[1] 4
[[5]]
[1] 5
[[6]]
[1] 6
[[7]]
[1] 7
[[8]]
[1] 8
[[9]]
[1] 9
[[10]]
[1] 10# matrix
mat <- matrix(data = 1:9, nrow = 3)
as.list(mat)[[1]]
[1] 1
[[2]]
[1] 2
[[3]]
[1] 3
[[4]]
[1] 4
[[5]]
[1] 5
[[6]]
[1] 6
[[7]]
[1] 7
[[8]]
[1] 8
[[9]]
[1] 9Use unlist() to coerce a list to a vector.
# unlist
unlist(l) vect11 vect12 vect13 vect14 vect15 vect16 vect17 vect18 vect19 vect110
"1" "2" "3" "4" "5" "6" "7" "8" "9" "10"
vect21 vect22 vect23 vect31 vect32 mat1 mat2 mat3 mat4 mat5
"Jack" "John" "Jill" "TRUE" "FALSE" "1" "2" "3" "4" "5"
mat6 mat7 mat8 mat9
"6" "7" "8" "9" # remove names
unlist(l, use.names = FALSE) [1] "1" "2" "3" "4" "5" "6" "7" "8" "9"
[10] "10" "Jack" "John" "Jill" "TRUE" "FALSE" "1" "2" "3"
[19] "4" "5" "6" "7" "8" "9" Knowledge Check: Why use a list at all? vect1 <- 1:10, vect2 <- c('Jack', 'John', 'Jill'), and vect3 <- c(TRUE, FALSE). Why not just c(vect1, vect2, vect3)?
Answer: That would coerce everything to character, because character is the most flexible type. A list keeps each element’s own type.
c(vect1, vect2, vect3) gives you c("1", "2", ..., "10", "Jack", ...). Vectors and matrices are homogeneous, so mixing types destroys information. list(vect1, vect2, vect3) stores the three vectors side by side with their types intact.
Knowledge Check: [[ versus [ Given l <- list(1:10, c('Jack', 'John')), what is the difference between l[[1]] and l[1]?
Answer: l[[1]] is the vector 1:10 itself. l[1] is a list containing one element, which is that same vector.
[[ extracts the element, so you get the object and can operate on it directly. [ builds a new list holding the selection, which is what you want when you want to delete or keep several elements. Square brackets are for selecting; double square brackets are for extracting.
Knowledge Check: $ needs names Why does l$vect1 fail before you run names(l) <- c('vect1', 'vect2', 'vect3', 'mat')?
Answer: The elements have no names, so there is nothing for $ to match.
$ looks up elements by name, not by position. Until the list has names, use l[[1]] or l[1]. Once named, l$vect1 and l[['vect1']] both return the vector.
Knowledge Check: Coercing to a list How do you turn the vector vect1 <- 1:10 into a list, one element per value?
Answer: as.list(vect1), which gives a list of ten single numbers.
as.list() splits the vector into its individual elements. The same call works on a matrix, in which case it splits the matrix into columns. The reverse direction is unlist().
Knowledge Check: Flattening a list What does unlist(l, use.names = FALSE) do differently from unlist(l)?
Answer: It returns the same values but without the names.
Explanation:unlist() concatenates every element into one vector and keeps the names it finds, so you get a named vector. Setting use.names = FALSE drops the names, which is usually what you want for further calculation.