11. Matrices and other arrays

The open-access textbook Deep R Programming by Marek Gagolewski is, and will remain, freely available for everyone’s enjoyment (also in PDF). It is a non-profit project. This book is still a work in progress. Beta versions of all chapters are already available (proofreading and copyediting pending). In the meantime, any bug/typos reports/fixes are appreciated. Although available online, this is a whole course. It should be read from the beginning to the end. Refer to the Preface for general introductory remarks. Also, check out my other book, Minimalist Data Wrangling with Python [26].

When we equip an atomic or generic vector with the dim attribute, it automatically becomes an object of the S3 class array. In particular, two-dimensional arrays (primary S3 class matrix) allow us to represent tabular data where items are aligned into rows and columns:

structure(1:6, dim=c(2, 3))  # a matrix with 2 rows and 3 columns
##      [,1] [,2] [,3]
## [1,]    1    3    5
## [2,]    2    4    6

Combined with the fact that there are many built-in functions overloaded for the matrix class, we have just opened up a whole world of new possibilities. We explore them in this chapter.

In particular, we discuss how to perform basic algebraic operations such as matrix multiplication, transpose, finding eigenvalues, and performing various decompositions. We also cover data wrangling operations such as array subsetting and column- and rowwise aggregation.

Important

Oftentimes, a numeric matrix with \(n\) rows and \(m\) columns is used to represent \(n\) points (samples) in an \(m\)-dimensional space (with \(m\) numeric features or variables), \(\mathbb{R}^m\).

Furthermore, in the next chapter, we will introduce data frames: matrix-like objects whose columns can be of any (not necessarily the same) type.

11.1. Creating arrays

11.1.1. matrix and array

A matrix can be conveniently created using the matrix function.

(A <- matrix(1:6, byrow=TRUE, nrow=2))
##      [,1] [,2] [,3]
## [1,]    1    2    3
## [2,]    4    5    6

The above converted an atomic vector of length six into a matrix with two rows. The number of columns was determined automatically (ncol=3 could have been passed to get the same result).

Important

By default, the elements of the input vector are read columnwisely:

matrix(1:6, ncol=3)  # byrow=FALSE
##      [,1] [,2] [,3]
## [1,]    1    3    5
## [2,]    2    4    6

A matrix can be equipped with dimension names, being a list of two character vectors of appropriate sizes, labelling each row and column in this order:

matrix(1:6, byrow=TRUE, nrow=2, dimnames=list(c("x", "y"), c("a", "b", "c")))
##   a b c
## x 1 2 3
## y 4 5 6

Alternatively, to create a matrix, we can use the array function, which requires the number of rows and columns to be specified explicitly.

array(1:6, dim=c(2, 3))
##      [,1] [,2] [,3]
## [1,]    1    3    5
## [2,]    2    4    6

The elements are consumed in a column-major manner.

Arrays of dimensionality other than 2 are also possible. Here is a one-dimensional array. When printed, it is indistinguishable from an atomic vector (but the class attribute is still set to array):

array(1:6, dim=6)
## [1] 1 2 3 4 5 6

And now for something completely different: a three-dimensional array of size \(3\times 4\times 2\):

array(1:24, dim=c(3, 4, 2))
## , , 1
##
##      [,1] [,2] [,3] [,4]
## [1,]    1    4    7   10
## [2,]    2    5    8   11
## [3,]    3    6    9   12
##
## , , 2
##
##      [,1] [,2] [,3] [,4]
## [1,]   13   16   19   22
## [2,]   14   17   20   23
## [3,]   15   18   21   24

which can be thought of as two matrices of size \(3\times 4\) (because how else can we print out a 3D object on a 2D console?).

The array function can be fed with the dimnames argument too. For instance, the above three-dimensional hypertable would require a list of three character vectors of sizes 3, 4, and 2, respectively.

Exercise 11.1

The readers are encouraged to try out themselves that 10-dimensional arrays are also possible.

11.1.2. Promoting and stacking vectors

We can promote an ordinary vector to a column vector, i.e., a matrix with one column, by calling:

as.matrix(1:2)
##      [,1]
## [1,]    1
## [2,]    2
cbind(1:2)
##      [,1]
## [1,]    1
## [2,]    2

and to a row vector:

t(1:3)  # transpose
##      [,1] [,2] [,3]
## [1,]    1    2    3
rbind(1:3)
##      [,1] [,2] [,3]
## [1,]    1    2    3

Actually, cbind and rbind stand for column- and row-bind; they allow multiple vectors and matrices to be stacked one after/below another:

rbind(1:4, 5:8, 9:10, 11)  # row bind
##      [,1] [,2] [,3] [,4]
## [1,]    1    2    3    4
## [2,]    5    6    7    8
## [3,]    9   10    9   10
## [4,]   11   11   11   11
cbind(1:4, 5:8, 9:10, 11)  # column bind
##      [,1] [,2] [,3] [,4]
## [1,]    1    5    9   11
## [2,]    2    6   10   11
## [3,]    3    7    9   11
## [4,]    4    8   10   11
cbind(1:2, 3:4, rbind(11:13, 21:23))  # vector, vector, 2x3 matrix
##      [,1] [,2] [,3] [,4] [,5]
## [1,]    1    3   11   12   13
## [2,]    2    4   21   22   23

and so forth.

Unfortunately, the generalised recycling rule is not implemented in full:

cbind(1:4, 5:8, cbind(9:10, 11))  # different from cbind(1:4, 5:8, 9:10, 11)
## Warning in cbind(1:4, 5:8, cbind(9:10, 11)): number of rows of result is
##     not a multiple of vector length (arg 1)
##      [,1] [,2] [,3] [,4]
## [1,]    1    5    9   11
## [2,]    2    6   10   11

The first two arguments are of length four.

11.1.3. Simplifying lists

simplify2array is an extension of the unlist function. Given a list of atomic vectors, each of length one, it will return a flat atomic vector. However, if a list of equisized vectors of greater lengths is given, these will be converted to a matrix.

simplify2array(list(1, 11, 21))  # each of length 1
## [1]  1 11 21
simplify2array(list(1:3, 11:13, 21:23, 31:33))  # each of length 3
##      [,1] [,2] [,3] [,4]
## [1,]    1   11   21   31
## [2,]    2   12   22   32
## [3,]    3   13   23   33
simplify2array(list(1, 11:12, 21:23))  # no can do
## [[1]]
## [1] 1
##
## [[2]]
## [1] 11 12
##
## [[3]]
## [1] 21 22 23

In the second example, each vector becomes a separate column of the resulting matrix[1].

See Section 12.3.7 for a few more examples.

Example 11.2

Quite a few functions call the above automatically (compare the simplify or SIMPLIFY (sic!) argument in sapply, tapply, mapply, replicate, etc.).

For instance:

min_mean_max <- function(x) c(Min=min(x), Mean=mean(x), Max=max(x))
sapply(split(iris[["Sepal.Length"]], iris[["Species"]]), min_mean_max)
##      setosa versicolor virginica
## Min   4.300      4.900     4.900
## Mean  5.006      5.936     6.588
## Max   5.800      7.000     7.900

Take note of what constitutes the columns of the return matrix.

Exercise 11.3

Inspect the behaviour of as.matrix on list arguments. Write your version of simplify2array named as.matrix.list that always returns a matrix. If a list of non-equisized vectors is given, fill the missing cells with NAs.

Important

Sometimes a call to do.call(cbind, x)) might be a better idea than a referral to simplify2array. Provided that x is a list of atomic vectors, it always returns a matrix: shorter vectors are recycled (which might be welcome, but not necessarily).

do.call(cbind, list(a=c(u=1), b=c(v=2, w=3), c=c(i=4, j=5, k=6)))
## Warning in (function (..., deparse.level = 1) : number of rows of result
##     is not a multiple of vector length (arg 2)
##   a b c
## i 1 2 4
## j 1 3 5
## k 1 2 6

Example 11.4

Consider a named toy list of numeric vectors:

x <- list(a=runif(10), b=rnorm(15))

Compare the results generated by sapply (which calls simplify2array):

sapply(x, function(e) c(Mean=mean(e)))
##  a.Mean  b.Mean
## 0.57825 0.12431
sapply(x, function(e) c(Min=min(e), Max=max(e)))
##            a       b
## Min 0.045556 -1.9666
## Max 0.940467  1.7869

with its version based on do.call and cbind:

sapply2 <- function(...)
    do.call(cbind, lapply(...))

sapply2(x, function(e) c(Mean=mean(e)))
##            a       b
## Mean 0.57825 0.12431
sapply2(x, function(e) c(Min=min(e), Max=max(e)))
##            a       b
## Min 0.045556 -1.9666
## Max 0.940467  1.7869

Notice that sapply may return an atomic vector with somewhat surprising names.

11.1.4. Beyond numeric arrays

Arrays built on atomic vectors other than numeric ones are possible too. For instance, we will later stress that comparisons featuring matrices are performed elementwisely. They spawn logical matrices:

A >= 3
##       [,1]  [,2] [,3]
## [1,] FALSE FALSE TRUE
## [2,]  TRUE  TRUE TRUE

Furthermore, matrices of character strings can be useful too:

matrix(strrep(LETTERS[1:6], 1:6), ncol=3)
##      [,1] [,2]   [,3]
## [1,] "A"  "CCC"  "EEEEE"
## [2,] "BB" "DDDD" "FFFFFF"

And, of course, complex matrices:

A + 1i
##      [,1] [,2] [,3]
## [1,] 1+1i 2+1i 3+1i
## [2,] 4+1i 5+1i 6+1i

We are not limited to atomic vectors: lists can be a basis for arrays as well:

matrix(list(1, 11:21, "A", list(1, 2, 3)), nrow=2)
##      [,1]       [,2]
## [1,] 1          "A"
## [2,] integer,11 list,3

Some elements are not displayed correctly, but they are still there.

11.1.5. Internal representation

An object of the S3 class array is an atomic vector or a list equipped with the dims attribute being a vector of nonnegative integers. Interestingly, we do not have to set the class attribute explicitly: the accessor function class will return an implicit[2] class anyway (compare Section 4.4.3).

class(1)  # atomic vector
## [1] "numeric"
class(structure(1, dim=rep(1, 1)))  # 1D array (vector)
## [1] "array"
class(structure(1, dim=rep(1, 2)))  # 2D array (matrix)
## [1] "matrix" "array"
class(structure(1, dim=rep(1, 3)))  # 3D array
## [1] "array"

Note that a two-dimensional array is additionally of the matrix class.

Optional dimension names are represented by means of the dimnames attribute, which is a list of \(d\) character vectors, where \(d\) is the array’s dimensionality.

(A <- structure(1:6, dim=c(2, 3), dimnames=list(letters[1:2], LETTERS[1:3])))
##   A B C
## a 1 3 5
## b 2 4 6
dim(A)  # or attr(A, "dim")
## [1] 2 3
dimnames(A)  # or attr(A, "dimnames")
## [[1]]
## [1] "a" "b"
##
## [[2]]
## [1] "A" "B" "C"

Important

Internally, elements in an array are always stored in the columnwise (column-major, Fortran) order:

as.numeric(A)  # drop all attributes to reveal the underlying numeric vector
## [1] 1 2 3 4 5 6

Setting byrow=TRUE in a call to the matrix only affects the order in which this function reads a given source vector, not the column/row-majorness.

(B <- matrix(1:6, ncol=3, byrow=TRUE))
##      [,1] [,2] [,3]
## [1,]    1    2    3
## [2,]    4    5    6
as.numeric(B)
## [1] 1 4 2 5 3 6

The two said special attributes can be modified through the replacement functions `dim<-` and `dimnames<-` (and, of course, `attr<-` as well). In particular, changing dim does not alter the underlying atomic vector; it only affects how other functions, including the corresponding print method, interpret their placement on a virtual grid:

`dim<-`(A, c(3, 2))  # not the same as transpose of A
##      [,1] [,2]
## [1,]    1    4
## [2,]    2    5
## [3,]    3    6

We obtained a different view of the same flat data vector. Also, the dimnames attribute was dropped as its size became incompatible with the newly requested dimensionality.

Exercise 11.5

Study the source code of the nrow, NROW, ncol, NCOL, rownames, row.names, and colnames functions.

Interestingly, for one-dimensional arrays, the names function returns a reasonable value (based on the dimnames attribute, which is a list featuring one character vector), despite the names attribute’s not being set.

What is more, dimnames itself can be named:

names(dimnames(A)) <- c("ROWS", "COLUMNS")
print(A)
##     COLUMNS
## ROWS A B C
##    a 1 3 5
##    b 2 4 6

It is still a numeric matrix, but the presentation thereof is slightly prettified.

Exercise 11.6

outer applies an elementwisely vectorised function on each pair of elements from two vectors, forming a two-dimensional result grid. Based on two calls to rep, implement it yourself.

Some examples:

outer(c(x=1, y=10, z=100), c(a=1, b=2, c=3, d=4), "*")  # multiplication
##     a   b   c   d
## x   1   2   3   4
## y  10  20  30  40
## z 100 200 300 400
outer(c("A", "B"), 1:8, paste, sep="-")  # concatenate strings
##      [,1]  [,2]  [,3]  [,4]  [,5]  [,6]  [,7]  [,8]
## [1,] "A-1" "A-2" "A-3" "A-4" "A-5" "A-6" "A-7" "A-8"
## [2,] "B-1" "B-2" "B-3" "B-4" "B-5" "B-6" "B-7" "B-8"

Exercise 11.7

Show how match(y, z) can be expressed using outer. Is its time and memory complexity optimal, though?

Exercise 11.8

table creates a contingency matrix/array that counts the number of unique pairs of corresponding elements from one or more vectors of equal lengths. Implement its one- and two-argument version based on tabulate.

For example:

tips <- read.csv(paste0("https://github.com/gagolews/teaching-data/raw/",
    "master/other/tips.csv"), comment.char="#")  # a data.frame (list)
table(tips[["day"]])
##
##  Fri  Sat  Sun Thur
##   19   87   76   62
table(tips[["smoker"]], tips[["day"]])
##
##       Fri Sat Sun Thur
##   No    4  45  57   45
##   Yes  15  42  19   17

11.2. Array indexing

Array subsetting can be performed by means of an overloaded[3] `[` method, which we will usually provide with many indexers: two in the matrix case; see help("[").

Throughout this section, we refer to the two following example matrices:

(A <- matrix(1:12, byrow=TRUE, nrow=3))
##      [,1] [,2] [,3] [,4]
## [1,]    1    2    3    4
## [2,]    5    6    7    8
## [3,]    9   10   11   12
B <- A
dimnames(B) <- list(
    c("a", "b", "c"),     # row labels
    c("x", "y", "z", "w") # column labels
)
B
##   x  y  z  w
## a 1  2  3  4
## b 5  6  7  8
## c 9 10 11 12

Subsetting higher-dimensional arrays will be covered at the end.

11.2.1. Arrays are built on basic vectors

Subsetting based on one indexer (as in Chapter 5) will refer to the underlying flat vector.

For instance:

A[6]
## [1] 10

It is the element in the third row, second column. Recall that values are stored in the column-major order.

11.2.2. Selecting individual elements

Mathematically, we say that our example \(3\times 4\) real matrix \(\mathbf{A}\in\mathbb{R}^{3\times 4}\) is like:

\[ \mathbf{A}= \left[ \begin{array}{cccc} a_{1, 1} & a_{1, 2} & a_{1, 3} & a_{1, 4} \\ a_{2, 1} & a_{2, 2} & a_{2, 3} & a_{2, 4} \\ a_{3, 1} & a_{3, 2} & a_{3, 3} & a_{3, 4} \\ \end{array} \right] = \left[ \begin{array}{cccc} 1 & 2 & 3 & 4\\ 5 & 6 & 7 & 8 \\ 9 & 10& 11& 12 \\ \end{array} \right]. \]

Matrix elements are aligned in a two-dimensional grid. They are organised into rows and columns. Hence, we can pinpoint a cell using two indexes: \(a_{i,j}\) refers to the \(i\)-th row and the \(j\)-th column.

Similarly in R:

A[3, 2]  # 3rd row, 2nd column
## [1] 10
B["c", "y"]  # using dimnames == B[3, 2]
## [1] 10

11.2.3. Selecting rows and columns

Some textbooks, and we are fond of this notation here as well, mark with \(\mathbf{a}_{i,\cdot}\) a vector that consists of all the elements in the \(i\)-th row and with \(\mathbf{a}_{\cdot,j}\) all items in the \(j\)-th column.

In R, these will correspond to one of the indexers being left out.

A[3, ]  # 3rd row
## [1]  9 10 11 12
A[, 2]  # 2nd column
## [1]  2  6 10
B["c", ]  # or B[3, ]
##  x  y  z  w
##  9 10 11 12
B[, "y"]  # or B[, 2]
##  a  b  c
##  2  6 10

Let us stress that A[1], A[1, ], and A[, 1] have all different meanings. Also, we see that the results’ dimnames are adjusted accordingly; see also unname, which can take care of them once and for all.

Exercise 11.9

Use duplicated to remove repeating rows in a given numeric matrix (see also unique).

11.2.4. Dropping dimensions

Extracting an individual element or a single row/column from a matrix brings about an atomic vector. If the dim attribute consists of 1s only, it will be removed whatsoever.

In order to obtain proper row and column vectors, we can request the preservation of the dimensionality of the output object (and, more precisely, the length of dim). It can be done by passing drop=FALSE to `[`.

A[1, 2, drop=FALSE]  # 1st row, 2nd columns
##      [,1]
## [1,]    2
A[1,  , drop=FALSE]  # 1st row
##      [,1] [,2] [,3] [,4]
## [1,]    1    2    3    4
A[ , 2, drop=FALSE]  # 2nd column
##      [,1]
## [1,]    2
## [2,]    6
## [3,]   10

Important

Unfortunately, the drop argument defaults to TRUE. Many bugs could be avoided otherwise, primarily when the indexers are generated programmatically.

See also the drop function, which removes the dimensions with only one level.

Note

For list-based matrices, we can also use a multi-argument version of `[[` to extract the individual elements.

C <- matrix(list(1, 11:12, 21:23, 31:34), nrow=2)
C[1, 2]  # for `[`, input type is the same as the output type, hence a list
## [[1]]
## [1] 21 22 23
C[1, 2, drop=FALSE]
##      [,1]
## [1,] integer,3
C[[1, 2]]  # extract
## [1] 21 22 23

11.2.5. Selecting submatrices

Indexing based on two vectors, both of length two or more, extracts a sub-block of a given matrix:

A[1:2, c(1, 2, 4)]  # rows 1 and 2, columns 1, 2, and 4
##      [,1] [,2] [,3]
## [1,]    1    2    4
## [2,]    5    6    8
B[c("a", "b"), -3]
##   x y w
## a 1 2 4
## b 5 6 8

Note again that drop=TRUE is the default, which affects the behaviour if one of the indexers is a scalar.

A[c(1, 3), 3]
## [1]  3 11
A[c(1, 3), 3, drop=FALSE]
##      [,1]
## [1,]    3
## [2,]   11

Exercise 11.10

Overload the split function for the matrix class so that, given a matrix with \(n\) rows and an object of the class factor of length \(n\) (or a list of such objects), a list of \(n\) matrices is returned. For example:

split.matrix <- ...to.do...
A <- matrix(1:12, nrow=3)  # matrix whose rows are to be split
s <- factor(c("a", "b", "a"))  # determines the grouping of rows
split(A, s)
## $a
##      [,1] [,2] [,3] [,4]
## [1,]    1    4    7   10
## [2,]    3    6    9   12
##
## $b
##      [,1] [,2] [,3] [,4]
## [1,]    2    5    8   11

11.2.6. Selecting elements based on logical vectors

Logical vectors can also be used as indexers, with consequences that are not hard to guess:

A[c(TRUE, FALSE, TRUE), -1]  # select 1st and 3rd row, all but 1st column
##      [,1] [,2] [,3]
## [1,]    4    7   10
## [2,]    6    9   12
B[B[, "x"]>1 & B[, "x"]<=9, ]  # all rows where x is in (1, 9]
##   x  y  z  w
## b 5  6  7  8
## c 9 10 11 12
A[2, colMeans(A)>6, drop=FALSE]  # 2nd row of the columns with means > 6
##      [,1] [,2]
## [1,]    8   11

Note

Section 11.3 notes that comparisons involving matrices are performed in an elementwise manner, for example:

A>7
##       [,1]  [,2]  [,3] [,4]
## [1,] FALSE FALSE FALSE TRUE
## [2,] FALSE FALSE  TRUE TRUE
## [3,] FALSE FALSE  TRUE TRUE

Such logical matrices can be used to index other matrices of the same size. This kind of indexing always gives rise to a (flat) vector:

A[A>7]
## [1]  8  9 10 11 12

It is nothing else than the single-indexer subsetting involving two flat vectors (a numeric and a logical one); the dim attributes are not considered here.

Exercise 11.11

Implement your versions of max.col, lower.tri, and upper.tri.

11.2.7. Selecting based on two-column numeric matrices

We can also index a matrix A with a two-column matrix of positive integers I, for instance:

(I <- cbind(
    c(1, 3, 2, 1, 2),
    c(2, 3, 2, 1, 4)
))
##      [,1] [,2]
## [1,]    1    2
## [2,]    3    3
## [3,]    2    2
## [4,]    1    1
## [5,]    2    4

Now A[I] gives easy access to:

• A[I[1, 1], I[1, 2]],

• A[I[2, 1], I[2, 2]],

• A[I[3, 1], I[3, 2]],

• …

and so forth. In other words, each row of I gives the coordinates of the elements to extract.

A[I]
## [1]  4  9  5  1 11

This is exactly A[1, 2], A[3, 3], A[2, 2], A[1, 1], A[2, 4]. The result is always a flat vector.

Note

which can also return a list of index matrices:

which(A>7, arr.ind=TRUE)
##      row col
## [1,]   2   3
## [2,]   3   3
## [3,]   1   4
## [4,]   2   4
## [5,]   3   4

Moreover, arrayInd can be used to convert flat indexes to multidimensional ones.

Exercise 11.12

Implement your version of arrayInd and a function performing the inverse operation.

Exercise 11.13

Write your version of diag.

11.2.8. Higher-dimensional arrays

For \(d\)-dimensional arrays, indexing can involve up to \(d\) indexes. It is particularly valuable for arrays with the dimnames attribute set representing contingency tables over a Cartesian product of multiple factors. The built-in datasets::Titanic object is an example of this:

str(dimnames(Titanic))  # for reference (note that dimnames are named)
## List of 4
##  $ Class   : chr [1:4] "1st" "2nd" "3rd" "Crew"
##  $ Sex     : chr [1:2] "Male" "Female"
##  $ Age     : chr [1:2] "Child" "Adult"
##  $ Survived: chr [1:2] "No" "Yes"
Titanic["Crew", "Male", "Adult", "Yes"]
## [1] 192

gives the number of adult male crew members who survived the accident. Also:

Titanic["Crew", , "Adult", ]
##         Survived
## Sex       No Yes
##   Male   670 192
##   Female   3  20

and so on.

Exercise 11.14

Check if the above four-dimensional array can be indexed using matrices with four columns.

11.2.9. Replacing elements

There is also a multidimensional version of the replacement subsetting function, `[<-`.

Generally, subsetting drops all attributes except names, dim, and dimnames (unless it does not make sense otherwise). The replacement variant of the index operator modifies vector values but generally preserves all the attributes.

This enables transforming matrix elements like:

B[B<10] <- A[B<10]^2
print(B)
##   x  y  z   w
## a 1 16 49 100
## b 4 25 64 121
## c 9 10 11  12
B[] <- rep(seq_len(NROW(B)), NCOL(B))  # NOT the same as B <- ...
print(B)
##   x y z w
## a 1 1 1 1
## b 2 2 2 2
## c 3 3 3 3

Take note of the preservation of dim and dimnames.

Exercise 11.15

Given a character matrix with entities that can be interpreted as numbers like:

(X <- rbind(x=c(a="1", b="2"), y=c("3", "4")))
##   a   b
## x "1" "2"
## y "3" "4"

convert it to a numeric matrix with a single line of code. Preserve all attributes.

11.3. Common operations

11.3.1. Matrix transpose

The matrix transpose, mathematically denoted with \(\mathbf{A}^T\), is available via a call to t:

(A <- matrix(1:6, byrow=TRUE, nrow=2))
##      [,1] [,2] [,3]
## [1,]    1    2    3
## [2,]    4    5    6
t(A)
##      [,1] [,2]
## [1,]    1    4
## [2,]    2    5
## [3,]    3    6

Hence, if \(\mathbf{B}=\mathbf{A}^T\), then it is a matrix such that \(b_{i,j}=a_{j,i}\). In other words, in the transposed matrix, rows become columns, and columns become rows.

For higher-dimensional arrays, a generalised transpose can be achieved with aperm (try permuting the dimensions of Titanic). Also, the conjugate transpose of a complex matrix \(\mathbf{A}\) is done via Conj(t(A)).

11.3.2. Vectorised mathematical functions

Vectorised functions such as sqrt, abs, round, log, exp, cos, sin, etc., operate on each array element[4].

A <- matrix(1/(1:6), nrow=2)
round(A, 2)  # rounds every element in A
##      [,1] [,2] [,3]
## [1,]  1.0 0.33 0.20
## [2,]  0.5 0.25 0.17

Exercise 11.16

Using a single call to matplot, which accepts the y argument to be a matrix, draw a plot of \(\sin(x)\), \(\cos(x)\), \(|\sin(x)|\), and \(|\cos(x)|\) for \(x\in[-2\pi, 6\pi]\).

11.3.3. Aggregating rows and columns

When we call an aggregation function on an array, it will reduce all elements to a single number:

(A <- matrix(1:12, byrow=TRUE, nrow=3))
##      [,1] [,2] [,3] [,4]
## [1,]    1    2    3    4
## [2,]    5    6    7    8
## [3,]    9   10   11   12
mean(A)
## [1] 6.5

The apply function may be used to summarise individual rows or columns in a matrix:

• apply(A, 1, f) applies a given function f on each row of a matrix A;

• apply(A, 2, f) applies f on each column of A.

For instance:

apply(A, 1, mean)  # synonym: rowMeans(A)
## [1]  2.5  6.5 10.5
apply(A, 2, mean)  # synonym: colMeans(A)
## [1] 5 6 7 8

The function being applied does not have to return a single number:

apply(A, 2, range)  # min and max
##      [,1] [,2] [,3] [,4]
## [1,]    1    2    3    4
## [2,]    9   10   11   12
apply(A, 1, function(row) c(Min=min(row), Mean=mean(row), Max=max(row)))
##      [,1] [,2] [,3]
## Min   1.0  5.0  9.0
## Mean  2.5  6.5 10.5
## Max   4.0  8.0 12.0

Take note of the columnwise order of the output values.

apply works on higher-dimensional arrays:

apply(Titanic, 1, mean)  # 1st dimension - Class
##     1st     2nd     3rd    Crew
##  40.625  35.625  88.250 110.625
apply(Titanic, c(1, 3), mean)  # w.r.t. Class (1st) and Age (3rd)
##       Age
## Class  Child  Adult
##   1st   1.50  79.75
##   2nd   6.00  65.25
##   3rd  19.75 156.75
##   Crew  0.00 221.25

11.3.4. Binary operators

In Section 5.5, we stated that binary elementwise operations, such as addition or multiplication, preserve the attributes of the longer input or both (with the first argument preferred to the second) if they are of equal sizes.

Taking into account that:

• an array is simply a flat vector equipped with the dim attribute, and

• we refer to the respective default methods when applying binary operators

allows us to deduce how `+`, `<=`, `&`, etc. behave in several different contexts.

Array-Array. First, let us note what happens when we operate on two arrays of identical dimensionalities.

(A <- rbind(c(1, 10, 100), c(-1, -10, -100)))
##      [,1] [,2] [,3]
## [1,]    1   10  100
## [2,]   -1  -10 -100
(B <- matrix(1:6, byrow=TRUE, nrow=2))
##      [,1] [,2] [,3]
## [1,]    1    2    3
## [2,]    4    5    6
A + B  # elementwise addition
##      [,1] [,2] [,3]
## [1,]    2   12  103
## [2,]    3   -5  -94
A * B  # elementwise multiplication (not: algebraic matrix multiply)
##      [,1] [,2] [,3]
## [1,]    1   20  300
## [2,]   -4  -50 -600

They are simply the addition and multiplication of the corresponding elements of two given matrices.

Array-Scalar. Second, we can apply scalar-matrix operations:

(-1)*B
##      [,1] [,2] [,3]
## [1,]   -1   -2   -3
## [2,]   -4   -5   -6
A^2
##      [,1] [,2]  [,3]
## [1,]    1  100 10000
## [2,]    1  100 10000

They multiplied each element in B by -1 and squared every element in A, respectively.

The behaviour of relational operators is similar:

A >= 1 & A <= 100
##       [,1]  [,2]  [,3]
## [1,]  TRUE  TRUE  TRUE
## [2,] FALSE FALSE FALSE

Array-Vector. Next, based on the recycling rule and the fact that elements are ordered columnwisely, we get that:

B * c(10, 100)
##      [,1] [,2] [,3]
## [1,]   10   20   30
## [2,]  400  500  600

It multiplied every element in the first row by 10 and each element in the second row by 100.

If we wish to multiply each element in the first, second, …, etc. column by the first, second, …, etc. value in a vector, we should not call:

B * c(1, 100, 1000)
##      [,1] [,2] [,3]
## [1,]    1 2000  300
## [2,]  400    5 6000

but rather:

t(t(B) * c(1, 100, 1000))
##      [,1] [,2] [,3]
## [1,]    1  200 3000
## [2,]    4  500 6000

or:

t(apply(B, 1, `*`, c(1, 100, 1000)))
##      [,1] [,2] [,3]
## [1,]    1  200 3000
## [2,]    4  500 6000

Exercise 11.17

Write a function that standardises the values in each column of a given matrix: for each column, from every element, subtract the mean and then divide it by the standard deviation. Try to do it in a few different ways, including via a call to apply, sweep, scale, or based solely on arithmetic operators.

Note

Some sanity checks are done on the dim attributes, so not every configuration is possible. Notice the following peculiarities:

getOption("error")
## NULL
A + t(B)  # dim==c(2, 3) vs dim==c(3, 2)
## Error in A + t(B): non-conformable arrays
A * cbind(1, 10, 100)  # this is too good to be true
## Error in A * cbind(1, 10, 100): non-conformable arrays
A * rbind(1, 10)  # but A * c(1, 10) works...
## Error in A * rbind(1, 10): non-conformable arrays
A + 1:12
## Error in eval(expr, envir, enclos): dims [product 6] do not match the
##     length of object [12]
A + 1:5  # partial recycling is okay
## Warning in A + 1:5: longer object length is not a multiple of shorter
##     object length
##      [,1] [,2] [,3]
## [1,]    2   13  105
## [2,]    1   -6  -99

11.4. Numerical matrix algebra (*)

Many data analysis and machine learning algorithms, in their essence, involve quite simple matrix algebra and numerical mathematics. Suffice it to say that anyone serious about data science and scientific computing should learn the necessary theory; see, for example, [30] and [31].

R is a convenient interface to the well-tested and stable algorithms from, amongst others, LAPACK and BLAS[5]. Below we mention only a few of them. Many third-party packages feature hundreds of algorithms tackling differential equations, constrained and unconstrained optimisation, etc. Exploring the relevant CRAN Task Views can give a comprehensive overview.

11.4.1. Matrix multiplication

`*` performs elementwise multiplication. For what we call (algebraic) matrix multiplication, we use the `%*%` operator.

Here is a quick refresher from a basic linear algebra course. The matrix multiplication can only be performed on two matrices of compatible sizes: the number of columns in the left matrix must match the number of rows in the right operand.

Given \(\mathbf{A}\in\mathbb{R}^{n\times p}\) and \(\mathbf{B}\in\mathbb{R}^{p\times m}\), their multiply is a matrix \(\mathbf{C}=\mathbf{A}\mathbf{B}\in\mathbb{R}^{n\times m}\) such that \(c_{i,j}\) is the dot product of the \(i\)-th row in \(\mathbf{A}\) and the \(j\)-th column in \(\mathbf{B}\):

\[ c_{i,j} = \mathbf{a}_{i,\cdot} \cdot \mathbf{b}_{\cdot,j} = \sum_{k=1}^p a_{i,k} b_{k, j}, \]

for \(i=1,\dots,n\) and \(j=1,\dots,m\).

For instance:

(A <- rbind(c(1, 1, 1), c(-1, 1, 0)))
##      [,1] [,2] [,3]
## [1,]    1    1    1
## [2,]   -1    1    0
(B <- rbind(c(3, -1), c(1, 2), c(6, 1)))
##      [,1] [,2]
## [1,]    3   -1
## [2,]    1    2
## [3,]    6    1
A %*% B
##      [,1] [,2]
## [1,]   10    2
## [2,]   -2    3

Note

When applying `%*%` on one or more flat vectors, their dimensionality will be promoted automatically to make the operation possible. However, c(a, b) %*% c(c, d) gives a scalar \(ac+bd\), and not a \(2\times 2\) matrix.

Further, crossprod(A, B) yields \(\mathbf{A}^T \mathbf{B}\) and tcrossprod(A, B) determines \(\mathbf{A} \mathbf{B}^T\) more efficiently than relying on `%*%`. We can omit the second argument and get \(\mathbf{A}^T \mathbf{A}\) and \(\mathbf{A} \mathbf{A}\), respectively.

crossprod(c(1, 1))  # Euclidean norm squared
##      [,1]
## [1,]    2
crossprod(c(1, 1), c(-1, 1))  # dot product of two vectors
##      [,1]
## [1,]    0
crossprod(A)  # same as t(A) %*% A, i.e., dot products of all column pairs
##      [,1] [,2] [,3]
## [1,]    2    0    1
## [2,]    0    2    1
## [3,]    1    1    1

Recall that if the dot product of two vectors equals 0, we say that they are orthogonal (perpendicular).

Exercise 11.18

(*) Write your versions of cov and cor: functions to compute the covariance and correlation matrices. Make use of the fact that the former can be determined with crossprod based on a centred version of an input matrix.

11.4.2. Solving systems of linear equations

The solve function can be used to solve \(m\) systems of \(n\) linear equations of the form \(\mathbf{A}\mathbf{X}=\mathbf{B}\), where \(\mathbf{A}\in\mathbb{R}^{n\times n}\) and \(\mathbf{X},\mathbf{B}\in\mathbb{R}^{n\times m}\) (via the LU decomposition with partial pivoting and row interchanges).

11.4.3. Norms and metrics

Given an \(n\times m\) matrix \(\mathbf{A}\), calling norm(A, "1"), norm(A, "2"), and norm(A, "I"), we can compute the operator norms:

\[ \begin{array}{lcl} \|\mathbf{A}\|_1 &=& \max_{j=1,\dots,m} \sum_{i=1}^n |a_{i,j}|,\\ \|\mathbf{A}\|_2 &=& \sigma_1(\mathbf{A}) = \sup_{\mathbf{0}\neq\mathbf{x}\in\mathbb{R}^m} \frac{\|\mathbf{A}\mathbf{x}\|_2}{\|\mathbf{x}\|_2} \\ \|\mathbf{A}\|_I &=& \max_{i=1,\dots,n} \sum_{j=1}^m |a_{i,j}|,\\ \end{array} \]

where \(\sigma_1\) gives the largest singular value (see below).

Also, passing "F" as the second argument yields the Frobenius norm, \(\|\mathbf{A}\|_F = \sqrt{\sum_{i=1}^n \sum_{j=1}^m a_{i,j}^2}\), and "M" computes the max norm, \(\|\mathbf{A}\|_M = \max_{{i=1,\dots,n\atop j=1,\dots,m}} |a_{i,j}|\).

If \(\mathbf{A}\) is a column vector, then \(\|\mathbf{A}\|_F\) and \(\|\mathbf{A}\|_2\) are equivalent. They are referred to as the Euclidean norm. Moreover, \(\|\mathbf{A}\|_M=\|\mathbf{A}\|_I\) give the supremum norm and outputs \(\|\mathbf{A}\|_1\) the Manhattan (taxicab) one.

Exercise 11.19

Given an \(n\times m\) matrix \(\mathbf{A}\) representing \(m\) vectors in \(\mathbb{R}^n\), normalise each column so that you obtain \(m\) unit vectors, i.e., whose Euclidean norm equals 1.

Further, dist determines all pairwise distances between a set of \(n\) vectors in \(\mathbb{R}^m\), written as an \(n\times m\) matrix.

For example, let us consider three vectors in \(\mathbb{R}^2\):

(X <- rbind(c(1, 1), c(1, -2), c(0, 0)))
##      [,1] [,2]
## [1,]    1    1
## [2,]    1   -2
## [3,]    0    0
as.matrix(dist(X, "euclidean"))
##        1      2      3
## 1 0.0000 3.0000 1.4142
## 2 3.0000 0.0000 2.2361
## 3 1.4142 2.2361 0.0000

Thus, the distance between the first and the third vector is roughly 1.41421. Euclidean, maximum, Manhattan, and Canberra distances/metrics are available, amongst others.

Exercise 11.20

dist returns an object of S3 class dist. Inspect how it is represented.

Example 11.21

adist implements a couple of string metrics. For example:

x <- c("spam", "bacon", "eggs", "spa", "spams", "legs")
names(x) <- x
(d <- adist(x))
##       spam bacon eggs spa spams legs
## spam     0     5    4   1     1    4
## bacon    5     0    5   5     5    5
## eggs     4     5    0   4     4    2
## spa      1     5    4   0     2    4
## spams    1     5    4   2     0    4
## legs     4     5    2   4     4    0

gives the Levenshtein distances between each pair of strings. In particular, we need two edit operations (character insertions, deletions, or replacements) to turn "eggs" into "legs" (add l and remove g).

Example 11.22

Objects of the class dist can be used to perform hierarchical clusterings of datasets. For example:

h <- hclust(as.dist(d), method="average")  # see also: plot(h, labels=x)
cutree(h, 3)
##  spam bacon  eggs   spa spams  legs
##     1     2     3     1     1     3

determines three clusters using the average linkage strategy ("legs" and "eggs" are grouped together, "spam", "spa", "spams" form another cluster, and "bacon" is a singleton).

11.4.4. Eigenvalues and eigenvectors

eigen returns a sequence of eigenvalues \((\lambda_1,\dots,\lambda_n)\) (ordered nondecreasingly w.r.t. \(|\lambda_i|\)) and a matrix \(\mathbf{V}\) whose columns define the corresponding eigenvectors (scaled to unit length) of a given matrix \(\mathbf{X}\). To recall, by definition, it holds that \(\mathbf{X}\mathbf{v}_{\cdot,i}=\lambda_i\mathbf{v}_{\cdot,i}\).

Here are the eigenvalues and the corresponding eigenvectors of an example matrix (defining rotation in 2D by \(\pi/3\)):

(R <- rbind(c(cos(pi/3), -sin(pi/3)), c(sin(pi/3), cos(pi/3))))
##         [,1]     [,2]
## [1,] 0.50000 -0.86603
## [2,] 0.86603  0.50000
eigen(R)
## eigen() decomposition
## $values
## [1] 0.5+0.86603i 0.5-0.86603i
##
## $vectors
##                  [,1]             [,2]
## [1,] 0.70711+0.00000i 0.70711+0.00000i
## [2,] 0.00000-0.70711i 0.00000+0.70711i

Example 11.23

Consider a pseudorandom sample from a bivariate[6] normal distribution; see Figure 11.1.

Z <- matrix(rnorm(2000), ncol=2)  # independent N(0, 1)
Z <- Z %*% rbind(c(1, 0), c(0, sqrt(5)))  # scaling
Z <- Z %*% R  # rotation
Z <- t(c(10, -5) + t(Z))  # translation
plot(Z, asp=1)

Figure 11.1 Example bivariate normal sample

It is known that eigenvectors of the covariance matrix correspond to the principal components of the original dataset. Furthermore, the eigenvalues give the variance explained by them.

eigen(cov(Z))
## eigen() decomposition
## $values
## [1] 5.18609 0.98386
##
## $vectors
##          [,1]     [,2]
## [1,] -0.86715  0.49804
## [2,] -0.49804 -0.86715

this roughly corresponds to the principal directions \([\sin(\pi/3), \cos(\pi/3)]\) and the thereto-orthogonal \([\cos(\pi/3), -\sin(\pi/3)]\) with variances of 5 and 1, respectively. Still, this method of performing a PCA is not particularly numerically stable; see below for an alternative.

11.4.5. QR decomposition

We say that a real \(n\times m\) matrix \(\mathbf{Q}\), \(n\ge m\), is orthogonal whenever \(\mathbf{Q}^T \mathbf{Q} = \mathbf{I}\) (identity matrix). This is equivalent to \(\mathbf{Q}\)’s being orthogonal unit vectors (if \(\mathbf{Q}\) is a square matrix, then \(\mathbf{Q}^T=\mathbf{Q}^{-1}\) if and only if \(\mathbf{Q}^T \mathbf{Q} = \mathbf{Q} \mathbf{Q}^T = \mathbf{I}\)).

Let \(\mathbf{A}\) be a real[7] \(n\times m\) matrix with \(n\ge m\). Then \(\mathbf{A}=\mathbf{Q}\mathbf{R}\) is its QR decomposition (in the so-called narrow form), if \(\mathbf{Q}\) is an orthogonal \(n\times m\) matrix and \(\mathbf{R}\) is an upper triangular \(m\times m\) one.

The qr function returns an object of S3 class qr from which we can extract the two components; see the qr.Q and qr.R functions.

Example 11.24

Let \(\mathbf{X}\) be an \(n\times m\) data matrix, representing \(n\) points in \(\mathbb{R}^m\), and a vector \(\mathbf{y}\in\mathbb{R}^n\) of the desired outputs corresponding to each input. For fitting a linear model \(\mathbf{x}^T \boldsymbol\theta\), where \(\boldsymbol\theta\) is a vector of \(m\) parameters, we can use the method of least squares, which minimises

\[ \mathcal{L}(\boldsymbol\theta) = \sum_{i=1}^n \left( \mathbf{x}_{i,\cdot}^T \boldsymbol\theta - y_i \right)^2 = \|\mathbf{X} \boldsymbol\theta - \mathbf{y}\|_2^2 \]

It might be shown that if \(\mathbf{X}=\mathbf{Q}\mathbf{R}\), then \(\boldsymbol\theta = \left(\mathbf{X}^T \mathbf{X} \right)^{-1} \mathbf{X}^T \mathbf{y} = \mathbf{R}^{-1}\mathbf{Q}^T\mathbf{y}\), which can conveniently be determined via a call to qr.coef.

In particular, we can fit a simple linear regression model \(y=ax+b\) by considering \(\mathbf{X} = [x, 1]\) and \(\boldsymbol\theta = [a, b]\), for example (see Figure 11.2):

x <- cars[["speed"]]
y <- cars[["dist"]]
X <- cbind(x, 1)  # the model is theta[1]*x + theta[2]*1
qrX <- qr(X)
(theta <- solve(qr.R(qrX)) %*% t(qr.Q(qrX)) %*% y)   # or: qr.coef(qrX, y)
##       [,1]
## x   3.9324
##   -17.5791
plot(x, y, xlab="speed", ylab="dist")  # scatter plot
abline(theta[2], theta[1], lty=2)  # add the regression line

Figure 11.2 The built-in cars dataset and the fitted regression line

solve with one argument determines the inverse of a given matrix. The fitted model is \(y=3.93241x-17.5791\).

The same approach is used by lm.fit, the workhorse behind the lm method accepting an R formula (which some readers might be familiar with; compare Section 17.6).

lm.fit(cbind(x, 1), y)[["coefficients"]]  # also: lm(dist~speed, data=cars)
##        x
##   3.9324 -17.5791

11.4.6. SVD decomposition

Given a real \(n\times m\) matrix \(\mathbf{X}\), its singular value decomposition (SVD) is given by \(\mathbf{X}=\mathbf{U} \mathbf{D} \mathbf{V}^T\), where \(\mathbf{D}\) is a \(p\times p\) diagonal matrix (featuring the so-called singular values of \(\mathbf{X}\), \(d_{1,1}\ge\dots\ge d_{p,p}\ge 0\), \(p=\min\{n,m\}\)), and \(\mathbf{U}\) and \(\mathbf{V}\) are orthogonal matrices of dimension \(n\times p\) and \(m\times p\), respectively.

svd may not only be used to determine the solution to linear regression[8] but also to perform the principal component analysis[9]. Namely, \(\mathbf{V}\) gives the eigenvectors of \(\mathbf{X}^T \mathbf{X}\). Assuming that \(\mathbf{X}\) is centred at \(\boldsymbol{0}\), the latter is precisely its scaled covariance matrix.

Example 11.25

Continuing the PCA example above, we can determine the principal directions also by calling:

Zc <- apply(Z, 2, function(x) x-mean(x))  # centred version of Z
svd(Zc)[["v"]]
##          [,1]     [,2]
## [1,] -0.86715  0.49804
## [2,] -0.49804 -0.86715

11.5. S4 classes (*)

The concept of the S3-style object-oriented programming is based on a brilliantly simple idea (see Chapter 10): calling a generic f(x) dispatches automatically to a method f.class_of_x(x) or f.default(x) in the case where the former does not exist. Naturally, it has some inherent limitations:

• classes cannot be formally defined; the class attribute may be assigned arbitrarily onto any object[10],

• argument dispatch is performed only[11] with regard to one data type[12].

In most cases, and with an appropriate level of mindfulness, this is not a problem at all. However, it is a typical condition of programmers who come to our world from more mainstream languages (e.g., C++; yours truly included) until they appreciate the true beauty of R’s being somewhat different. Before they fully develop such an acquired taste, though, they grow restless as “R is not a real object-oriented system because it lacks polymorphism, encapsulation, formal inheritance, and so on, and something must be done about it”. The truth is that it had not had to, but with high probability, it would have anyway in one way or another.

And so the fourth version of the S language was introduced in 1998 (see [9]). It brought a new object-oriented system, which we are used to referring to as S4. Its R version is given by the methods package. Below we discuss it briefly; for more details, see help("Classes_Details") and help("Methods_Details") as well as [10] and [11].

Note

(*) S4 was loosely inspired by the Common Lisp Object System (with its defclass, defmethod, etc.; see, e.g., [20]). In the current author’s opinion, the S4 system is somewhat of an afterthought. Due to appendages like this, R seems like a patchwork language; suffice it to say that it was not the last attempt to do a somewhat more real OOP in the overall functional R: the story will resume in Section 16.1.5.

The main problem with all the OOP approaches is that each of them is parallel to S3 which never lost its popularity and is still at the very core of our language. We are thus covering them for the sake of completeness because that’s what must be done. After all, with non-zero probability, the reader will sooner or later come across such objects (e.g., below, we explain the meaning of a notation like x@slot). Also, yours truly rebelliously suggests taking statements such as “for new projects, it is recommended to use the more flexible and robust S4 scheme provided in the methods package” (see help("UseMethod")) with a pinch of salt.

11.5.1. Defining S4 classes

An S4 class can formally be registered through a call to setClass. For instance:

library("methods")  # in the case where it is not auto-loaded
setClass("categorical", slots=c(data="integer", levels="character"))

defines a class named categorical. It has two slots: data and levels being integer and character vectors, respectively. This notation is already quite peculiar. There is no assignment suggesting that we have introduced something novel.

An object of the above class can be instantiated by calling new:

z <- new("categorical", data=c(1L, 2L, 2L, 1L, 1L), levels=c("a", "b"))
print(z)
## An object of class "categorical"
## Slot "data":
## [1] 1 2 2 1 1
##
## Slot "levels":
## [1] "a" "b"

That z is of the recently-introduced class can be verified as follows:

is(z, "categorical")
## [1] TRUE
class(z)  # also: attr(z, "class")
## [1] "categorical"
## attr(,"package")
## [1] ".GlobalEnv"

Important

Some R packages require the methods package only for the sake of being able to access the handy is function. It does not mean they are defining new S4 classes.

Note

S4 objects are marked as being of the following basic type:

typeof(z)
## [1] "S4"

See Section 1.12 of [65] for technical details on how they are internally represented. In particular, in our case, all the slots are simply stored as object attributes:

attributes(z)
## $data
## [1] 1 2 2 1 1
##
## $levels
## [1] "a" "b"
##
## $class
## [1] "categorical"
## attr(,"package")
## [1] ".GlobalEnv"

11.5.2. Accessing slots

Reading or writing slot contents can be done via the `@` operator and the slot function or their replacement versions.

z@data  # or slot(z, "data")
## [1] 1 2 2 1 1
z@levels <- c("A", "B")

Note

The `@` operator can only be used on S4 objects, and some sanity checks are automatically performed:

z@unknown <- "spam"
## Error in (function (cl, name, valueClass) : 'unknown' is not a slot in
##     class "categorical"
z@data <- "spam"
## Error in (function (cl, name, valueClass) : assignment of an object of
##     class "character" is not valid for @'data' in an object of class
##     "categorical"; is(value, "integer") is not TRUE

11.5.3. Defining methods

For the S4 counterparts of the S3 generics (Section 10.2), see help("setGeneric"). Luckily, there is an reasonable degree of interoperability between the S3 and S4 systems.

Let us start by introducing a new method for the well-known as.character generic. Instead of defining as.character.categorical, we need to register a new routine with setMethod.

setMethod(
    "as.character",     # name of the generic
    "categorical",      # class of 1st arg; or: signature=c(x="categorical")
    function(x, ...)    # method definition
        x@levels[x@data]
)

Testing:

as.character(z)
## [1] "A" "B" "B" "A" "A"

show is the S4 counterpart of print:

setMethod(
    "show",
    "categorical",
    function(object)
    {
        x_character <- as.character(object)
        print(x_character)  # calls `print.default`
        cat(sprintf("Categories: %s\n",
            paste(object@levels, collapse=", ")))
    }
)

Interestingly, it is involved automatically on a call to print:

print(z)  # calls `show` for `categorical`
## [1] "A" "B" "B" "A" "A"
## Categories: A, B

Methods that dispatch on the type of multiple arguments are also possible, for example:

setMethod(
    "split",
    c(x="ANY", f="categorical"),
    function (x, f, drop=FALSE, ...)
        split(x, as.character(f), drop=drop, ...)
)

allows the first argument to be of any type (like a default method), and:

setMethod(
    "split",
    c(x="matrix", f="categorical"),
    function (x, f, drop=FALSE, ...)
        lapply(
            split(seq_len(NROW(x)), f, drop=drop, ...),  # calls the above
            function(i) x[i, , drop=FALSE])
)

is a version tailored for matrices. Testing:

A <- matrix(1:35, nrow=5)  # whatever
split(A, z)    # matrix,categorical
## $A
##      [,1] [,2] [,3] [,4] [,5] [,6] [,7]
## [1,]    1    6   11   16   21   26   31
## [2,]    4    9   14   19   24   29   34
## [3,]    5   10   15   20   25   30   35
##
## $B
##      [,1] [,2] [,3] [,4] [,5] [,6] [,7]
## [1,]    2    7   12   17   22   27   32
## [2,]    3    8   13   18   23   28   33
split(1:5, z)  # ANY,categorical
## $A
## [1] 1 4 5
##
## $B
## [1] 2 3

Exercise 11.26

Overload the `[` operator for the categorical class.

11.5.4. Defining constructors

We can also overload the initialize method, which is automatically called by new:

setMethod(
    "initialize",   # class name
    "categorical",  # method name
    function(.Object, x)
    {  # the method itself
        x <- as.character(x)  # see above
        xu <- unique(sort(x))  # drops NAs

        .Object@data <- match(x, xu)
        .Object@levels <- xu

        .Object  # return value - a modified object
    }
)

This allows for constructing new objects of the class categorical based on an object like x above, for instance:

w <- new("categorical", c("a", "c", "a", "a", "d", "c"))
print(w)
## [1] "a" "c" "a" "a" "d" "c"
## Categories: a, c, d

We have not set the two slots directly. They were automatically taken care of by initialize (note the American spelling).

Exercise 11.27

Set up a validating method for our class; see help("setValidity").

11.5.5. Inheritance

New S4 classes can be derived from existing ones, for instance:

setClass("binary", contains="categorical")

is a child class inhering all slots from its parent. We can, for example, overload the initialisation method for it:

setMethod(
    "initialize",
    "binary",
    function(.Object, x)
    {
        x <- as.character(as.integer(as.logical(x)))
        xu <- c("0", "1")
        .Object@data <- match(x, xu)
        .Object@levels <- xu
        .Object
    }
)

Testing:

new("binary", c(TRUE, FALSE, TRUE, FALSE, NA, TRUE))
## [1] "1" "0" "1" "0" NA  "1"
## Categories: 0, 1

We still used the show method of the parent class.

11.5.6. A note on the Matrix package

The Matrix package is perhaps the most widely known showcase of the S4 object orientation (and that is the reason why we cover S4 in this very chapter). It defines classes and methods for dense and sparse matrices, including rectangular, symmetric, triangular, band, and diagonal ones.

For instance, large graph (e.g., in network sciences) or preference (e.g., in recommender systems) data can be represented using sparse matrices (those which feature many 0s; after all, it is extremely more common for two vertices in a network not to be joined by an edge than to be connected).

For example:

library("Matrix")
(A <- Diagonal(x=1:5))
## 5 x 5 diagonal matrix of class "ddiMatrix"
##      [,1] [,2] [,3] [,4] [,5]
## [1,]    1    .    .    .    .
## [2,]    .    2    .    .    .
## [3,]    .    .    3    .    .
## [4,]    .    .    .    4    .
## [5,]    .    .    .    .    5

It created a real diagonal matrix. Moreover:

B <- as(A, "sparseMatrix")
B[1, 2] <- 7
B[4, 1] <- 42
print(B)
## 5 x 5 sparse Matrix of class "dgCMatrix"
##
## [1,]  1 7 . . .
## [2,]  . 2 . . .
## [3,]  . . 3 . .
## [4,] 42 . . 4 .
## [5,]  . . . . 5

yields a general sparse real matrix in the CRC (compressed, sparse, column-oriented) format.

For more information on the package, see vignette(package="Matrix").

11.6. Exercises

Exercise 11.28

Let X be a matrix with dimnames set, e.g.:

X <- matrix(1:12, byrow=TRUE, nrow=3)      # example matrix
dimnames(X)[[2]] <- c("a", "b", "c", "d")  # set column names
print(X)
##      a  b  c  d
## [1,] 1  2  3  4
## [2,] 5  6  7  8
## [3,] 9 10 11 12

Explain (in your own words) the meaning of the following expressions involving matrix subsetting. Note that not each of them is valid.

• X[1, ],

• X[, 3],

• X[, 3, drop=FALSE],

• X[3],

• X[, "a"],

• X[, c("a", "b", "c")],

• X[, -2],

• X[X[,1] > 5, ],

• X[X[,1]>5, c("a", "b", "c")],

• X[X[,1]>=5 & X[,1]<=10, ],

• X[X[,1]>=5 & X[,1]<=10, c("a", "b", "c")],

• X[, c(1, "b", "d")].

Exercise 11.29

Assuming that X is an array, what are the differences between the following indexing schemes?

• X["1", ] vs X[1, ],

• X[, "a", "b", "c"] vs X["a", "b", "c"] vs X[, c("a", "b", "c")] vs X[c("a", "b", "c")],

• X[1] vs X[, 1] vs X[1, ],

• X[X>0] vs X[X>0, ] vs X[, X>0],

• X[X[, 1]>0] vs X[X[, 1]>0,] vs X[,X[,1]>0],

• X[X[, 1]>5, X[1, ]<10] vs X[X[1, ]>5, X[, 1]<10].

Exercise 11.30

Give a few ways to create a matrix like:

##      [,1] [,2]
## [1,]    1    1
## [2,]    1    2
## [3,]    1    3
## [4,]    2    1
## [5,]    2    2
## [6,]    2    3

and one like:

##       [,1] [,2] [,3]
##  [1,]    1    1    1
##  [2,]    1    1    2
##  [3,]    1    2    1
##  [4,]    1    2    2
##  [5,]    1    3    1
##  [6,]    1    3    2
##  [7,]    2    1    1
##  [8,]    2    1    2
##  [9,]    2    2    1
## [10,]    2    2    2
## [11,]    2    3    1
## [12,]    2    3    2

Exercise 11.31

For a given real \(n\times m\) matrix \(\mathbf{X}\), determine the bounding hyperrectangle of thusly encoded \(n\) input points in an \(m\)-dimensional space. Return a \(2\times m\) matrix \(\mathbf{B}\) with \(b_{1,j}=\min_i x_{i,j}\) and \(b_{2,j}=\max_i x_{i,j}\).

Exercise 11.32

Let \(\mathbf{t}\) be a vector of \(n\) integers in \(\{1,\dots,k\}\). Write a function to one-hot-encode each \(t_i\): return a 0–1 matrix \(\mathbf{R}\) of size \(n\times k\) such that \(r_{i,j}=1\) if and only if \(j = t_i\). For example, if \(\mathbf{t}=[1, 2, 3, 2, 4]\) and \(k=4\), then:

\[ \mathbf{R} = \left[ \begin{array}{cccc} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 1 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 1 \\ \end{array} \right]. \]

On a side note, such a representation is beneficial when solving, e.g., a multiclass classification problem by means of \(k\) binary classifiers.

Then, compose another function, but this time setting \(r_{i,j}=1\) if and only if \(j\ge t_i\), e.g.:

\[ R = \left[ \begin{array}{cccc} 1 & 1 & 1 & 1 \\ 0 & 1 & 1 & 1 \\ 0 & 0 & 1 & 1 \\ 0 & 1 & 1 & 1 \\ 0 & 0 & 0 & 1 \\ \end{array} \right]. \]

Important

Kind reminder: as usual, try to solve all the exercises without using explicit for and while loops (provided that it is possible).

Exercise 11.33

Given an \(n\times k\) real matrix, apply the softmax function on each row, i.e., map \(x_{i,j}\) to \(\frac{\exp(x_{i,j})}{\sum_{l=1}^k \exp(x_{i,l})}\). Then, one-hot decode the values in each row, i.e., find the column number with the greatest value. Return a vector of size \(n\) with elements in \(\{1,\dots,k\}\).

Exercise 11.34

Assume that an \(n\times m\) real matrix \(\mathbf{X}\) represents \(n\) points in \(\mathbb{R}^m\). Write a function (but do not refer to dist) that determines the pairwise distances between all the \(n\) points and a given \(\mathbf{y}\in\mathbb{R}^m\). Return a vector \(\mathbf{d}\) of length \(n\) with \(d_{i}=\|\mathbf{x}_{i,\cdot}-\mathbf{y}\|_2\).

Exercise 11.35

Let \(\mathbf{X}\) and \(\mathbf{Y}\) be two real-valued matrices of sizes \(n\times m\) and \(k\times m\), respectively, representing two sets of points in \(\mathbb{R}^m\). Return an integer vector \(\mathbf{r}\) of length \(k\) such that \(r_i\) indicates the index of the point in \(\mathbf{X}\) with the least distance to (the closest to) the \(i\)-th point in \(\mathbf{Y}\), i.e., \(r_i = \mathrm{arg}\min_j \|\mathbf{x}_{j,\cdot}-\mathbf{y}_{i,\cdot}\|_2\).

Exercise 11.36

Write your version of the built-in utils::combn.

Exercise 11.37

Time series are vectors or matrices of the class ts equipped with the tsp attribute, amongst others. Refer to help("ts") for more information about how they are represented and what S3 methods have been overloaded for them.

Exercise 11.38

(*) Numeric matrices can be stored in a CSV file, amongst others. Usually, we will be loading them via read.csv, which returns a data frame (see Chapter 12), for example:

X <- as.matrix(read.csv(
    paste0(
        "https://github.com/gagolews/teaching-data/",
        "raw/master/marek/eurxxx-20200101-20200630.csv"
    ),
    comment.char="#",
    sep=","
))

Write a function read_numeric_matrix(file_name, comment, sep) which is based on a few calls to scan instead. Use file to establish a file connection so that you can ignore the comment lines and fetch the column names before reading the actual numeric values.

Exercise 11.39

(*) Using readBin, read the t10k-images-idx3-ubyte.gz from the MNIST database homepage. The output object should be a three-dimensional, \(10000\times 28\times 28\) array with real elements between 0 and 255. Refer to the File Formats section therein for more details.

Exercise 11.40

(**) Circular convolution of discrete-valued multidimensional signals can be performed by means of fft and matrix multiplication, whereas affine transformations require only the latter. Apply various image transformations such as sharpening, shearing, and rotating on the MNIST digits and plot the results using the image function.

Exercise 11.41

(*) Using constrOptim, find the minimum of the Constrained Betts Function \(f(x_1, x_2) = 0.01 x_1^2 + x_2^2 - 100\) with linear constraints \( 2\le x_1 \le 50\), \(-50 \le x_2 \le 50\), and \(10 x_1 \ge 10 + x_2\). (**) Also, use solve.QP from the quadprog package to find the minimum.

---

[1]

Which can easily be explained by the fact that matrix elements are stored in a columnwise order.

[2]

Calling unclass on a matrix has no effect.

[3]

Hidden deeply at the C language level.

[4]

They are simply applied on each element of the underlying flat vector. Section 5.5 mentioned that unary functions preserve all attributes of their inputs, hence also dim and dimnames.

[5]

(*) We can select the underlying implementation of BLAS at R’s compile time; see Section A.3 in [64]. Some of them are faster than others.

[6]

For drawing random samples from any multivariate distribution, refer to the theory of copulas, e.g., [48]. There are a few R packages on CRAN that implement the most popular models.

[7]

\(\mathbf{A}\) can also be a complex matrix,. In this case, its QR decomposition spawns \(\mathbf{Q}\) that is a unitary matrix.

[8]

As the pseudoinverse \(\mathbf{X}^+=\left(\mathbf{X}^T \mathbf{X} \right)^{-1} \mathbf{X}^T =\mathbf{V} \mathbf{D}^+ \mathbf{U}^T=\mathbf{R}^{-1}\mathbf{Q}^T\), with \(\mathbf{X}^+ \mathbf{X}=\mathbf{I}\). Here \(\mathbf{D}^+\) is a transposed version of \(\mathbf{D}\) featuring the reciprocals of its non-zero elements.

[9]

See the source code of getS3method("prcomp", "default").

[10]

A partial solution to this could involve defining a method like validate.class_name, which is called frequently and which checks whether a given object enjoys some desired constraints.

[11]

Although there are functions featuring some workarounds (see, e.g., cbind, which dispatches to cbind.data.frame if one argument is a data frame and the remaining ones are vectors or matrices). Also, we said in the previous chapter that binary operators consider the classes of both operands.

[12]

Hypothetically, we can imagine an OOP system relying on methods named like method.class_name1.class_name2 where dispatching is based on two argument types. It would be beautiful, but it is not the case in R.