Linear Algebra Notes

David A. SANTOS
[email protected]

July 2, 2006 REVISION

ii

Contents
Preface iii 4.5 Vectors in R3 . . . . . . . . . . . . . . . 86
4.6 Planes and Lines in R3 . . . . . . . . . . 91
To the Student v 4.7 Rn . . . . . . . . . . . . . . . . . . . . 95
1 Preliminaries 1 5 Vector Spaces 100
1.1 Sets and Notation . . . . . . . . . . . . . 1 5.1 Vector Spaces . . . . . . . . . . . . . . . 100
1.2 Partitions and Equivalence Relations . . . 4 5.2 Vector Subspaces . . . . . . . . . . . . . 103
1.3 Binary Operations . . . . . . . . . . . . . 6 5.3 Linear Independence . . . . . . . . . . . 105
1.4 Zn . . . . . . . . . . . . . . . . . . . . . 9 5.4 Spanning Sets . . . . . . . . . . . . . . . 107
1.5 Fields . . . . . . . . . . . . . . . . . . . 12 5.5 Bases . . . . . . . . . . . . . . . . . . . 110
1.6 Functions . . . . . . . . . . . . . . . . . 14 5.6 Coordinates . . . . . . . . . . . . . . . . 114
2 Matrices and Matrix Operations 17 6 Linear Transformations 120
2.1 The Algebra of Matrices . . . . . . . . . . 17 6.1 Linear Transformations . . . . . . . . . . 120
2.2 Matrix Multiplication . . . . . . . . . . . 20 6.2 Kernel and Image of a Linear Transformation122
2.3 Trace and Transpose . . . . . . . . . . . 24 6.3 Matrix Representation . . . . . . . . . . 127
2.4 Special Matrices . . . . . . . . . . . . . . 26
2.5 Matrix Inversion . . . . . . . . . . . . . 32 7 Determinants 133
2.6 Block Matrices . . . . . . . . . . . . . . 40 7.1 Permutations . . . . . . . . . . . . . . . 133
2.7 Rank of a Matrix . . . . . . . . . . . . . 41 7.2 Cycle Notation . . . . . . . . . . . . . . 136
2.8 Rank and Invertibility . . . . . . . . . . . 50 7.3 Determinants . . . . . . . . . . . . . . . 141
7.4 Laplace Expansion . . . . . . . . . . . . 150
3 Linear Equations 57 7.5 Determinants and Linear Systems . . . . . 156
3.1 Definitions . . . . . . . . . . . . . . . . 57
3.2 Existence of Solutions . . . . . . . . . . . 61 8 Eigenvalues and Eigenvectors 157
3.3 Examples of Linear Systems . . . . . . . . 63 8.1 Similar Matrices . . . . . . . . . . . . . . 157
8.2 Eigenvalues and Eigenvectors . . . . . . . 158
4 R2 , R3 and Rn 67 8.3 Diagonalisability . . . . . . . . . . . . . 162
4.1 Points and Bi-points in R2 . . . . . . . . 67 8.4 The Minimal Polynomial . . . . . . . . . 165
4.2 Vectors in R2 . . . . . . . . . . . . . . . 70
4.3 Dot Product in R2 . . . . . . . . . . . . 76 A Answers and Hints 167
4.4 Lines on the Plane . . . . . . . . . . . . 81 Answers and Hints . . . . . . . . . . . . . . . 167
Preface

These notes started during the Spring of 2002, when John MAJEWICZ and I each taught a section of Linear Algebra.
I would like to thank him for numerous suggestions on the written notes.

The students of my class were: Craig BARIBAULT, Chun CAO, Jacky CHAN, Pho DO, Keith HARMON,
Nicholas SELVAGGI, Sanda SHWE, and Huong VU.

John’s students were David HERNÁNDEZ, Adel JAILILI, Andrew KIM, Jong KIM, Abdelmounaim LAAYOUNI,
Aju MATHEW, Nikita MORIN, Thomas NEGRÓN, Latoya ROBINSON, and Saem SOEURN.

Linear Algebra is often a student’s first introduction to abstract mathematics. Linear Algebra is well suited for
this, as it has a number of beautiful but elementary and easy to prove theorems. My purpose with these notes is to
introduce students to the concept of proof in a gentle manner.

David A. Santos [email protected]

iii
Legal Notice

This material may be distributed only subject to the terms and conditions set forth in the Open Publication License,
version 1.0 or later (the latest version is presently available at
http://www.opencontent.org/openpub/.

THIS WORK IS LICENSED AND PROVIDED “AS IS” WITHOUT WARRANTY OF ANY KIND, EXPRESS
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
AND FITNESS FOR A PARTICULAR PURPOSE OR A WARRANTY OF NON-INFRINGEMENT.

THIS DOCUMENT MAY NOT BE SOLD FOR PROFIT OR INCORPORATED INTO COMMERCIAL DOC-
UMENTS WITHOUT EXPRESS PERMISSION FROM THE AUTHOR(S). THIS DOCUMENT MAY BE FREELY
DISTRIBUTED PROVIDED THE NAME OF THE ORIGINAL AUTHOR(S) IS(ARE) KEPT AND ANY CHANGES
TO IT NOTED.

iv
 v

To the Student
These notes are provided for your benefit as an attempt to organise the salient points of the course. They are a
very terse account of the main ideas of the course, and are to be used mostly to refer to central definitions and
theorems. The number of examples is minimal, and here you will find few exercises. The motivation or informal
ideas of looking at a certain topic, the ideas linking a topic with another, the worked-out examples, etc., are given
in class. Hence these notes are not a substitute to lectures: you must always attend to lectures. The order of
the notes may not necessarily be the order followed in the class.

There is a certain algebraic fluency that is necessary for a course at this level. These algebraic prerequisites
would be difficult to codify here, as they vary depending on class response and the topic lectured. If at any stage
you stumble in Algebra, seek help! I am here to help you!

Tutoring can sometimes help, but bear in mind that whoever tutors you may not be familiar with my conventions.
Again, I am here to help! On the same vein, other books may help, but the approach presented here is at times
unorthodox and finding alternative sources might be difficult.

Here are more recommendations:

 Read a section before class discussion, in particular, read the definitions.
 Class provides the informal discussion, and you will profit from the comments of your classmates, as well as
gain confidence by providing your insights and interpretations of a topic. Don’t be absent!
 Once the lecture of a particular topic has been given, take a fresh look at the notes of the lecture topic.
 Try to understand a single example well, rather than ill-digest multiple examples.
 Start working on the distributed homework ahead of time.
 Ask questions during the lecture. There are two main types of questions that you are likely to ask.
1. Questions of Correction: Is that a minus sign there? If you think that, for example, I have missed
out a minus sign or wrote P where it should have been Q,1 then by all means, ask. No one likes to carry
an error till line XLV because the audience failed to point out an error on line I. Don’t wait till the end
of the class to point out an error. Do it when there is still time to correct it!
2. Questions of Understanding: I don’t get it! Admitting that you do not understand something is an act
requiring utmost courage. But if you don’t, it is likely that many others in the audience also don’t. On
the same vein, if you feel you can explain a point to an inquiring classmate, I will allow you time in the
lecture to do so. The best way to ask a question is something like: “How did you get from the second
step to the third step?” or “What does it mean to complete the square?” Asseverations like “I don’t
understand” do not help me answer your queries. If I consider that you are asking the same questions
too many times, it may be that you need extra help, in which case we will settle what to do outside the
lecture.
 Don’t fall behind! The sequence of topics is closely interrelated, with one topic leading to another.
 The use of calculators is allowed, especially in the occasional lengthy calculations. However, when graphing, you
will need to provide algebraic/analytic/geometric support of your arguments. The questions on assignments
and exams will be posed in such a way that it will be of no advantage to have a graphing calculator.
 Presentation is critical. Clearly outline your ideas. When writing solutions, outline major steps and write in
complete sentences. As a guide, you may try to emulate the style presented in the scant examples furnished
in these notes.

1 My doctoral adviser used to say “I said A, I wrote B, I meant C and it should have been D!
Chapter 1
Preliminaries

1.1 Sets and Notation

1 Definition We will mean by a set a collection of well defined members or elements.

2 Definition The following sets have special symbols.

N = {0, 1, 2, 3, . . .} denotes the set of natural numbers.

Z = {. . . , −3, −2, −1, 0, 1, 2, 3, . . .} denotes the set of integers.

Q denotes the set of rational numbers.

R denotes the set of real numbers.

C denotes the set of complex numbers.

∅ denotes the empty set.

3 Definition (Implications) The symbol =⇒ is read “implies”, and the symbol ⇐⇒ is read “if and only if.”

4 Example Prove that between any two rational numbers there is always a rational number.

a c
Solution: Let (a, c) ∈ Z2 , (b, d) ∈ (N \ {0})2 , b
< d
. Then da < bc. Now

a a+c
ab + ad < ab + bc =⇒ a(b + d) < b(a + c) =⇒ < ,
b b+d
a+c c
da + dc < cb + cd =⇒ d(a + c) < c(b + d) =⇒ < ,
b+d d
a+c a c
whence the rational number lies between b
and d
.
b+d

☞ We can also argue that the average

r +r
of two distinct numbers lies between the numbers and so
1 2
if r1 < r2 are rational numbers, then lies between them.
2

5 Definition Let A be a set. If a belongs to the set A, then we write a ∈ A, read “a is an element of A.” If a does
not belong to the set A, we write a 6∈ A, read “a is not an element of A.”

1
2 Chapter 1

6 Definition (Conjunction, Disjunction, and Negation) The symbol ∨ is read “or” (disjunction), the symbol ∧ is
read “and” (conjunction), and the symbol ¬ is read “not.”

7 Definition (Quantifiers) The symbol ∀ is read “for all” (the universal quantifier), and the symbol ∃ is read “there
exists” (the existential quantifier).

We have
¬(∀x ∈ A, P(x)) ⇐⇒ (∃ ∈ A, ¬P(x)) (1.1)
¬(∃ ∈ A, P(x)) ⇐⇒ (∀x ∈ A, ¬P(x)) (1.2)

8 Definition (Subset) If ∀a ∈ A we have a ∈ B, then we write A ⊆ B, which we read “A is a subset of B.”

In particular, notice that for any set A, ∅ ⊆ A and A ⊆ A. Also

N ⊆ Z ⊆ Q ⊆ R ⊆ C.

☞A=B ⇐⇒ (A ⊆ B) ∧ (B ⊆ A).

9 Definition The union of two sets A and B, is the set

A ∪ B = {x : (x ∈ A) ∨ (x ∈ B)}.

This is read “A union B.” See figure 1.1.

10 Definition The intersection of two sets A and B, is

A ∩ B = {x : (x ∈ A) ∧ (x ∈ B)}.

This is read “A intersection B.” See figure 1.2.

11 Definition The difference of two sets A and B, is

A \ B = {x : (x ∈ A) ∧ (x 6∈ B)}.

This is read “A set minus B.” See figure 1.3.

A B A B A B

Figure 1.1: A ∪ B Figure 1.2: A ∩ B Figure 1.3: A \ B

12 Example Prove by means of set inclusion that

(A ∪ B) ∩ C = (A ∩ C) ∪ (B ∩ C).
Sets and Notation 3

Solution: We have,

x ∈ (A ∪ B) ∩ C ⇐⇒ x ∈ (A ∪ B) ∧ x ∈ C

⇐⇒ (x ∈ A ∨ x ∈ B) ∧ x ∈ C

⇐⇒ (x ∈ A ∧ x ∈ C) ∨ (x ∈ B ∧ x ∈ C)

⇐⇒ (x ∈ A ∩ C) ∨ (x ∈ B ∩ C)

⇐⇒ x ∈ (A ∩ C) ∪ (B ∩ C),

which establishes the equality.

13 Definition Let A1 , A2 , . . . , An , be sets. The Cartesian Product of these n sets is defined and denoted by

A1 × A2 × · · · × An = {(a1 , a2 , . . . , an ) : ak ∈ Ak },

that is, the set of all ordered n-tuples whose elements belong to the given sets.

☞ In the particular case when all the A k are equal to a set A, we write

A1 × A2 × · · · × An = An .

If a ∈ A and b ∈ A we write (a, b) ∈ A2 .

14 Definition Let x ∈ R. The absolute value of x—denoted by |x|—is defined by




 −x if x < 0,
|x| =


 x if x ≥ 0.

It follows from the definition that for x ∈ R,

− |x| ≤ x ≤ |x|. (1.3)
t ≥ 0 =⇒ |x| ≤ t ⇐⇒ −t ≤ x ≤ t. (1.4)
√
∀a ∈ R =⇒ a2 = |a|. (1.5)

15 Theorem (Triangle Inequality) Let (a, b) ∈ R2 . Then

|a + b| ≤ |a| + |b|. (1.6)

Proof: From 1.3, by addition,

−|a| ≤ a ≤ |a|
to
−|b| ≤ b ≤ |b|
we obtain
−(|a| + |b|) ≤ a + b ≤ (|a| + |b|),
whence the theorem follows by 1.4. ❑
4 Chapter 1

16 Problem Prove that between any two rational numbers 20 Problem Prove that (A ∪ B) \ (A ∩ B) = (A \ B) ∪ (B \ A).
there is an irrational number.
21 Problem Shew how to write the union A ∪ B ∪ C as a
17 Problem Prove that X \ (X \ A) = X ∩ A. disjoint union of sets.

22 Problem Prove that a set with n ≥ 0 elements has 2n

18 Problem Prove that X \ (A ∪ B) = (X \ A) ∩ (X \ B).
subsets.

19 Problem Prove that X \ (A ∩ B) = (X \ A) ∪ (X \ B). 23 Problem Let (a, b) ∈ R2 . Prove that ||a| − |b|| ≤ |a − b|.

1.2 Partitions and Equivalence Relations

24 Definition Let S 6= ∅ be a set. A partition of S is a collection of non-empty, pairwise disjoint subsets of S whose
union is S.

25 Example Let
2Z = {. . . , −6, −4, −2, 0, 2, 4, 6, . . .} = 0
be the set of even integers and let

2Z + 1 = {. . . , −5, −3, −1, 1, 3, 5, . . .} = 1

be the set of odd integers. Then

(2Z) ∪ (2Z + 1) = Z, (2Z) ∩ (2Z + 1) = ∅,

and so {2Z, 2Z + 1} is a partition of Z.

26 Example Let
3Z = {. . . − 9, , −6, −3, 0, 3, 6, 9, . . .} = 0
be the integral multiples of 3, let

3Z + 1 = {. . . , −8, −5, −2, 1, 4, 7, . . .} = 1

be the integers leaving remainder 1 upon division by 3, and let

3Z + 2 = {. . . , −7, −4, −1, 2, 5, 8, . . .} = 2

be integers leaving remainder 2 upon division by 3. Then

(3Z) ∪ (3Z + 1) ∪ (3Z + 2) = Z,

(3Z) ∩ (3Z + 1) = ∅, (3Z) ∩ (3Z + 2) = ∅, (3Z + 1) ∩ (3Z + 2) = ∅,
and so {3Z, 3Z + 1, 3Z + 2} is a partition of Z.

☞ Notice that 0 and 1 do not mean the same in examples 25 and 26. Whenever we make use
of this notation, the integral divisor must be made explicit.

27 Example Observe
R = (Q) ∪ (R \ Q), ∅ = (Q) ∩ (R \ Q),
which means that the real numbers can be partitioned into the rational and irrational numbers.

28 Definition Let A, B be sets. A relation R is a subset of the Cartesian product A × B. We write the fact that
(x, y) ∈ R as x ∼ y.
Partitions and Equivalence Relations 5

29 Definition Let A be a set and R be a relation on A × A. Then R is said to be

 reflexive if (∀x ∈ A), x ∼ x,
 symmetric if (∀(x, y) ∈ A2 ), x ∼ y =⇒ y ∼ x,
 anti-symmetric if (∀(x, y) ∈ A2 ), (x ∼ y) ∧ (y ∼ x) =⇒ x = y,
 transitive if (∀(x, y, z) ∈ A3 ), (x ∼ y) ∧ (y ∼ z) =⇒ (x ∼ z).
A relation R which is reflexive, symmetric and transitive is called an equivalence relation on A. A relation R which
is reflexive, anti-symmetric and transitive is called a partial order on A.

30 Example Let S ={All Human Beings}, and define ∼ on S as a ∼ b if and only if a and b have the same
mother. Then a ∼ a since any human a has the same mother as himself. Similarly, a ∼ b =⇒ b ∼ a and
(a ∼ b) ∧ (b ∼ c) =⇒ (a ∼ c). Therefore ∼ is an equivalence relation.

31 Example Let L be the set of all lines on the plane and write l1 ∼ l2 if l1 ||l2 (the line l1 is parallel to the line
l2 ). Then ∼ is an equivalence relation on L.

a x
32 Example In Q define the relation b ∼ y ⇐⇒ ay = bx, where we will always assume that the denominators
are non-zero. Then ∼ is an equivalence relation. For a
b
∼ a
b
since ab = ab. Clearly
a x x a
∼ =⇒ ay = bx =⇒ xb = ya =⇒ ∼ .
b y y b
a x x s
Finally, if b ∼ y
and y
∼ t
then we have ay = bx and xt = sy. Multiplying these two equalities ayxt = bxsy.
This gives
ayxt − bxsy = 0 =⇒ xy(at − bs) = 0.
Now if x = 0, we will have a = s = 0, in which case trivially at = bs. Otherwise we must have at − bs = 0 and so
a
b
∼ st .

33 Example Let X be a collection of sets. Write A ∼ B if A ⊆ B. Then ∼ is a partial order on X.

34 Example For (a, b) ∈ R2 define

a ∼ b ⇔ a2 + b2 > 2.
Determine, with proof, whether ∼ is reflexive, symmetric, and/or transitive. Is ∼ an equivalence relation?

Solution: Since 02 + 02 ≯ 2, we have 0 ≁ 0 and so ∼ is not reflexive. Now,

a∼b ⇔ a2 + b2

⇔ b2 + a2

⇔ b ∼ a,

so ∼ is symmetric. Also 0 ∼ 3 since 02 + 32 > 2 and 3 ∼ 1 since 32 + 12 > 2. But 0 ≁ 1 since 02 + 12 ≯ 2. Thus
the relation is not transitive. The relation, therefore, is not an equivalence relation.

35 Definition Let ∼ be an equivalence relation on a set S. Then the equivalence class of a is defined and denoted
by
[a] = {x ∈ S : x ∼ a}.

36 Lemma Let ∼ be an equivalence relation on a set S. Then two equivalence classes are either identical or disjoint.
6 Chapter 1

Proof: We prove that if (a, b) ∈ S2 , and [a] ∩ [b] 6= ∅ then [a] = [b]. Suppose that x ∈ [a] ∩ [b].
Now x ∈ [a] =⇒ x ∼ a =⇒ a ∼ x, by symmetry. Similarly, x ∈ [b] =⇒ x ∼ b. By transitivity

(a ∼ x) ∧ (x ∼ b) =⇒ a ∼ b.

Now, if y ∈ [b] then b ∼ y. Again by transitivity, a ∼ y. This means that y ∈ [a]. We have shewn
that y ∈ [b] =⇒ y ∈ [a] and so [b] ⊆ [a]. In a similar fashion, we may prove that [a] ⊆ [b]. This
establishes the result. ❑

37 Theorem Let S 6= ∅ be a set. Any equivalence relation on S induces a partition of S. Conversely, given a
partition of S into disjoint, non-empty subsets, we can define an equivalence relation on S whose equivalence classes
are precisely these subsets.

Proof: By Lemma 36, if ∼ is an equivalence relation on S then

[
S= [a],
a∈S

and [a] ∩ [b] = ∅ if a ≁ b. This proves the first half of the theorem.

Conversely, let [
S= Sα , Sα ∩ Sβ = ∅ if α 6= β,
α

be a partition of S. We define the relation ≈ on S by letting a ≈ b if and only if they belong to

the same Sα . Since the Sα are mutually disjoint, it is clear that ≈ is an equivalence relation on
S and that for a ∈ Sα , we have [a] = Sα . ❑

38 Problem For (a, b) ∈ (Q \ {0})2 define the relation ∼ as 40 Problem Define the relation ∼ in R by x ∼ y ⇐⇒ xey =
follows: a ∼ b ⇔ ba
∈ Z. Determine whether this relation is yex . Prove that ∼ is an equivalence relation.
reflexive, symmetric, and/or transitive.
41 Problem Define the relation ∼ in Q by x ∼ y ⇐⇒ ∃h ∈ Z
39 Problem Give an example of a relation on Z \ {0} which is 3y + h
such that x = . [A] Prove that ∼ is an equivalence re-
reflexive, but is neither symmetric nor transitive. 3
lation. [B] Determine [x], the equivalence of x ∈ Q. [C] Is
2
3
∼ 45 ?

1.3 Binary Operations

42 Definition Let S, T be sets. A binary operation is a function

S×S → T
⊗: .
(a, b) 7→ (a, b)

We usually use the “infix” notation a ⊗ b rather than the “prefix” notation ⊗(a, b). If S = T then we say that the
binary operation is internal or closed and if S 6= T then we say that it is external. If

a⊗b =b⊗a

then we say that the operation ⊗ is commutative and if

a ⊗ (b ⊗ c) = (a ⊗ b) ⊗ c,

we say that it is associative. If ⊗ is associative, then we can write

a ⊗ (b ⊗ c) = (a ⊗ b) ⊗ c = a ⊗ b ⊗ c,

without ambiguity.
Binary Operations 7

☞We usually omit the sign ⊗ and use juxtaposition to indicate the operation ⊗. Thus we write
ab instead of a ⊗ b.

43 Example The operation + (ordinary addition) on the set Z × Z is a commutative and associative closed binary
operation.

44 Example The operation − (ordinary subtraction) on the set N × N is a non-commutative, non-associative non-
closed binary operation.

45 Example The operation ⊗ defined by a ⊗ b = 1 + ab on the set Z × Z is a commutative but non-associative

internal binary operation. For
a ⊗ b = 1 + ab = 1 + ba = ba,
proving commutativity. Also, 1 ⊗ (2 ⊗ 3) = 1 ⊗ (7) = 8 and (1 ⊗ 2) ⊗ 3 = (3) ⊗ 3 = 10, evincing non-associativity.

46 Definition Let S be a set and ⊗ : S × S → S be a closed binary operation. The couple hS, ⊗i is called an
algebra.

☞ When we desire to drop the sign ⊗ and indicate the binary operation by juxtaposition, we
simply speak of the “algebra S.”

47 Example Both hZ, +i and hQ, ·i are algebras. Here + is the standard addition of real numbers and · is the
standard multiplication.

48 Example hZ, −i is a non-commutative, non-associative algebra. Here − is the standard subtraction operation
on the real numbers

49 Example (Putnam Exam, 1972) Let S be a set and let ∗ be a binary operation of S satisfying the laws ∀(x, y) ∈
S2
x ∗ (x ∗ y) = y, (1.7)
(y ∗ x) ∗ x = y. (1.8)
Shew that ∗ is commutative, but not necessarily associative.

Solution: By (1.8)
x ∗ y = ((x ∗ y) ∗ x) ∗ x.
By (1.8) again
((x ∗ y) ∗ x) ∗ x = ((x ∗ y) ∗ ((x ∗ y) ∗ y)) ∗ x.
By (1.7)
((x ∗ y) ∗ ((x ∗ y) ∗ y)) ∗ x = (y) ∗ x = y ∗ x,
which is what we wanted to prove.

To shew that the operation is not necessarily associative, specialise S = Z and x ∗ y = −x − y (the opposite of
x minus y). Then clearly in this case ∗ is commutative, and satisfies (1.7) and (1.8) but

0 ∗ (0 ∗ 1) = 0 ∗ (−0 − 1) = 0 ∗ (−1) = −0 − (−1) = 1,

and
(0 ∗ 0) ∗ 1 = (−0 − 0) ∗ 1 = (0) ∗ 1 = −0 − 1 = −1,
evincing that the operation is not associative.

50 Definition Let S be an algebra. Then l ∈ S is called a left identity if ∀s ∈ S we have ls = s. Similarly r ∈ S

is called a right identity if ∀s ∈ S we have sr = s.
8 Chapter 1

51 Theorem If an algebra S possesses a left identity l and a right identity r then l = r.

Proof: Since l is a left identity

r = lr.
Since r is a right identity
l = lr.
Combining these two, we gather
r = lr = l,
whence the theorem follows. ❑

52 Example In hZ, +i the element 0 ∈ Z acts as an identity, and in hQ, ·i the element 1 ∈ Q acts as an identity.

53 Definition Let S be an algebra. An element a ∈ S is said to be left-cancellable or left-regular if ∀(x, y) ∈ S2

ax = ay =⇒ x = y.

Similarly, element b ∈ S is said to be right-cancellable or right-regular if ∀(x, y) ∈ S2

xb = yb =⇒ x = y.

Finally, we say an element c ∈ S is cancellable or regular if it is both left and right cancellable.

54 Definition Let hS, ⊗i and hS, ⊤i be algebras. We say that ⊤ is left-distributive with respect to ⊗ if

∀(x, y, z) ∈ S3 , x⊤(y ⊗ z) = (x⊤y) ⊗ (x⊤z).

Similarly, we say that ⊤ is right-distributive with respect to ⊗ if

∀(x, y, z) ∈ S3 , (y ⊗ z)⊤x = (y⊤x) ⊗ (z⊤x).

We say that ⊤ is distributive with respect to ⊗ if it is both left and right distributive with respect to ⊗.

55 Problem Let ➌ Find an element e ∈ R such that (∀a ∈ Q∩] −

1; 1[) (e ⊗ a = a).
S = {x ∈ Z : ∃(a, b) ∈ Z2 , x = a3 + b3 + c3 − 3abc}.

Prove that S is closed under multiplication, that is, if x ∈ S ➍ Given e as above and an arbitrary element a ∈
and y ∈ S then xy ∈ S. Q∩] − 1; 1[, solve the equation a ⊗ b = e for b.

56 Problem Let hS, ⊗i be an associative algebra, let a ∈ S

be a fixed element and define the closed binary operation ⊤ 58 Problem On R \ {1} define the a binary operation ⊗
by
a ⊗ b = a + b − ab,
x⊤y = x ⊗ a ⊗ y.
Prove that ⊤ is also associative over S × S. where juxtaposition means ordinary multiplication and + is
the ordinary addition of real numbers. Clearly ⊗ is a closed
binary operation. Prove that
57 Problem On Q∩] − 1; 1[ define the a binary operation ⊗
➊ Prove that ⊗ is both commutative and associative.
a+b
a⊗b= ,
1 + ab
➋ Find an element e ∈ R \ {1} such that (∀a ∈
where juxtaposition means ordinary multiplication and + is R \ {1}) (e ⊗ a = a).
the ordinary addition of real numbers. Prove that
➊ Prove that ⊗ is a closed binary operation on Q∩]−1; 1[. ➌ Given e as above and an arbitrary element a ∈ R \ {1},
solve the equation a ⊗ b = e for b.
➋ Prove that ⊗ is both commutative and associative.
Zn 9

59 Problem (Putnam Exam, 1971) Let S be a set and let Shew that ◦ is commutative.
◦ be a binary operation on S satisfying the two laws
(∀x ∈ S)(x ◦ x = x),
60 Problem Define the symmetric difference of the sets A, B
and
as A△B = (A \ B) ∪ (B \ A). Prove that △ is commutative
(∀(x, y, z) ∈ S3 )((x ◦ y) ◦ z = (y ◦ z) ◦ x). and associative.

1.4 Zn
61 Theorem (Division Algorithm) Let n > 0 be an integer. Then for any integer a there exist unique integers q
(called the quotient) and r (called the remainder) such that a = qn + r and 0 ≤ r < q.

Proof: In the proof of this theorem, we use the following property of the integers, called the
well-ordering principle: any non-empty set of non-negative integers has a smallest element.

Consider the set

S = {a − bn : b ∈ Z ∧ a ≥ bn}.
Then S is a collection of nonnegative integers and S 6= ∅ as ±a − 0 · n ∈ S and this is non-negative
for one choice of sign. By the Well-Ordering Principle, S has a least element, say r. Now, there
must be some q ∈ Z such that r = a − qn since r ∈ S. By construction, r ≥ 0. Let us prove
that r < n. For assume that r ≥ n. Then r > r − n = a − qn − n = a − (q + 1)n ≥ 0, since
r − n ≥ 0. But then a − (q + 1)n ∈ S and a − (q + 1)n < r which contradicts the fact that r is the
smallest member of S. Thus we must have 0 ≤ r < n. To prove that r and q are unique, assume
that q1 n + r1 = a = q2 n + r2 , 0 ≤ r1 < n, 0 ≤ r2 < n. Then r2 − r1 = n(q1 − q2 ), that is, n
divides (r2 − r1 ). But |r2 − r1 | < n, whence r2 = r1 . From this it also follows that q1 = q2 . This
completes the proof. ❑

62 Example If n = 5 the Division Algorithm says that we can arrange all the integers in five columns as follows:

.. .. .. .. ..
. . . . .

−10 −9 −8 −7 −6

−5 −4 −3 −2 −1

0 1 2 3 4

5 6 7 8 9
.. .. .. .. ..
. . . . .

The arrangement above shews that any integer comes in one of 5 flavours: those leaving remainder 0 upon division
by 5, those leaving remainder 1 upon division by 5, etc. We let

5Z = {. . . , −15, −10, −5, 0, 5, 10, 15, . . .} = 0,

5Z + 1 = {. . . , −14, −9, −4, 1, 6, 11, 16, . . .} = 1,

5Z + 2 = {. . . , −13, −8, −3, 2, 7, 12, 17, . . .} = 2,
5Z + 3 = {. . . , −12, −7, −2, 3, 8, 13, 18, . . .} = 3,
5Z + 4 = {. . . , −11, −6, −1, 4, 9, 14, 19, . . .} = 4,
and
Z5 = {0, 1, 2, 3, 4}.
10 Chapter 1

Let n be a fixed positive integer. Define the relation ≡ by x ≡ y if and only if they leave the same remainder upon
division by n. Then clearly ≡ is an equivalence relation. As such it partitions the set of integers Z into disjoint
equivalence classes by Theorem 37. This motivates the following definition.

63 Definition Let n be a positive integer. The n residue classes upon division by n are

0 = nZ, 1 = nZ + 1, 2 = nZ + 2, . . . , n − 1 = nZ + n − 1.

The set of residue classes modulo n is

Zn = {0, 1, . . . , n − 1}.

Our interest is now to define some sort of “addition” and some sort of “multiplication” in Zn .

64 Theorem (Addition and Multiplication Modulo n) Let n be a positive integer. For (a, b) ∈ (Zn )2 define
a + b = r, where r is the remainder of a + b upon division by n. and a · b = t, where t is the remainder of ab
upon division by n. Then these operations are well defined.

Proof: We need to prove that given arbitrary representatives of the residue classes, we always
obtain the same result from our operations. That is, if a = a ′ and b = b ′ then we have a + b =
a ′ + b ′ and a · b = a ′ · b ′ .

Now

a = a ′ =⇒ ∃(q, q ′ ) ∈ Z2 , r ∈ N, a = qn + r, a ′ = q ′ n + r, 0 ≤ r < n,
′
b=b =⇒ ∃(q1 , q1′ ) ∈ Z2 , r1 ∈ N, b = q1 n + r1 , b ′ = q1′ n + r1 , 0 ≤ r1 < n.

Hence
a + b = (q + q1 )n + r + r1 , a ′ + b ′ = (q ′ + q1′ )n + r + r1 ,
meaning that both a + b and a ′ + b ′ leave the same remainder upon division by n, and therefore

a + b = a + b = a ′ + b ′ = a ′ + b ′.

Similarly

ab = (qq1 n + qr1 + rq1 )n + rr1 , a ′ b ′ = (q ′ q1′ n + q ′ r1 + rq1′ )n + rr1 ,

and so both ab and a ′ b ′ leave the same remainder upon division by n, and therefore

a · b = ab = a ′ b ′ = a ′ · b ′ .

This proves the theorem. ❑

65 Example Let
Z6 = {0, 1, 2, 3, 4, 5}
be the residue classes modulo 6. Construct the natural addition + table for Z6 . Also, construct the natural
multiplication · table for Z6 .

Solution: The required tables are given in tables 1.1 and 1.2.

We notice that even though 2 6= 0 and 3 6= 0 we have 2 · 3 = 0. This prompts the following definition.

66 Definition (Zero Divisor) An element a 6= 0 of Zn is called a zero divisor if ab = 0 for some b ∈ Zn .

Zn 11

+ 0 1 2 3 4 5 · 0 1 2 3 4 5

0 0 1 2 3 4 5 0 0 0 0 0 0 0

1 1 2 3 4 5 0 1 0 1 2 3 4 5

2 2 3 4 5 0 1 2 0 2 4 0 2 4

3 3 4 5 0 1 2 3 0 3 0 3 0 3

4 4 5 0 1 2 3 4 0 4 2 0 4 2

5 5 0 1 2 3 4 5 0 5 4 3 2 1

Table 1.1: Addition table for Z6 . Table 1.2: Multiplication table for Z6 .

We will extend the concept of zero divisor later on to various algebras.

67 Example Let
Z7 = {0, 1, 2, 3, 4, 5, 6}
be the residue classes modulo 7. Construct the natural addition + table for Z7 . Also, construct the natural multi-
plication · table for Z7

Solution: The required tables are given in tables 1.3 and 1.4.

+ 0 1 2 3 4 5 6 · 0 1 2 3 4 5 6

0 0 1 2 3 4 5 6 0 0 0 0 0 0 0 0

1 1 2 3 4 5 6 0 1 0 1 2 3 4 5 6

2 2 3 4 5 6 0 1 2 0 2 4 6 1 3 5

3 3 4 5 6 0 1 2 3 0 3 6 2 5 1 4

4 4 5 6 0 1 2 3 4 0 4 1 5 2 6 3

5 5 6 0 1 2 3 4 5 0 5 3 1 6 4 2

6 6 0 1 2 3 4 5 6 0 6 5 4 3 2 1

Table 1.3: Addition table for Z7 . Table 1.4: Multiplication table for Z7 .

68 Example Solve the equation

5x = 3
in Z11 .
12 Chapter 1

Solution: Multiplying by 9 on both sides

45x = 27,
that is,
x = 5.

We will use the following result in the next section.

69 Definition Let a, b be integers with one of them different from 0. The greatest common divisor d of a, b,
denoted by d = gcd(a, b) is the largest positive integer that divides both a and b.

70 Theorem (Bachet-Bezout Theorem) The greatest common divisor of any two integers a, b can be written as a
linear combination of a and b, i.e., there are integers x, y with

gcd(a, b) = ax + by.

Proof: Let A = {ax + by : ax + by > 0, x, y ∈ Z}. Clearly one of ±a, ±b is in A, as one of

a, b is not zero. By the Well Ordering Principle, A has a smallest element, say d. Therefore,
there are x0 , y0 such that d = ax0 + by0 . We prove that d = gcd(a, b). To do this we prove that
d divides a and b and that if t divides a and b, then t must also divide then d.

We first prove that d divides a. By the Division Algorithm, we can find integers q, r, 0 ≤ r < d
such that a = dq + r. Then
r = a − dq = a(1 − qx0 ) − by0 .
If r > 0, then r ∈ A is smaller than the smaller element of A, namely d, a contradiction. Thus
r = 0. This entails dq = a, i.e. d divides a. We can similarly prove that d divides b.

Assume that t divides a and b. Then a = tm, b = tn for integers m, n. Hence d = ax0 + bx0 =
t(mx0 + ny0 ), that is, t divides d. The theorem is thus proved. ❑

71 Problem Write the addition and multiplication tables of in Z11 .

Z11 under natural addition and multiplication modulo 11.

72 Problem Solve the equation

73 Problem Prove that if n > 0 is a composite integer, Zn has
5x2 = 3 zero divisors.

1.5 Fields
74 Definition Let F be a set having at least two elements 0F and 1F (0F 6= 1F ) together with two operations
· (multiplication, which we usually represent via juxtaposition) and + (addition). A field hF, ·, +i is a triplet
satisfying the following axioms ∀(a, b, c) ∈ F3 :
F1 Addition and multiplication are associative:

(a + b) + c = a + (b + c), (ab)c = a(bc) (1.9)

F2 Addition and multiplication are commutative:

a + b = b + a, ab = ba (1.10)

F3 The multiplicative operation distributes over addition:

a(b + c) = ab + ac (1.11)
Fields 13

F4 0F is the additive identity:

0F + a = a + 0F = a (1.12)

F5 1F is the multiplicative identity:

1F a = a1F = a (1.13)

F6 Every element has an additive inverse:

∃ − a ∈ F, a + (−a) = (−a) + a = 0F (1.14)

F7 Every non-zero element has a multiplicative inverse: if a 6= 0F

∃a−1 ∈ F, aa−1 = a−1 a = 1F (1.15)

The elements of a field are called scalars.

An important property of fields is the following.

75 Theorem A field does not have zero divisors.

Proof: Assume that ab = 0F . If a 6= 0F then it has a multiplicative inverse a−1 . We deduce

a−1 ab = a−1 0F =⇒ b = 0F .

This means that the only way of obtaining a zero product is if one of the factors is 0F . ❑

76 Example hQ, ·, +i, hR, ·, +i, and hC, ·, +i are all fields. The multiplicative identity in each case is 1 and the
additive identity is 0.

77 Example Let √ √
Q( 2) = {a + 2b : (a, b) ∈ Q2 }
and define addition on this set as
√ √ √
(a + 2b) + (c + 2d) = (a + c) + 2(b + d),

and multiplication as √ √ √
(a + 2b)(c + 2d) = (ac + 2bd) + 2(ad + bc).
√ √ √
Then hQ + 2Q, , +i is a field. √ Observe 0F = 0, 1F = 1, that the additive inverse of a + 2b is −a − 2b, and
the multiplicative inverse of a + 2b, (a, b) 6= (0, 0) is
√ √
√ −1 1 a − 2b a 2b
(a + 2b) = √ = 2 2
= 2 2
− 2 .
a + 2b a − 2b a − 2b a − 2b2
√
Here a2 − 2b2 6= 0 since 2 is irrational.

78 Theorem If p is a prime, hZp , ·, +i is a field under · multiplication modulo p and + addition modulo p.

Proof: Clearly the additive identity is 0 and the multiplicative identity is 1. The additive
inverse of a is p − a. We must prove that every a ∈ Zp \ {0} has a multiplicative inverse. Such
an a satisfies gcd(a, p) = 1 and by the Bachet-Bezout Theorem 70, there exist integers x, y with
px + ay = 1. In such case we have

1 = px + ay = ay = a · y,

whence (a)−1 = y. ❑
14 Chapter 1

79 Definition A field is said to be of characteristic p 6= 0 if for some positive integer p we have ∀a ∈ F, pa = 0F ,

and no positive integer smaller than p enjoys this property.

If the field does not have characteristic p 6= 0 then we say that it is of characteristic 0. Clearly Q, R and C are of
characteristic 0, while Zp for prime p, is of characteristic p.

80 Theorem The characteristic of a field is either 0 or a prime.

Proof: If the characteristic of the field is 0, there is nothing to prove. Let p be the least positive
integer for which ∀a ∈ F, pa = 0F . Let us prove that p must be a prime. Assume that instead we
had p = st with integers s > 1, t > 1. Take a = 1F . Then we must have (st)1F = 0F , which entails
(s1F )(t1F ) = 0F . But in a field there are no zero-divisors by Theorem 75, hence either s1F = 0F or
t1F = 0F . But either of these equalities contradicts the minimality of p. Hence p is a prime. ❑

81 Problem Consider the set of numbers 83 Problem Let F be a field and a 6= 0F . Prove that
√ √ √ √ √ √ 4
Q( 2, 3, 6) = {a+b 2+c 3+d 6 : (a, b, c, d) ∈ Q }.
√ √ √
Assume that Q( 2, 3, 6) is a field under ordinary addi- (−a)−1 = −(a−1 ).
tion and multiplication.
√ √ What √ is the multiplicative inverse of
the element 2 + 2 3 + 3 6?
84 Problem Let F be a field and a, b two non-zero elements
of F. Prove that
82 Problem Let F be a field and a, b two non-zero elements
of F. Prove that
−(ab−1 ) = (−a)b−1 = a(−b−1 ). ab−1 = (−a)(−b−1 ).

1.6 Functions
85 Definition By a function or a mapping from one set to another, we mean a rule or mechanism that assigns to
every input element of the first set a unique output element of the second set. We shall call the set of inputs the
domain of the function, the set of possible outputs the target set of the function, and the set of actual outputs the
image of the function.

We will generally refer to a function with the following notation:

D → T
f: .
x 7→ f(x)

Here f is the name of the function, D is its domain, T is its target set, x is the name of a typical input and f(x) is
the output or image of x under f. We call the assignment x 7→ f(x) the assignment rule of the function. Sometimes
x is also called the independent variable. The set f(D) = {f(a)|a ∈ D} is called the image of f. Observe that
f(D) ⊆ T.

α β
1 2 1 4
2 8 2 2
3 4 3

Figure 1.4: An injection. Figure 1.5: Not an injection

Functions 15

X → Y
86 Definition A function f : is said to be injective or one-to-one if ∀(a, b) ∈ X2 , we have
x 7→ f(x)

a 6= b =⇒ f(a) 6= f(b).

This is equivalent to saying that

f(a) = f(b) =⇒ a = b.

87 Example The function α in the diagram 1.4 is an injective function. The function β represented by the diagram
1.5, however, is not injective, β(3) = β(1) = 4, but 3 6= 1.

88 Example Prove that

R \ {1} → R \ {1}
t:
x+1
x 7→
x−1
is an injection.

Solution: Assume t(a) = t(b). Then

a+1 b+1
t(a) = t(b) =⇒ =
a−1 b−1
=⇒ (a + 1)(b − 1) = (b + 1)(a − 1)

=⇒ ab − a + b − 1 = ab − b + a − 1

=⇒ 2a = 2b

=⇒ a = b

We have proved that t(a) = t(b) =⇒ a = b, which shews that t is injective.

β γ
1 4 1 4
2 2 2 2
3
8

Figure 1.6: A surjection Figure 1.7: Not a surjection

89 Definition A function f : A → B is said to be surjective or onto if (∀b ∈ B) (∃a ∈ A) : f(a) = b. That is,
each element of B has a pre-image in A.

☞A function is surjective if its image coincides with its target set. It is easy to see that a
graphical criterion for a function to be surjective is that every horizontal line passing through a
point of the target set (a subset of the y-axis) of the function must also meet the curve.

90 Example The function β represented by diagram 1.6 is surjective. The function γ represented by diagram 1.7 is
not surjective as 8 does not have a preimage.
16 Chapter 1

R → R
91 Example Prove that t : is a surjection.
3
x 7→ x

Solution: Since the graph of t is that of a cubic polynomial with only one zero, every horizontal line passing through
a point in R will eventually meet the graph of g, whence t is surjective. To prove this analytically, proceed as follows.
We must prove that (∀ b ∈ R) (∃a) such that t(a) = b. We choose a so that a = b1/3 . Then

t(a) = t(b1/3 ) = (b1/3 )3 = b.

Our choice of a works and hence the function is surjective.

92 Definition A function is bijective if it is both injective and surjective.

93 Problem Prove that 94 Problem Shew that


3
R → R R\ → R \ {3}
h: f: 2
x 7→ x3 6x
x 7→
2x − 3
is an injection. is a bijection.
Chapter 2
Matrices and Matrix Operations

2.1 The Algebra of Matrices

95 Definition Let hF, ·, +i be a field. An m × n (m by n) matrix A with m rows and n columns with entries over
F is a rectangular array of the form
2 3
6 a11 a12 ··· a1n 7
6 7
6 7
6a a22 ··· a2n 7
6 21 7
A=6
6 .
7,
6 . .. .. 7
7
6 . . ··· . 7
6 7
4 5
am1 am2 ··· amn

where ∀(i, j) ∈ {1, 2, . . . , m} × {1, 2, . . . , n}, aij ∈ F.

☞ As a shortcut, we often use the notation A = [a ij ] to denote the matrix A with entries aij .
Notice that when we refer to the matrix we put parentheses—as in “[aij ],” and when we refer to a
specific entry we do not use the surrounding parentheses—as in “aij .”

96 Example
2 3
60 −1 17
A=6
4
7
5
1 2 3

is a 2 × 3 matrix and
2 3
6−2 17
6 7
6 7
B=6
6 1
7
27
6 7
4 5
0 3

is a 3 × 2 matrix.

97 Example Write out explicitly the 4 × 4 matrix A = [aij ] where aij = i2 − j2 .

17
18 Chapter 2

Solution: This is

2 3 2 3
2 1
61 − 1 12 − 22 12 − 32 12 − 42 7 60 −3 −8 −157
6 7 6 7
6 7 6 7
6 22 − 12 22 − 22 22 − 32 22 − 42 7 6 −5 −127
6 7 63 0 7
A=6
6
7=6
7 6
7.
7
6 2 7 6 7
6 3 − 12 32 − 22 32 − 32 32 − 42 7 6 8 5 0 −7 7
6 7 6 7
4 5 4 5
42 − 12 42 − 22 42 − 32 42 − 42 15 12 7 0

98 Definition Let hF, ·, +i be a field. We denote by Mm×n (F) the set of all m × n matrices with entries over F.
If m = n we use the abbreviated notation Mn (F) = Mn×n (F). Mn (F) is thus the set of all square matrices of size
n with entries over F.

99 Definition The m × n zero matrix 0m×n ∈ Mm×n (F) is the matrix with 0F ’s everywhere,

2 3
60F 0F 0F ··· 0F 7
6 7
6 7
6 7
60F 0F 0F ··· 0F 7
6 7
6 7
6 7
0m×n = 60F 0F 0F ··· 0F 7 .
6 7
6 7
6 . .. .. .. 7
6 .. . . ··· .7
6 7
6 7
4 5
0F 0F 0F ··· 0F

When m = n we write 0n as a shortcut for 0n×n .

100 Definition The n × n identity matrix In ∈ Mn (F) is the matrix with 1F ’s on the main diagonal and 0F ’s
everywhere else,
2 3
61 F 0F 0F ··· 0F 7
6 7
6 7
6 7
60 F 1F 0F ··· 0F 7
6 7
6 7
6 7
In = 60 F 0F 1F ··· 0F 7 .
6 7
6 7
6 . .. .. .. 7
6 .. . . ··· .7
6 7
6 7
4 5
0F 0F 0F ··· 1F

101 Definition (Matrix Addition and Multiplication of a Matrix by a Scalar) Let A = [aij ] ∈ Mm×n (F), B =
[bij ] ∈ Mm×n (F) and α ∈ F. The matrix A + αB is the matrix C ∈ Mm×n (F) with entries C = [cij ] where
cij = aij + αbij .
The Algebra of Matrices 19

2 3 2 3
6 1 17 6−1 1 7
6 7 6 7
6 7 6 7
102 Example For A = 6
6−1
7 and B = 6
17 6 2
7
1 7 we have
6 7 6 7
4 5 4 5
0 2 0 −1

2 3
6−1 37
6 7
6 7
A + 2B = 6
6 3
7
37.
6 7
4 5
0 0.

103 Theorem Let (A, B, C) ∈ (Mm×n (F))3 and (α, β) ∈ F2 . Then

M1 Mm×n (F) is close under matrix addition and scalar multiplication

A + B ∈ Mm×n (F), αA ∈ Mm×n (F) (2.1)

M2 Addition of matrices is commutative

A+B=B+A (2.2)

M3 Addition of matrices is associative

A + (B + C) = (A + B) + C (2.3)

M4 There is a matrix 0m×n such that

A + 0m×n (2.4)

M5 There is a matrix −A such that

A + (−A) = (−A) + A = 0m×n (2.5)

M6 Distributive law
α(A + B) = αA + αB (2.6)

M7 Distributive law
(α + β)A = αA + βB (2.7)

M8
1F A = A (2.8)

M9
α(βA) = (αβ)A (2.9)

Proof: The theorem follows at once by reducing each statement to an entry-wise and appealing
to the field axioms. ❑

104 Problem Write out explicitly the 3 × 3 matrix A = [aij ] 106 Problem Let
where aij = ij . 2 3 2 3
6 a −2a c 7 6 1 2a c 7
6 7 6 7
6 7 6 7
M=6 0 −a b7 , N=6 a b−a −b7
6 7 6 7
4 5 4 5
105 Problem Write out explicitly the 3 × 3 matrix A = [aij ] a+b 0 −1 a−b 0 −1
where aij = ij.
be square matrices with entries over R. Find M + N and 2M.
20 Chapter 2

107 Problem Determine x and y such that shortest giant. Another person goes along the rows and asks
2 3 2 3 2 3 the shortest person to stand up and from these he selects the
tallest, which we will call the tallest midget. Who is taller,
63 x 17 62 1 37 6 7 3 77
the tallest midget or the shortest giant?
4 5 + 24 5=4 5.
1 2 0 5 x 4 11 y 8
111 Problem (Putnam Exam, 1959) Choose five elements
from the matrix
108 Problem Determine 2 × 2 matrices A and B such that
2 3
2 3 2 3
−27 611 17 25 19 167
61 64 27
6 7
2A − 5B = 4 5, −2A + 6B = 4 5. 6 7
624 10 13 15 37
0 1 6 0 6 7
6 7
6 7
612 5 14 2 187 ,
6 7
109 Problem Let A = [aij ] ∈ Mn (R). Prove that 6 7
6 7
623 4 1 8 227
min max aij ≥ max min aij . 6 7
j i i j 4 5
6 20 7 21 9
110 Problem A person goes along the rows of a movie theater
and asks the tallest person of each row to stand up. Then no two coming from the same row or column, so that the
he selects the shortest of these people, who we will call the minimum of these five elements is as large as possible.

2.2 Matrix Multiplication

112 Definition Let A = [aij ] ∈ Mm×n (F) and B = P [bij ] ∈ Mn×p (F). Then the matrix product AB is defined as
the matrix C = [cij ] ∈ Mm×p (F) with entries cij = nl=1 ail blj :

2 3 2 3
6 a11 a12 ··· a1n 7 6 c11 ··· c1p 7
6 72 3 6 7
6 7 6 7
6 7 6 7
6 a21 a22 ··· a2n 7 6 b11 ··· b1j ··· b1p 7 6 c21 ··· c2p 7
6 76 7 6 7
6 7 7 6 7
6 .. .. .. 7 6
6b 6 .. .. 7
6 .
6 . ··· . 7
76 ··· b2j ··· b2p 7
7 6 . ··· . 7
7=6 7
21
6 766 7 6 7.
6 76 . .. .. 7 6 7
6 ai1 ain 7 .. . . 7 6 ··· 7
6 ai2 ··· 766
··· ···
7 6
··· cij 7
6 74 5 6 7
6 . .. .. 7 6 . .. 7
6 .. . ··· . 7 ··· ··· 6 .. ··· . 7
6 7 bn1 bnj bnp 6 7
6 7 6 7
4 5 4 5
am1 am2 ··· amn cm1 ··· cmp

☞ Observe that we use juxtaposition rather than a special symbol to denote matrix multiplica-
tion. This will simplify notation.In order to obtain the ij-th entry of the matrix AB we multiply
elementwise the i-th row of A by the j-th column of B. Observe that AB is a m × p matrix.

2 3 2 3
61 27 65 67
113 Example Let M = 6
4
7 and N = 6
5 4
7 be matrices over R. Then
5
3 4 7 8

2 32 3 2 3 2 3
61 27 65 6761 · 5 + 2 · 7 1 · 6 + 2 · 87 619 227
MN = 6
4
76
54
7=6
5 4
7=6
5 4
7,
5
3 4 7 8 3·5+4·7 3·6+4·8 43 50
Matrix Multiplication 21

and 2 32 3 2 3 2 3
65 67 61 27 65 · 1 + 6 · 3 5 · 2 + 6 · 47 623 347
NM = 6
4
76
54
7=6
5 4
7=6
5 4
7.
5
7 8 3 4 7·1+8·3 7·2+8·4 31 46

Hence, in particular, matrix multiplication is not necessarily commutative.

114 Example We have

2 32 3 2 3

17 6 2 −1 −17 60 0 07
61 1
6 76 7 6 7
6 76 7 6 7
6 7 6−1 7 6 7
61 1 17 6 2 −17 = 60 0 07 ,
6 76 7 6 7
4 564
7 6
5 4
7
5
1 1 1 −1 −1 2 0 0 0

over R. Observe then that the product of two non-zero matrices may be the zero matrix.

115 Example Consider the matrix

2 3
62 1 37
6 7
6 7
A=6
60 1
7
17
6 7
4 5
4 4 0

with entries over Z5 . Then 2 32 3

62 1 37 62 1 37
6 76 7
6 76 7
A2 = 6 76 7
60 1 17 60 1 17
6 76 7
4 54 5
4 4 0 4 4 0
2 3
61 0 27
6 7
6 7
6
= 64 0 17
7.
6 7
4 5
3 3 1

☞ Even though matrix multiplication is not necessarily commutative, it is associative.

116 Theorem If (A, B, C) ∈ Mm×n (F) × Mn×r (F) × Mr×s (F) we have

(AB)C = A(BC),

i.e., matrix multiplication is associative.

Proof: To shew this we only need to consider the ij-th entry of each side, appeal to the associa-
tivity of the underlying field F and verify that both sides are indeed equal to
n X
X r
aik bkk ′ ck ′ j .
k=1 k ′ =1

❑
22 Chapter 2

☞ By virtue of associativity, a square matrix commutes with its powers, that is, if A ∈ Mn (F),
and (r, s) ∈ N2 , then (Ar )(As ) = (As )(Ar ) = Ar+s .

117 Example Let A ∈ M3 (R) be given by

2 3
61 1 17
6 7
6 7
A=6
61 1
7
17 .
6 7
4 5
1 1 1

Demonstrate, using induction, that An = 3n−1 A for n ∈ N, n ≥ 1.

Solution: The assertion is trivial for n = 1. Assume its truth for n − 1, that is, assume An−1 = 3n−2 A. Observe
that
2 32 3 2 3
61 1 17 6 1 1 17 63 3 37
6 76 7 6 7
6 76 7 6 7
A =6
2
61 1
76
17 6 1 1
7 6
17 = 6 3 3
7
37 = 3A.
6 76 7 6 7
4 54 5 4 5
1 1 1 1 1 1 3 3 3

Now

An = AAn−1 = A(3n−2 A) = 3n−2 A2 = 3n−2 3A = 3n−1 A,

and so the assertion is proved by induction.

118 Theorem Let A ∈ Mn (F). Then there is a unique identity matrix. That is, if E ∈ Mn (F) is such that
AE = EA = A, then E = In .

Proof: It is clear that for any A ∈ Mn (F), AIn = In A = A. Now because E is an identity,
EIn = In . Because In is an identity, EIn = E. Whence

In = EIn = E,

demonstrating uniqueness. ❑

119 Example Let A = [aij ] ∈ Mn (R) be such that aij = 0 for i > j and aij = 1 if i ≤ j. Find A2 .

Solution: Let A2 = B = [bij ]. Then

n
X
bij = aik akj .
k=1

Observe that the i-th row of A has i − 1 0’s followed by n − i + 1 1’s, and the j-th column of A has j 1’s followed
by n − j 0’s. Therefore if i − 1 > j, then bij = 0. If i ≤ j + 1, then

j
X
bij = aik akj = j − i + 1.
k=i
Matrix Multiplication 23

This means that 2 3

61 2 3 4 ··· n−1 n 7
6 7
6 7
6 7
60 1 2 3 ··· n−2 n − 17
6 7
6 7
6 7
60 0 1 2 ··· n−3 n − 27
6 7
A2 = 6 7.
6. .. .. .. .. .. 7
6 .. . . . ··· . . 7
6 7
6 7
6 7
60 0 0 0 ··· 1 2 7
6 7
6 7
4 5
0 0 0 0 ··· 0 1

120 Problem Determine the product 126 Problem Prove, using mathematical induction, that
2 3n 2 3
2 32 32 3
−17 6−2 61 17 61 n7
61 1 7 61 17 4 5 =4 5.
4 54 54 5
0 1 0 1
1 1 0 −1 1 2
.
2 3
2 3 2 3 6 1 −17
6
127 Problem Let M = 4 5. Find M .
61 0 07 6a b c7
6 7 6 7 −1 1
6 7 6 7
121 Problem Let A = 61 1 07 , B = 6c a b7 . Find
6 7 6 7
4 5 4 5 2 3
1 1 1 b c a
60 37
2003
AB and BA. 128 Problem Let A = 4 5. Find, with proof, A .
2 0
122 Problem Let
2 3 2 3
129 Problem Let (A, B, C) ∈ Ml×m (F) × Mm×n (F) ×
62 3 4 17 61 1 1 17
Mm×n (F) and α ∈ F. Prove that
6 7 6 7
6 7 6 7
61 2 3 47 61 1 1 17
6 7 6 7 A(B + C) = AB + AC,
A=6 7, B=6 7
6 7 6 7
64 1 2 37 61 1 1 17
6 7 6 7 (A + B)C = AC + BC,
4 5 4 5
3 4 1 2 1 1 1 1 α(AB) = (αA)B = A(αB).

be matrices in M4 (Z5 ) . Find the products AB and BA.

130 Problem Let A ∈ M2 (R) be given by

123 Problem Solve the equation 2 3

2 32 2 3 6cos α − sin α7
A=4 5.
6−4 x7 6−1 0 7
sin α cos α
4 5 =4 5
−x 4 0 −1
Demonstrate, using induction, that for n ∈ N, n ≥ 1.
over R. 2 3
6cos nα − sin nα7
124 Problem Prove or disprove! If (A, B) ∈ (Mn (F))2 are An = 4 5.
such that AB = 0n , then also BA = 0n . sin nα cos nα

125 Problem Prove or disprove! For all matrices (A, B) ∈

(Mn (F))2 , 131 Problem A matrix A = [aij ] ∈ Mn (R) is said to be
checkered if aij = 0 when (j − i) is odd. Prove that the sum
(A + B)(A − B) = A2 − B2 . and the product of two checkered matrices is checkered.
24 Chapter 2

132 Problem Let A ∈ M3 (R), 134 Problem Let (A, B) ∈ (Mn (F))2 and k be a positive
2 3 integer such that Ak = 0n . If AB = B prove that B = 0n .

61 1 17
6 7 2 3
6 7
A = 60 17 .
6 1 7 6a b7
4 5 135 Problem Let A = 4 5. Demonstrate that
0 0 1 c d

Prove that 2 3 A2 − (a + d)A + (ad − bc)I2 = 02

n(n+1)
61 n 2 7
6 7 .
6 7
An = 60 1 n 7.
6 7
4 5
0 0 1 136 Problem Let A ∈ M2 (F) and let k ∈ Z, k > 2. Prove
that Ak = 02 if and only if A2 = 02 .

133 Problem Prove, by means of induction that for the fol-

137 Problem Find all matrices A ∈ M2 (R) such that A2 =
lowing n × n we have
02
2 33 2 3
n(n+1)
61 1 1 ··· 17 61 3 6 ··· 2 7
6 7 6 7 138 Problem Find all matrices A ∈ M2 (R) such that A2 = I2
6 7 6 7
60 1 1 ··· 17 7 60 1 3 ··· (n−1)n 7
6 6 2 7
6 7 6 7
6 7 6 7 139 Problem Find a solution X ∈ M2 (R) for
60 0 1 ··· 17 = 60 0 1 ··· (n−2)(n−1) 7 .
6 7 6 2 7
6 7 6 7 2 3
6 .. .. .. .. 7 6 .. .. .. 7
6. . . ··· .7 6. . ··· ··· . 7
6 7 6 7 6 −1 0 7
4 5 4 5 X2 − 2X = 4 5.
0 0 0 ··· 1 0 0 0 ··· 1 6 3

2.3 Trace and Transpose

140 Definition Let A = [aij ] ∈ Mn (F). Then the trace of A, denoted by tr (A) is the sum of the diagonal elements
of A, that is
Xn
tr (A) = akk .
k=1

141 Theorem Let A = [aij ] ∈ Mn (F), B = [bij ] ∈ Mn (F). Then

tr (A + B) = tr (A) + tr (B) , (2.10)

tr (AB) = tr (BA) . (2.11)

Pn
Proof: P The first assertion is trivial. To prove the second, observe that AB = ( k=1 aik bkj ) and
n
BA = ( k=1 bik akj ). Then
n X
X n n X
X n
tr (AB) = aik bki = bki aik = tr (BA) ,
i=1 k=1 k=1 i=1

whence the theorem follows. ❑

142 Example Let A ∈ Mn (R). Shew that A can be written as the sum of two matrices whose trace is different
from 0.
Trace and Transpose 25

Solution: Write
A = (A − αIn ) + αIn .
tr (A)
Now, tr (A − αIn ) = tr (A) − nα and tr (αIn ) = nα. Thus it suffices to take α 6= , α 6= 0. Since R has
n
infinitely many elements, we can find such an α.

143 Example Let A, B be square matrices of the same size and over the same field of characteristic 0. Is it possible
that AB − BA = In ? Prove or disprove!

Solution: This is impossible. For if, taking traces on both sides

0 = tr (AB) − tr (BA) = tr (AB − BA) = tr (In ) = n

a contradiction, since n > 0.

144 Definition The transpose of a matrix of a matrix A = [aij ] ∈ Mm×n (F) is the matrix AT = B = [bij ] ∈
Mn×m (F), where bij = aji .

145 Example If
2 3
6a b c7
6 7
6 7
M=6
6d e f7
7,
6 7
4 5
g h i

with entries in R, then

2 3
6a d g7
6 7
6 7
M =6
T
6b e
7
h7 .
6 7
4 5
c f i

146 Theorem Let

A = [aij ] ∈ Mm×n (F), B = [bij ] ∈ Mm×n (F), C = [cij ] ∈ Mn×r (F), α ∈ F, u ∈ N.

Then
ATT = A, (2.12)
T T T
(A + αB) = A + αB , (2.13)
(AC)T = CT AT , (2.14)
(Au )T = (AT )u . (2.15)

Proof: The first two assertions are obvious, and the fourth follows from the third by using
induction. To prove the third put AT = (αij ), αij = aji , CT = (γij ), γij = cji , AC = (uij ) and
CT AT = (vij ). Then
Xn n
X Xn
uij = aik ckj = αki γjk = γjk αki = vji ,
k=1 k=1 k=1

whence the theorem follows. ❑

147 Definition A square matrix A ∈ Mn (F) is symmetric if AT = A. A matrix B ∈ Mn (F) is skew-symmetric

if BT = −B.
26 Chapter 2

148 Example Let A, B be square matrices of the same size, with A symmetric and B skew-symmetric. Prove that
the matrix A2 BA2 is skew-symmetric.

Solution: We have
(A2 BA2 )T = (A2 )T (B)T (A2 )T = A2 (−B)A2 = −A2 BA2 .

149 Theorem Let F be a field of characteristic different from 2. Then any square matrix A can be written as the
sum of a symmetric and a skew-symmetric matrix.

Proof: Observe that

(A + AT )T = AT + ATT = AT + A,
and so A + AT is symmetric. Also,
(A − AT )T = AT − ATT = −(A − AT ),
and so A − AT is skew-symmetric. We only need to write A as
A = (2−1 )(A + AT ) + (2−1 )(A − AT )
to prove the assertion. ❑

150 Problem Write 155 Problem

Given a square matrix A ∈ M4 (R) such that
2 3 tr A2 = −4, and
61 2 37
6 7 (A − I4 )2 = 3I4 ,
6 7
A = 62 3 17 ∈ M3 (R)
6 7
4 5 find tr (A).
3 1 2
156 Problem Prove or disprove! If A, B are square matri-
as the sum of two 3 × 3 matrices E1 , E2 , with tr (E2 ) = 10. ces of the same size, then it is always true that tr (AB) =
tr (A) tr (B).
151 Problem Shew that there are no matrices (A, B, C, D) ∈
(Mn (R))4 such that
157 Problem Prove or disprove! If (A, B, C) ∈ (M3 (F))3
AC + DB = In , then tr (ABC) = tr (BAC).

CA + BD = 0n .
158 Problem Let A be a square matrix. Prove that the matrix
AAT is symmetric.
152 Problem Let (A, B) ∈ (M2 (R))2 be symmetric matri-
ces. Must their product AB be symmetric? Prove or disprove!
159 Problem Let A, B be square matrices of the same size,
2
with A symmetric and B skew-symmetric. Prove that the
153 Problem Given

square
matrices (A, B) ∈ (M7 (R)) matrix AB − BA is symmetric.
such that tr A2 = tr B2 = 1, and

(A − B)2 = 3I7 , 160 Problem

P Let APn ∈ 2Mn (F), A = [aij ]. Prove that
tr AAT = n i=1 j=1 aij .
find tr (BA).

2 3 161 Problem Let X ∈ Mn (R). Prove that if XXT = 0n then

X = 0n .
6a b7
154 Problem Consider the matrix A = 4 5 ∈ M2 (R).
c d 162 Problem Let m, n, p be positive integers and A ∈
Find necessary

and sufficient conditions on a, b, c, d so that Mm×n (R), B ∈ Mn×p (R), C ∈ Mp×m (R). Prove that
tr A2 = (tr (A))2 . (BA)T A = (CA)T A =⇒ BA = CA.

2.4 Special Matrices

163 Definition The main diagonal of a square matrix A = [aij ] ∈ Mn (F) is the set {aii : i ≤ n}. The counter
diagonal of a square matrix A = [aij ] ∈ Mn (F) is the set {a(n−i+1)i : i ≤ n}.
Special Matrices 27

164 Example The main diagonal of the matrix

2 3
60 1 57
6 7
6 7
A=6
63 2
7
47
6 7
4 5
9 8 7

is the set {0, 2, 7}. The counter diagonal of A is the set {5, 2, 9}.

165 Definition A square matrix is a diagonal matrix if every entry off its main diagonal is 0F .

166 Example The matrix

2 3
61 0 07
6 7
6 7
A=6
60 2
7
07
6 7
4 5
0 0 3

is a diagonal matrix.

167 Definition A square matrix is a scalar matrix if it is of the form αIn for some scalar α.

168 Example The matrix

2 3
64 0 07
6 7
6 7
A=6
60 4
7
07 = 4I3
6 7
4 5
0 0 4

is a scalar matrix.

169 Definition A ∈ Mm×n (F) is said to be upper triangular if

(∀(i, j) ∈ {1, 2, · · · , n}2 ), (i > j, aij = 0F ),

that is, every element below the main diagonal is 0F . Similarly, A is lower triangular if

(∀(i, j) ∈ {1, 2, · · · , n}2 ), (i < j, aij = 0F ),

that is, every element above the main diagonal is 0F .

170 Example The matrix A ∈ M3×4(R) shewn is upper triangular and B ∈ M4 (R) is lower triangular.
2 3
2 3
61 0 0 07
6 7
61 a b c7 6 7
6 7 61 a 0 07
6 7 6 7
A=6
60 2 3
7 B=6
07 6
7
7
6 7 6 7
4 5 60 2 3 07
6 7
0 0 0 1 4 5
1 1 t 1
28 Chapter 2

171 Definition The Kronecker delta δij is defined by




 1F if i = j
δij =


 0F if i 6= j

172 Definition The set of matrices Eij ∈ Mm×n (F), Eij = (ers ) such that eij = 1F and ei ′ j ′ = 0F , (i ′ , j ′ ) 6= (i, j)
is called the set of elementary matrices. Observe that in fact ers = δir δsj .

Elementary matrices have interesting effects when we pre-multiply and post-multiply a matrix by them.

173 Example Let

2 3
2 3
60 0 07
6 7
61 2 3 47 6 7
6 7 60 0 17
6 7 6 7
A=6
65 6 7
7
8 7, E23 =6
6
7.
7
6 7 6 7
4 5 60 0 07
6 7
9 10 11 12 4 5
0 0 0

Then
2 3 2 3
60 0 0 07 60 0 27
6 7 6 7
6 7 6 7
E23 A = 6
69 10 11
7
127 , AE23 =6
60 0
7
6 7.
6 7 6 7
4 5 4 5
0 0 0 0 0 0 10

174 Theorem (Multiplication by Elementary Matrices) Let Eij ∈ Mm×n (F) be an elementary matrix, and let
A ∈ Mn×m (F). Then Eij A has as its i-th row the j-th row of A and 0F ’s everywhere else. Similarly, AEij has as
its j-th column the i-th column of A and 0F ’s everywhere else.

Proof: Put (αuv ) = Eij A. To obtain Eij A we multiply the rows of Eij by the columns of A.
Now
n
X n
X
αuv = euk akv = δui δkj akv = δui ajv .
k=1 k=1

Therefore, for u 6= i, αuv = 0F , i.e., off of the i-th row the entries of Eij A are 0F , and αiv = αjv ,
that is, the i-th row of Eij A is the j-th row of A. The case for AEij is similarly argued.❑

The following corollary is immediate.

175 Corollary Let (Eij , Ekl ) ∈ (Mn (F))2 , be square elementary matrices. Then

Eij Ekl = δjk Eil .

176 Example Let M ∈ Mn (F) be a matrix such that AM = MA for all matrices A ∈ Mn (F). Demonstrate that
M = aIn for some a ∈ F, i.e. M is a scalar matrix.
Special Matrices 29

Solution: Assume (s, t) ∈ {1, 2, . . . , n}2 . Let M = (mij ) and Est ∈ Mn (F). Since M commutes with Est we have
2 3 2 3
6 0 0 ... 0 7 60 0 ... m1s . . . 07
6 7 6 7
6
6 .. .. .. 77
6
6 .. .. 7
7
6 . . ... . 7 60 0 . m2s . 07
6 7 6 7
6 7 6 7
6 7 6. .. .. .. .. .. 7
6mt1 mt2 ... mtn 7 = Est M = MEst = 6 .. . . . . .7
6 7 6 7
6 7 6 7
6 . .. .. 7 6 .. .. 7
6 .. . ... 7
. 7 60 0 . m(n−1)s . 07
6 6 7
6 7 6 7
4 5 4 .. .. 5
0 0 ... 0 0 0 . mns . 0

For arbitrary s 6= t we have shown that mst = mts = 0, and that mss = mtt . Thus the entries off the main diagonal
are zero and the diagonal entries are all equal to one another, whence M is a scalar matrix.

177 Definition Let λ ∈ F and Eij ∈ Mn (F). A square matrix in Mn (F) of the form In + λEij is called a
transvection.

178 Example The matrix

2 3
61 0 47
6 7
6 7
T = I3 + 4E13 =6
60 1
7
07
6 7
4 5
0 0 1

is a transvection. Observe that if 2 3

61 1 17
6 7
6 7
A=6
65 6
7
77
6 7
4 5
1 2 3

then 2 32 3 2 3
61 0 47 61 1 17 65 9 137
6 76 7 6 7
6 76 7 6 7
TA = 6
60 1
76
07 65 6
7 6
7 7 = 65 6
7
7 7,
6 76 7 6 7
4 54 5 4 5
0 0 1 1 2 3 1 2 3

that is, pre-multiplication by T adds 4 times the third row of A to the first row of A. Similarly,
2 32 3 2 3
61 1 17 61 0 47 61 1 57
6 76 7 6 7
6 76 7 6 7
AT = 6
65 6
76
77 60 1
7 6
07 = 65 6
7
277 ,
6 76 7 6 7
4 54 5 4 5
1 2 3 0 0 1 1 2 7

that is, post-multiplication by T adds 4 times the first column of A to the third row of A.

In general, we have the following theorem.

179 Theorem (Multiplication by a Transvection Matrix) Let In + λEij ∈ Mn (F) be a transvection and let A ∈
Mn×m (F). Then (In + λEij )A adds the j-th row of A to its i-th row and leaves the other rows unchanged.
30 Chapter 2

Similarly, if B ∈ Mp×n (F), B(In + λEij ) adds the i-th column of B to the j-th column and leaves the other
columns unchanged.

Proof: Simply observe that (In + λEij )A = A + λEij A and A(In + λEij ) = A + λAEij and apply
Theorem 174. ❑

Observe that the particular transvection In + (λ − 1F )Eii ∈ Mn (F) consists of a diagonal matrix with 1F ’s
everywhere on the diagonal, except on the ii-th position, where it has a λ.

180 Definition If λ 6= 0F , we call the matrix In + (λ − 1F )Eii a dilatation matrix.

181 Example The matrix

2 3
64 0 07
6 7
6 7
S = I3 + (4 − 1)E11 =6
60 1 07
7
6 7
4 5
0 0 1

is a dilatation matrix. Observe that if 2 3

61 1 17
6 7
6 7
A=6
65 6
7
77
6 7
4 5
1 2 3

then 2 32 3 2 3
64 0 0 7 61 1 17 64 4 47
6 76 7 6 7
6 76 7 6 7
SA = 6
60 1
76
0 7 65 6
7 6
77 = 65 6
7
77 ,
6 76 7 6 7
4 54 5 4 5
0 0 1 1 2 3 1 2 3

that is, pre-multiplication by S multiplies by 4 the first row of A. Similarly,

2 32 3 2 3
61 1 17 64 0 07 64 1 17
6 76 7 6 7
6 76 7 6 7
AS = 6
65 6
76
77 60 1
7 6
07 = 620 6
7
77 ,
6 76 7 6 7
4 54 5 4 5
1 2 3 0 0 1 4 2 3

that is, post-multiplication by S multiplies by 4 the first column of A.

182 Theorem (Multiplication by a Dilatation Matrix) Pre-multiplication of

the matrixA ∈ Mn×m (F) by the dilatation matrix In + (λ − 1F )Eii ∈ Mn (F) multiplies the i-th row of A by λ
and leaves the other rows of A unchanged. Similarly, if B ∈ Mp×n (F) post-multiplication of B by In + (λ − 1F )Eii
multiplies the i-th column of B by λ and leaves the other columns of B unchanged.

Proof: This follows by direct application of Theorem 179. ❑

183 Definition We write Iij

n for the matrix which permutes the i-th row with the j-th row of the identity matrix.
We call Iij
n a transposition matrix.
Special Matrices 31

184 Example We have

2 3
61 0 0 07
6 7
6 7
60 0 1 07
6 7
=6 7.
(23)
I4 6 7
6 7
60 1 0 07
6 7
4 5
0 0 0 1
If 2 3
61 2 3 47
6 7
6 7
65 6 7 87
6 7
A=6
6
7,
7
6 7
69 10 11 127
6 7
4 5
13 14 15 16
then 2 3
61 2 3 47
6 7
6 7
69 10 11 127
6 7
I4 A = 6 7,
(23)
6 7
6 7
65 6 7 87
6 7
4 5
13 14 15 16
and 2 3
61 3 2 47
6 7
6 7
65 7 6 87
6 7
=6 7.
(23)
AI4 6 7
6 7
69 11 10 127
6 7
4 5
13 15 14 16

185 Theorem (Multiplication by a Transposition Matrix) If A ∈ Mn×m (F), then Iij n A is the matrix obtained
from A permuting the the i-th row with the j-th row of A. Similarly, if B ∈ Mp×n (F), then BIij
n is the matrix
obtained from B by permuting the i-th column with the j-th column of B.

Proof: We must prove that Iijn A exchanges the i-th and j-th rows but leaves the other rows
unchanged. But this follows upon observing that
Iij
n = In + Eij + Eji − Eii − Ejj

and appealing to Theorem 174.

❑

186 Definition A square matrix which is either a transvection matrix, a dilatation matrix or a transposition matrix
is called an elimination matrix.

☞ In a very loose way, we may associate pre-multiplication of a matrix A by another matrix

with an operation on the rows of A, and post-multiplication of a matrix A by another with an
operation on the columns of A.
32 Chapter 2

187 Problem Consider the matrices is transformed into the matrix

2 3 2 3 2 3

6 1 0 1 07 64 −2 4 27 6 h−g g i7
6 7 6 7 6 7
6 7 6 7 6 7
B =6 e−d d f7
6 0 1 0 17 60 1 0 17 6 7
6 7 6 7 4 5
A=6 7, B=6 7.
6 7 6 7 2b − 2a 2a 2c
6−1 1 1 17 61 1 −1 17
6 7 6 7
4 5 4 5
by a series of row and column operations. Find explicit per-
1 −1 1 1 1 −1 1 1
mutation matrices P, P ′ , an explicit dilatation matrix D, and
an explicit transvection matrix T such that
Find a specific dilatation matrix D, a specific transposition
matrix P, and a specific transvection matrix T such that B = DPAP ′ T.
B = TDAP.
189 Problem Let A ∈ Mn (F). Prove that if

(∀X ∈ Mn (F)), (tr (AX) = tr (BX)),

188 Problem The matrix
2 3 then A = B.

6a b c7
6 7 190 Problem Let A ∈ Mn (R) be such that
6 7
A = 6d e f7
6 7 (∀X ∈ Mn (R)), ((XA)2 = 0n ).
4 5
g h i Prove that A = 0n .

2.5 Matrix Inversion

191 Definition Let A ∈ Mm×n (F). Then A is said to be left-invertible if ∃L ∈ Mn×m (F) such that LA = In . A
is said to be right-invertible if ∃R ∈ Mn×m (F) such that AR = Im . A matrix is said to be invertible if it possesses
a right and a left inverse. A matrix which is not invertible is said to be singular.

192 Example The matrix A ∈ M2×3(R)

2 3
61 0 07
A=6
4
7
5
0 1 0

has infinitely many right-inverses of the form

2 3
61 07
6 7
6 7
R(x,y) =6
60 17
7.
6 7
4 5
x y

For
2 3
2 3 1 07 2 3
6
61 0 07 6
6
7
7 61 07
6 76 7 6 7,
4 5 60 17 = 4 5
6 7
0 1 0 4 5 0 1
x y
Matrix Inversion 33

regardless of the values of x and y. Observe, however, that A does not have a left inverse, for
2 3 2 3
6a b7 2 3
6a b 07
6 7 1 0 07 6 7
6 76 6 7
6
6c d7
74
6 7=6
5 6c d 07
7,
6 7 6 7
4 5 0 1 0 4 5
f g f g 0

which will never give I3 regardless of the values of a, b, c, d, f, g.

193 Example If λ 6= 0, then the scalar matrix λIn is invertible, for

(λIn ) λ−1 In = In = λ−1 In (λIn ) .

194 Example The zero matrix 0n is singular.

195 Theorem Let A ∈ Mn (F) a square matrix possessing a left inverse L and a right inverse R. Then L = R. Thus
an invertible square matrix possesses a unique inverse.

Proof: Observe that we have LA = In = AR. Then

L = LIn = L(AR) = (LA)R = In R = R.

196 Definition The subset of Mn (F) of all invertible n × n matrices is denoted by GLn (F), read “the linear group
of rank n over F.”

197 Corollary Let (A, B) ∈ (GLn (F))2 . Then AB is also invertible and

(AB)−1 = B−1 A−1 .

Proof: Since AB is a square matrix, it suffices to notice that

B−1 A−1 (AB) = (AB)B−1 A−1 = In

and that since the inverse of a square matrix is unique, we must have B−1 A−1 = (AB)−1 . ❑

198 Corollary If a square matrix S ∈ Mn (F) is invertible, then S−1 is also invertible and (S−1 )−1 = S, in view of
the uniqueness of the inverses of square matrices.

199 Corollary If a square matrix A ∈ Mn (F) is invertible, then AT is also invertible and (AT )−1 = (A−1 )T .

Proof: We claim that (AT )−1 = (A−1 )T . For

AA−1 = In =⇒ (AA−1 )T = ITn =⇒ (A−1 )T AT = In ,

where we have used Theorem 146. ❑

The next few theorems will prove that elimination matrices are invertible matrices.

200 Theorem (Invertibility of Transvections) Let In + λEij ∈ Mn (F) be a transvection, and let i 6= j. Then

(In + λEij )−1 = In − λEij .

34 Chapter 2

Proof: Expanding the product

(In + λEij )(In − λEij ) = In + λEij − λEij − λ2 Eij Eij

= In − λ2 δij Eij

= In ,

since i 6= j. ❑

201 Example By Theorem 200, we have

2 32 3 2 3
61 0 37 61 0 −37 61 0 07
6 76 7 6 7
6 76 7 6 7
6 76 7 6 7
60 1 07 60 1 0 7 = 60 1 07 .
6 76 7 6 7
4 54 5 4 5
0 0 1 0 0 1 0 0 1

202 Theorem (Invertibility of Dilatations) Let λ 6= 0F . Then

(In + (λ − 1F )Eii )−1 = In + (λ−1 − 1F )Eii .

Proof: Expanding the product

(In + (λ − 1F )Eii )(In + (λ−1 − 1F )Eii ) = In + (λ − 1F )Eii

+(λ−1 − 1F )Eii

+(λ − 1F )(λ−1 − 1F )Eii

= In + (λ − 1F )Eii

+(λ−1 − 1F )Eii

+(λ − 1F )(λ−1 − 1F ))Eii

= In + (λ − 1F + λ−1 − 1F + 1F

−λ − λ−1 − 1F ))Eii

= In ,

proving the assertion. ❑

203 Example By Theorem 202, we have

2 32 3 2 3
61 0 07 6 1 0 07 61 0 07
6 76 7 6 7
6 76 7 6 7
6 76 7 6 7
60 2 07 6 0 1
07 = 6 0 1 07 .
6 76 2 7 6 7
4 54 5 4 5
0 0 1 0 0 1 0 0 1
Matrix Inversion 35

Repeated applications of Theorem 202 gives the following corollary.

204 Corollary If λ1 λ2 λ3 · · · λn 6= 0F , then

2 3
6λ1 0 0 0 ··· 07
6 7
6 7
6 7
60 λ2 0 0 ··· 0 7
6 7
6 7
6 7
60 0 λ3 0 ··· 0 7
6 7
6 7
6 . .. .. .. .. 7
6 .. . . . ··· . 7
6 7
6 7
4 5
0 0 0 0 ··· λn

is invertible and 2 3−1 2 3

6 λ1 0 0 0 ··· 07 6λ−1 0 0 0 ··· 0 7
6 7 6 1 7
6 7 6 7
6 7 6 7
60 λ2 0 0 ··· 0 7 6 0 λ−1 0 0 ··· 0 7
6 7 6 2 7
6 7 6 7
6 7 6 7
60 0 λ3 0 ··· 0 7 =6 0 0 λ−1 0 ··· 0 7
6 7 6 3 7
6 7 6 7
6 . .. .. .. .. 7 6 .. .. .. .. .. 7
6 .. . . . ··· . 7 6 . . . . ··· . 7
6 7 6 7
6 7 6 7
4 5 4 5
0 0 0 0 ··· λn 0 0 0 0 ··· λ−1
n

205 Theorem (Invertibility of Permutation Matrices) Let τ ∈ Sn be a permutation. Then

−1
(Iij
n) = (Iij T
n) .

Proof: By Theorem 185 pre-multiplication of Iij n by In exchanges the i-th row with the j-th row,
ij

meaning that they return to the original position in In . Observe in particular that Iij ij T
n = (In ) , and
so In (In ) = In . ❑
ij ij T

206 Example By Theorem 205, we have

2 32 3 2 3
61 0 07 61 0 07 61 0 07
6 76 7 6 7
6 76 7 6 7
6 76 7 6 7
60 0 17 60 0 17 = 60 1 07 .
6 76 7 6 7
4 54 5 4 5
0 1 0 0 1 0 0 0 1

207 Corollary If a square matrix can be represented as the product of elimination matrices of the same size, then it
is invertible.

Proof: This follows from Corollary 197, and Theorems 200, 202, and 205. ❑

208 Example Observe that

2 3
61 0 07
6 7
6 7
A=6
60 3
7
47
6 7
4 5
0 0 1
36 Chapter 2

is the transvection I3 + 4E23 followed by the dilatation of the second column of this transvection by 3. Thus

2 3 2 32 3
61 0 07 61 0 0 7 61 0 07
6 7 6 76 7
6 7 6 76 7
6 7 6 76 7
60 3 47 = 60 1 4 7 60 3 07 ,
6 7 6 76 7
4 5 4 54 5
0 0 1 0 0 1 0 0 1

and so

2 3−1 2 3−1 2 3−1

61 0 07 61 0 07 61 0 07
6 7 6 7 6 7
6 7 6 7 6 7
6 7 = 6 7 6 7
60 3 47 60 3 07 60 1 47
6 7 6 7 6 7
4 5 4 5 4 5
0 0 1 0 0 1 0 0 1
2 32 3
61 0 07 6 1 0 0 7
6 76 7
6 76 7
= 6 1 76 7
60 07 6 0 1 −47
6 3 76 7
4 54 5
0 0 1 0 0 1
2 3
61 0 0 7
6 7
6 7
= 6 7
60
1
− 43 7 .
6 3 7
4 5
0 0 1

209 Example We have

2 3 2 32 3
61 1 17 61 1 0 7 61 0 07
6 7 6 76 7
6 7 6 76 7
6 7 6 76 7
60 1 17 = 60 1 0 7 60 1 17 ,
6 7 6 76 7
4 5 4 54 5
0 0 1 0 0 1 0 0 1
Matrix Inversion 37

hence
2 3−1 2 3−1 2 3−1
61 1 17 61 0 07 61 1 07
6 7 6 7 6 7
6 7 6 7 6 7
6 7 = 6 7 6 7
60 1 17 60 1 17 60 1 07
6 7 6 7 6 7
4 5 4 5 4 5
0 0 1 0 0 1 0 0 1
2 32 3
61 0 0 7 61 −1 07
6 76 7
6 76 7
= 6 76 7
60 1 −17 60 1 07
6 76 7
4 54 5
0 0 1 0 0 1
2 3
61 −1 0 7
6 7
6 7
6
= 60 1 −17
7.
6 7
4 5
0 0 1

In the next section we will give a general method that will permit us to find the inverse of a square matrix when
it exists.
2 3
6a b7
210 Example Let T = 6
4
7 ∈ M2 (R). Then
5
c d

2 32 3 2 3
6a b7 6 d −b7 61 07
6 76 7 = (ad − bc) 6 7
4 54 5 4 5
c d −c a 0 1

Thus if ad − bc 6= 0 we see that

2 3
d b
6 ad−bc
− ad−bc 7
T −1 = 6
4
7.
5
c a
− ad−bc ad−bc

211 Example If
2 3
61 1 1 1 7
6 7
6 7
61 1 −1 −17
6 7
A=6
6
7,
7
6 7
61 −1 1 −17
6 7
4 5
1 −1 −1 1
38 Chapter 2

then A is invertible, for an easy computation shews that

2 32
61 1 1 1 7
6 7
6 7
61 1 −1 −17
6 7
A2 = 6
6
7 = 4I4 ,
7
6 7
61 −1 1 −17
6 7
4 5
1 −1 −1 1

whence the inverse sought is

2 3 2 3
61 1 1 1 761/4 1/4 1/4 1/4 7
6 7 6 7
6 7 6 7
6 −1 −17 6 −1/4 −1/47
1 1 61 1 7 61/4 1/4 7
A−1 = A= 6 7=6 7.
4 466
7 6
7 6
7
7
61 −1 1 −17 61/4 −1/4 1/4 −1/47
6 7 6 7
4 5 4 5
1 −1 −1 1 1/4 −1/4 −1/4 1/4

212 Example A matrix A ∈ Mn (R) is said to be nilpotent of index k if satisfies A 6= 0n , A2 6= 0n , . . . , Ak−1 6= 0n

and Ak = 0n for integer k ≥ 1. Prove that if A is nilpotent, then In − A is invertible and find its inverse.

Solution: To motivate the solution, think that instead of a matrix, we had a real number x with |x| < 1. Then the
inverse of 1 − x is
1
(1 − x)−1 = = 1 + x + x2 + x3 + · · · .
1−x
Notice now that since Ak = 0n , then Ap = 0n for p ≥ k. We conjecture thus that
(In − A)−1 = In + A + A2 + · · · + Ak−1 .
The conjecture is easily verified, as

(In − A)(In + A + A2 + · · · + Ak−1 ) = In + A + A2 + · · · + Ak−1

−(A + A2 + A3 + · · · + Ak )

= In

and
(In + A + A2 + · · · + Ak−1 )(In − A) = In − A + A − A2 + A3 − A4 + · · ·

· · · + Ak−2 − Ak−1 + Ak−1 − Ak

= In .

213 Example The inverse of A ∈ M3 (Z5 ),

2 3
62 0 07
6 7
6 7
A=6
60 3
7
07
6 7
4 5
0 0 4
Matrix Inversion 39

is 2 3
63 0 07
6 7
6 7
A−1 =6
60 2
7
07 ,
6 7
4 5
0 0 4
as 2 32 3 2 3
62 0 07 63 0 07 61 0 07
6 76 7 6 7
6 76 7 6 7
AA−1 =6
60 3
76
07 60 2
7 6
07 = 60 1
7
07
6 76 7 6 7
4 54 5 4 5
0 0 4 0 0 4 0 0 1

214 Example (Putnam Exam, 1991) Let A and B be different n × n matrices with real entries. If A3 = B3 and
A2 B = B2 A, prove that A2 + B2 is not invertible.

Solution: Observe that

(A2 + B2 )(A − B) = A3 − A2 B + B2 A − B3 = 0n .
If A2 + B2 were invertible, then we would have

A − B = (A2 + B2 )−1 (A2 + B2 )(A − B) = 0n ,

contradicting the fact that A and B are different matrices.

215 Lemma If A ∈ Mn (F) has a row or a column consisting all of 0F ’s, then A is singular.

Proof: If A were invertible, the (i, i)-th Pnentry of the product of its inverse with A would be 1F .
But if the i-th row of A is all 0F ’s, then k=1 aik bki = 0F , so the (i, i) entry of any matrix product
with A is 0F , and never 1F . ❑
2 3
k a positive integer. Prove that if B = SAS−1 then Bk =
61 1 17
SAk S−1 .
6 7
6 7
216 Problem The inverse of the matrix A = 61 1 27 is
6 7
4 5 220 Problem Let A ∈ Mn (F) and let k be a positive integer.
1 2 3 Prove that A is invertible if and only if Ak is invertible.
2 3
6a 1 −17 221 Problem Let S ∈ GLn (C), A ∈ Mn (C) with Ak = 0n
6 7 for some positive integer k. Prove that both In − SAS−1 and
6 7
the matrix A−1 =6 1 b 1 7 . Determine a and b. In − S−1 AS are invertible and find their inverses.
6 7
4 5
−1 1 0 222 Problem Let A and B be square matrices of the same
size such that both A − B and A + B are invertible. Put
C = (A − B)−1 + (A + B)−1 . Prove that
217 Problem A square matrix A satisfies A3 6= 0n but
A4 = 0n . Demonstrate that In + A is invertible and find, ACA − ACB + BCA − BCB = 2A.
with proof, its inverse.
223 Problem Let A, B, C be non-zero square matrices of the
same size over the same field and such that ABC = 0n . Prove
218 Problem Prove or disprove! If (A, B, A + B) ∈ that at least two of these three matrices are not invertible.
(GLn (R))3 then (A + B)−1 = A−1 + B−1 .
224 Problem Let (A, B) ∈ (Mn (F))2 be such that A2 =
219 Problem Let S ∈ GLn (F), (A, B) ∈ (Mn (F)) , and 2 B2 = (AB)2 = In . Prove that AB = BA.
40 Chapter 2

2 3
A + B = AB. Demonstrate that A − In is invertible and find
6a b b ··· b7
6 7 this inverse.
6 7
6b a b ··· b7
6 7
6 7 227 Problem Let S ∈

GLn (F) and A ∈ Mn (F). Prove that
6 7 tr (A) = tr SAS−1 .
225 Problem Let A = 6b b a ··· b7 ∈ Mn (F),
6 7
6 7
6 .. .. .. .. 7
6. . . ··· .7 228 Problem Let A ∈ Mn (R) be a skew-symmetric ma-
6 7
4 5 trix. Prove that In + A is invertible. Furthermore, if
b b b ··· a B = (In − A)(In + A)−1 , prove that B−1 = BT .
2
n > 1, (a, b) ∈ F . Determine when A is invertible and
find this inverse when it exists. 229 Problem A matrix A ∈ Mn (F) is said to be a magic
square if the sum of each individual row equals the sum of
each individual column. Assume that A is a magic square
226 Problem Let (A, B) ∈ (Mn (F))2 be matrices such that and invertible. Prove that A−1 is also a magic square.

2.6 Block Matrices

230 Definition Let A ∈ Mm×n (F), B ∈ Mm×s (F), C ∈ Mr×n (F), D ∈ Mr×s (F). We use the notation
2 3
6A B7
L=6
4
7
5
C D

for the block matrix L ∈ M(m+r)×(n+s) (F).

☞ If (A, A ′ ) ∈ (Mm (F))2 , (B, B ′ ) ∈ (Mm×n (F))2 , (C, C ′ ) ∈ (Mn×m (F))2 , (D, D ′ ) ∈
(Mm (F))2 , and
2 3 2 3
′ ′
6A B7 6A B 7
S=6
4
7,
5 T =6
4
7,
5
C D C′ D′

then it is easy to verify that

2 3
′ ′ ′ ′
6 AA + BC AB + BD 7
ST = 6
4
7.
5
CA ′ + DC ′ CB ′ + DD ′

231 Lemma Let L ∈ M(m+r)×(m+r) (F) be the square block matrix

2 3
6 A C7
L=6
4
7,
5
0r×m B

with square matrices A ∈ Mm (F) and B ∈ Mr (F), and a matrix C ∈ Mm×r (F). Then L is invertible if and only
if A and B are, in which case
2 3
−1 −1 −1
6A −A CB 7
L−1 = 6
4
7
5
0r×m B−1
Rank of a Matrix 41

Proof: Assume first that A, and B are invertible. Direct calculation yields
2 32 3 2 3
−1 −1 −1 −1 −1 −1 −1
6 A C7 6 A −A CB 7 6AA −AA CB + CB 7
6 76 7 = 6 7
4 54 5 4 5
0r×m B 0r×m B−1 0r×m BB−1
2 3
6 Im 0m×r 7
= 6
4
7
5
0r×m Ir

= Im+r .

2 3
6E H7
Assume now that L is invertible, L−1 = 6
4
7, with E ∈ Mm (F) and K ∈ Mr (F), but that, say,
5
J K
B is singular. Then
2 3
6 Im 0m×r 7
6 7 = LL−1
4 5
0r×m Ir
2 32 3
6 A C 7 6E H 7
= 6
4
76
54
7
5
0r×m B J K
2 3
6AE + CJ AH + BK7
= 6
4
7,
5
BJ BK

which gives BK = Ir , i.e., B is invertible, a contradiction. ❑

2.7 Rank of a Matrix

232 Definition Let (A, B) ∈ (Mm×n (F))2 . We say that A is row-equivalent to B if there exists a matrix R ∈
GLm (F) such that B = RA. Similarly, we say that A is column-equivalent to B if there exists a matrix C ∈ GLm (F)
such that B = AC. We say that A and B are equivalent if ∃(P, Q) ∈ GLm (F) × GLn (F) such that B = PAQ.

233 Theorem Row equivalence, column equivalence, and equivalence are equivalence relations.

Proof: We prove the result for row equivalence. The result for column equivalence, and equiva-
lence are analogously proved.

Since Im ∈ GLm (F) and A = Im A, row equivalence is a reflexive relation. Assume (A, B) ∈
(Mm×n (F))2 and that ∃P ∈ GLm (F) such that B = PA. Then A = P−1 B and since P−1 ∈
GLm (F), we see that row equivalence is a symmetric relation. Finally assume (A, B, C) ∈
(Mm×n (F))3 and that ∃P ∈ GLm (F), ∃P ′ ∈ GLm (F) such that A = PB, B = P ′ C. Then
A = PP ′ C. But PP ′ ∈ GLm (F) in view of Corollary 197. This completes the proof. ❑

234 Theorem Let A ∈ Mm×n (F). Then A can be reduced, by means of pre-multiplication and post-multiplication
42 Chapter 2

by elimination matrices, to a unique matrix of the form

2 3
6 Ir 0r×(n−r) 7
Dm,n,r = 6
4
7,
5 (2.16)
0(m−r)×r 0(m−r)×(n−r)

called the Hermite normal form of A. Thus there exist P ∈ GLm (F), Q ∈ GLn (F) such that Dm,n,r = PAQ.
The integer r ≥ 0 is called the rank of the matrix A which we denote by rank (A).

Proof: If A is the m × n zero matrix, then the theorem is obvious, taking r = 0. Assume hence
that A is not the zero matrix. We proceed as follows using the Gauß-Jordan Algorithm.

GJ-1 Since A is a non-zero matrix, it has a non-zero column. By means of permutation matrices
we move this column to the first column.
GJ-2 Since this column is a non-zero column, it must have an entry a 6= 0F . Again, by means of
permutation matrices, we move the row on which this entry is to the first row.
GJ-3 By means of a dilatation matrix with scale factor a−1 , we make this new (1, 1) entry into a
1F .
GJ-4 By means of transvections (adding various multiples of row 1 to the other rows) we now
annihilate every entry below the entry (1, 1).

This process ends up in a matrix of the form

2 3
61F ∗ ∗ ··· ∗ 7
6 7
6 7
6 7
60F b22 b23 ··· b2n 7
6 7
6 7
6 7
P1 AQ1 = 60F b32 b33 ··· b3n 7 . (2.17)
6 7
6 7
6 .. .. .. 7
60 . . ··· . 7
6 F 7
6 7
4 5
0F bm2 bm3 ··· bmn

Here the asterisks represent unknown entries. Observe that the b’s form a (m − 1) × (n − 1)
matrix.

GJ-5 Apply GJ-1 through GJ-4 to the matrix of the b’s.

Observe that this results in a matrix of the form

2 3
61F ∗ ∗ ··· ∗ 7
6 7
6 7
6 7
60F 1F ∗ ··· ∗ 7
6 7
6 7
6 7
P2 AQ2 = 60F 0F c33 ··· c3n 7 . (2.18)
6 7
6 7
6 .. .. .. 7
60 . . ··· . 7
6 F 7
6 7
4 5
0F 0F cm3 ··· cmn

GJ-6 Add the appropriate multiple of column 1 to column 2, that is, apply a transvection, in order
to make the entry in the (1, 2) position 0F .
Rank of a Matrix 43

This now gives a matrix of the form

2 3
61F 0F ∗ ··· ∗ 7
6 7
6 7
6 7
60F 1F ∗ ··· ∗ 7
6 7
6 7
6 7
P3 AQ3 = 60F 0F c33 ··· c3n 7 . (2.19)
6 7
6 7
6 .. .. .. 7
60 . . ··· . 7
6 F 7
6 7
4 5
0F 0F cm3 ··· cmn

The matrix of the c’s has size (m − 2) × (n − 2).

GJ-7 Apply GJ-1 through GJ-6 to the matrix of the c’s, etc.
Observe that this process eventually stops, and in fact, it is clear that rank (A) ≤ min(m, n).

Suppose now that A were equivalent to a matrix Dm,n,s with s > r. Since matrix equivalence is an
equivalence relation, Dm,n,s and Dm,n,r would be equivalent, and so there would be R ∈ GLm (F),
S ∈ GLn (F), such that RDm,n,r S = Dm,n,s , that is, RDm,n,r = Dm,n,s S−1 . Partition R and S−1
as follows
2 3
2 3
6S11 S12 S13 7
6 7
6R11 R12 7 6 7
R=6
4
7,
5 S −1
=6
6S21 S22
7
S23 7 ,
6 7
R21 R22 4 5
S31 S32 S33

with (R11 , S11 )2 ∈ (Mr (F))2 , S22 ∈ M(s−r)×(s−r) (F). We have

2 32 3 2 3
6R11 R12 7 6 Ir 0(m−r)×r 7 6R11 0(m−r)×r 7
RDm,n,r = 6
4
76
54
7=6
5 4
7,
5
R21 R22 0(m−r)×r 0r×(m−r) R21 0r×(m−r)

and 2 32 3
6 Ir 0r×(s−r) 0r×(n−s) 7 6S11 S12 S13 7
6 76 7
6 76 7
Dm,n,s S −1
= 6 76 7
6 0(s−r)×r Is−r 0(s−r)×(n−s) 7 6S21 S22 S23 7
6 76 7
4 54 5
0(m−s)×r 0(m−s)×(s−r) 0(m−s)×(n−s) S31 S32 S33
2 3
6 S11 S12 S13 7
6 7
6 7
= 6 7.
6 S21 S22 S23 7
6 7
4 5
0(m−s)×r 0(m−s)×(s−r) 0(m−s)×(n−s)
Since we are assuming
2 3
2 3 S11 S12 S13
6 7
0(m−r)×r 7 6 7
6R11 6 7
6 7=6 7,
4 5 6 S21 S22 S23 7
6 7
R21 0r×(m−r) 4 5
0(m−s)×r 0(m−s)×(s−r) 0(m−s)×(n−s)
44 Chapter 2

we must have S12 = 0r×(s−r) , S13 = 0r×(n−s) , S22 = 0(s−r)×(s−r) , S23 = 0(s−r)×(n−s) . Hence
2 3
6S11 0r×(s−r) 0r×(n−s) 7
6 7
6 7
S −1
=6
6S21 0(s−r)×(s−r)
7
0(s−r)×(n−s) 7 .
6 7
4 5
S31 S32 S33

The matrix 2 3
60(s−r)×(s−r) 0(s−r)×(n−s) 7
6 7
4 5
S32 S33

is non-invertible, by virtue of Lemma 215. This entails that S−1 is non-invertible by virtue of
Lemma 231. This is a contradiction, since S is assumed invertible, and hence S−1 must also be
invertible. ❑

☞ Albeit the rank of a matrix is unique, the matrices P and Q appearing in Theorem 234 are
not necessarily unique. For example, the matrix
2 3
61 07
6 7
6 7
6 7
60 17
6 7
4 5
0 0

has rank 2, the matrix

2 3
61 0 x7
6 7
6 7
6 7
60 1 y7
6 7
4 5
0 0 1

is invertible, and an easy computation shews that

2 32 3 2 3
61 0 x 7 61 07 2 3
6 1 07
6 76 7 1 07 6 7
6 76 76 6 7
6 76 76 7=6 7
60 1 y 7 60 17 4 5 60 17 ,
6 76 7 6 7
4 54 5 0 1 4 5
0 0 1 0 0 0 0

regardless of the values of x and y.

235 Corollary Let A ∈ Mm×n (F). Then rank (A) = rank AT .

Proof: Let P, Q, Dm,n,r as in Theorem 234. Observe that PT , QT are invertible. Then

PAQ = Dm,n,r =⇒ QT AT PT = DTm,n,r = Dn,m,r ,

and since this last matrix has the same number of 1F ’s as Dm,n,r , the corollary is proven. ❑
Rank of a Matrix 45

236 Example Shew that

2 3
60 2 37
A=6
4
7
5
0 1 0

has rank (A) = 2 and find invertible matrices P ∈ GL2 (R) and Q ∈ GL3 (R) such that
2 3
61 0 07
PAQ = 6
4
7.
5
0 1 0

Solution: We first transpose the first and third columns by effecting

2 3
2 3 0 0 17 2 3
6
60 2 37 6
6
7
7 63 2 07
6 76 7 6 7.
4 5 60 1 07 = 4 5
6 7
0 1 0 4 5 0 1 0
1 0 0

We now subtract twice the second row from the first, by effecting
2 32 3 2 3
61 −27 63 2 07 63 0 07
6 76 7=6 7.
4 54 5 4 5
0 1 0 1 0 0 1 0

Finally, we divide the first row by 3,

2 32 3 2 3
61/3 0 7 63 0 07 61 0 07
6 76 7=6 7.
4 54 5 4 5
0 1 0 1 0 0 1 0

We conclude that 2 3
2 32 32 3 0 0 17 2 3
6
61/3 07 61 −27 60 2 37 6
6
7
7 61 0 07
6 76 76 76
4 54 54 5 60 1 07 6
7=4
7,
5
6 7
0 1 0 1 0 1 0 4 5 0 1 0
1 0 0

from where we may take

2 32 3 2 3
61/3 07 61 −27 61/3 −2/37
P=6
4
76
54
7=6
5 4
7
5
0 1 0 1 0 1

and
2 3
60 0 17
6 7
6 7
Q=6
60 1
7
07 .
6 7
4 5
1 0 0
46 Chapter 2

In practice it is easier to do away with the multiplication by elimination matrices and perform row and column
operations on the augmented (m + n) × (m + n) matrix
2 3
6In 0n×m 7
6 7.
4 5
A Im

237 Definition Denote the rows of a matrix A ∈ Mm×n (F) by R1 , R2 , . . . , Rm , and its columns by C1 , C2 , . . . , Cn .
The elimination operations will be denoted as follows.
 Exchanging the i-th row with the j-th row, which we denote by Ri ↔ Rj , and the s-th column by the t-th
column by Cs ↔ Ct .
 A dilatation of the i-th row by a non-zero scalar α ∈ F \ {0F }, we will denote by αRi → Ri . Similarly,
βCj → Cj denotes the dilatation of the j-th column by the non-zero scalar β.
 A transvection on the rows will be denoted by Ri + αRj → Ri , and one on the columns by Cs + βCt → Cs .

238 Example Find the Hermite normal form of

2 3
6−1 07
6 7
6 7
6 0 07
6 7
A=6
6
7.
7
6 7
6 1 17
6 7
4 5
1 2

Solution: First observe that rank (A) ≤ min(4, 2) = 2, so the rank can be either 1 or 2 (why not 0?). Form the
augmented matrix 2 3
6 1 0 0 0 0 07
6 7
6 7
6 7
6 0 1 0 0 0 07
6 7
6 7
6 7
6−1 0 1 0 0 07
6 7
6 7.
6 7
6 0 0 0 1 0 07
6 7
6 7
6 7
6 1 1 0 0 1 07
6 7
6 7
4 5
1 2 0 0 0 1
Perform R5 + R3 → R5 and R6 + R3 → R6 successively, obtaining
2 3
6 1 0 0 0 0 07
6 7
6 7
6 7
6 0 1 0 0 0 07
6 7
6 7
6 7
6−1 0 1 0 0 07
6 7
6 7.
6 7
6 0 0 0 1 0 07
6 7
6 7
6 7
6 0 1 1 0 1 07
6 7
6 7
4 5
0 2 1 0 0 1
Rank of a Matrix 47

Perform R6 − 2R5 → R6
2 3
6 1 0 0 0 0 07
6 7
6 7
6 7
6 0 1 0 0 0 07
6 7
6 7
6 7
6−1 0 1 0 0 07
6 7
6 7.
6 7
6 0 0 0 1 0 07
6 7
6 7
6 7
6 0 1 1 0 1 07
6 7
6 7
4 5
0 0 −1 0 −2 1

Perform R4 ↔ R5
2 3
6 1 0 0 0 0 07
6 7
6 7
6 7
6 0 1 0 0 0 07
6 7
6 7
6 7
6−1 0 1 0 0 07
6 7
6 7.
6 7
6 0 1 1 0 1 07
6 7
6 7
6 7
6 0 0 0 1 0 07
6 7
6 7
4 5
0 0 −1 0 −2 1

Finally, perform −R3 → R3

2 3
61 0 0 0 0 07
6 7
6 7
6 7
60 1 0 0 0 07
6 7
6 7
6 7
61 0 −1 0 0 07
6 7
6 7.
6 7
60 1 1 0 1 07
6 7
6 7
6 7
60 0 0 1 0 07
6 7
6 7
4 5
0 0 −1 0 −2 1

We conclude that
2 32 3 2 3
6−1 0 0 07 6−1 07 61 07
6 76 72 3 6 7
6 76 7 6 7
6 1 0 1 07 6 07 07 6 17
6 76 0 7 61 60 7
6 76 76 7=6 7.
6 76 74 5 6 7
6 76 7 6 7
6 0 1 0 07 6 1 17 0 1 60 07
6 76 7 6 7
4 54 5 4 5
−1 0 −2 1 1 2 0 0

239 Theorem Let A ∈ Mm×n (F), B ∈ Mn×p (F). Then

rank (AB) ≤ min(rank (A) , rank (B)).

48 Chapter 2

Proof: We prove that rank (A) ≥ rank (AB). The proof that rank (B) ≥ rank (AB) is similar
and left to the reader. Put r = rank (A) , s = rank (AB). There exist matrices P ∈ GLm (F),
Q ∈ GLn (F), S ∈ GLm (F), T ∈ GLp (F) such that

PAQ = Dm,n,r , SABT = Dm,p,s .

Now

Dm,p,s = SABT = SP−1 Dm,n,r Q−1 BT,

from where it follows that

PS−1 Dm,p,s = Dm,n,r Q−1 BT.

Now the proof is analogous to the uniqueness proof of Theorem 234. Put U = PS−1 ∈ GLm (R)
and V = Q−1 BT ∈ Mn×p (F), and partition U and V as follows:
2 3 2 3
6U11 U12 7 6V11 V12 7
U=6
4
7,
5 V=6
4
7,
5
U21 U22 V21 V22

with U11 ∈ Ms (F), V11 ∈ Mr (F). Then

2 32 3
6U11 U12 7 6 Is 0s×(p−s) 7
UDm,p,s = 6
4
76
54
7 ∈ Mm×p (F),
5
U21 U22 0(m−s)×s 0(m−s)×(p−s)

and 2 32 3
6 Ir 0r×(n−r) 7 6V11 V12 7
Dm,p,s V = 6
4
76
54
7 ∈ Mm×p(F).
5
0(m−r)×r 0(m−r)×(n−r) V21 V22

From the equality of these two m × p matrices, it follows that

2 3 2 3
6U11 0s×(p−s) 7 6 V11 V12 7
6 7=6 7.
4 5 4 5
U21 0(m−s)×(p−s) 0(m−r)×r 0(m−r)×(n−r)

If s > r then (i) U11 would have at least one row of 0F ’s meaning that U11 is non-invertible
by Lemma 215. (ii) U21 = 0(m−s)×s . Thus from (i) and (ii) and from Lemma 231, U is not
invertible, which is a contradiction. ❑

240 Corollary Let A ∈ Mm×n (F), B ∈ Mn×p (F). If A is invertible then rank (AB) = rank (B). If B is invertible
then rank (AB) = rank (A).

Proof: Using Theorem 239, if A is invertible

rank (AB) ≤ rank (B) = rank A−1 AB ≤ rank (AB) ,

and so rank (B) = rank (AB). A similar argument works when B is invertible.
❑
Rank of a Matrix 49

241 Example Study the various possibilities for the rank of the matrix
2 3
6 1 1 1 7
6 7
6 7
A=6
6b + c c+a
7
a + b7 .
6 7
4 5
bc ca ab

Solution: Performing R2 − (b + c)R1 → R2 and R3 − bcR1 → R3 , we find

2 3
61 1 1 7
6 7
6 7
6 7.
60 a−b a−c 7
6 7
4 5
0 0 (b − c)(a − c)

Performing C2 − C1 → C2 and C3 − C1 → C3 , we find

2 3
61 0 0 7
6 7
6 7
6 7.
60 a−b a−c 7
6 7
4 5
0 0 (b − c)(a − c)

We now examine the various ways of getting rows consisting only of 0’s. If a = b = c, the last two rows are 0-rows
and so rank (A) = 1. If exactly two of a, b, c are equal, the last row is a 0-row, but the middle one is not, and so
rank (A) = 2 in this case. If none of a, b, c are equal, then the rank is clearly 3.

242 Problem On a symmetric matrix A ∈ Mn (R) with the matrix

n ≥ 3, 2 3
R3 − 3R1 → R3
61 a 1 b7
6 7
successively followed by 6 7
6a 1 b 17
6 7
6 7
C3 − 3C1 → C3 6 7
61 b 1 a7
6 7
4 5
are performed. Is the resulting matrix still symmetric?
b 1 a 1

243 Problem Find the rank of

when (a, b) ∈ R2 .
2 3
6a + 1 a+2 a+3 a+4 a + 57
6 7 2 3
6 7
6a + 2 a + 67 −1
6 a+3 a+4 a+5 7 61 0 1 7
6 7 ∈ M5 (R). 6 7
6 7 6 7
6a + 3 a+4 a+5 a+6 a + 77 6m 1 −1 −17
6 7 6 7
4 5 246 Problem Find the rank of 6 7 as a
6 7
a+4 a+5 a+6 a+7 a+8 61 −m 1 0 7
6 7
4 5
1 −1 m 2
244 Problem Let A, B be arbitrary n × n matrices over R. function of m ∈ C.
Prove or disprove! rank (AB) = rank (BA) .

245 Problem Study the various possibilities for the rank of 247 Problem Determine the rank of the matrix
50 Chapter 2

2 3 2 3
2
6a ab ab b2 7 6 1 1 27
6 7 6 7
6 7 6 7
6ab a2 b2 ab7 M2×3 (R) be such that ABC = 6−2 x 17 . Find x.
6 7 6 7
6 7. 4 5
6 7
6ab b2 a2 ab7 1 −2 1
6 7
4 5
b2 ab ab a2
250 Problem Let A, B be matrices of the same size. Prove
that rank (A + B) ≤ rank (A) + rank (B).
248
2 Problem Determine
3 the rank of the matrix
251 Problem Let B be the matrix obtained by adjoining a
61 1 1 1 7
row (or column) to a matrix A. Prove that either rank (B) =
6 7
6 7 rank (A) or rank (B) = rank (A) + 1.
6a b a b7
6 7
6 7.
6 7
6c c d d7 252 Problem
Let A ∈ Mn (R). Prove that rank (A) =
6 7 rank AAT . Find a counterexample in the case A ∈
4 5
ac bc ad bd Mn (C).

253 Problem Prove that the rank of a skew-symmetric matrix

249 Problem Let A ∈ M3×2 (R), B ∈ M2 (R), and C ∈ is an even number.

2.8 Rank and Invertibility

254 Theorem A matrix A ∈ Mm×n (F) is left-invertible if and only if rank (A) = n. A matrix A ∈ Mm×n (F) is
right-invertible if and only if rank (A) = m.

Proof: Observe that we always have rank (A) ≤ n. If A is left invertible, then ∃L ∈ Mn×m (F)
such that LA = In . By Theorem 239,

n = rank (In ) = rank (LA) ≤ rank (A) ,

whence the two inequalities give rank (A) = n.

Conversely, assume that rank (A) = n. Then rank AT = n by Corollary 235, and so by Theorem
234 there exist P ∈ GLm (F), Q ∈ GLn (F), such that
2 3
In  
6 7
PAQ = 6
4
7,
5 QT AT PT = In 0n×(m−n) .
0(m−n)×n

This gives

QT AT PT PAQ = In =⇒ AT PT PA = (QT )−1 Q−1

=⇒ ((QT )−1 Q−1 )−1 AT PT PA = In ,

and so ((QT )−1 Q−1 )−1 AT PT P is a left inverse for A.

The right-invertibility case is argued similarly. ❑

By combining Theorem 254 and Theorem 195, the following corollary is thus immediate.

255 Corollary If A ∈ Mm×n (F) possesses a left inverse L and a right inverse R then m = n and L = R.
Rank and Invertibility 51

We use Gauß-Jordan Reduction to find the inverse of A ∈ GLn (F). We form the augmented matrix T = [A|In ]
which is obtained by putting A side by side with the identity matrix In . We perform permissible row operations
on T until instead of A we obtain In , which will appear if the matrix is invertible. The matrix on the right will be
A−1 . We finish with [In |A−1 ].

☞ If A ∈ Mn (R) is non-invertible, then the left hand side in the procedure above will not reduce
to In .

256 Example Find the inverse of the matrix B ∈ M3 (Z7 ),

2 3
66 0 17
6 7
6 7
B=6
63 2
7
07 .
6 7
4 5
1 0 1

Solution: We have
2 3 2 3
66 0 1 1 0 07 61 0 1 0 0 17
6 7 6 7
6 7 R1 ↔ R3 6 7
6 7 6 7
63 2 0 0 1 07 63 2 0 0 1 07
6 7 6 7
4 5 4 5
1 0 1 0 0 1 6 0 1 1 0 0
2 3
61 0 1 0 0 17
6 7
R3 −6R1 → R3 6 7
6 7
60 2 4 0 1 47
R2 −3R1 → R2 6 7
4 5
0 0 2 1 0 1
2 3
65 0 0 1 0 67
6 7
R2 −2R3 → R2 6 7
6 7
60 2 0 5 1 27
5R1 +R3 → R1 6 7
4 5
0 0 2 1 0 1
2 3
61 0 0 3 0 47
6 7
3R1 → R1 ; 4R3 → R3 6 7
6
60 1 0 6 4 17
7.
4R2 → R2 6 7
4 5
0 0 1 4 0 4

We conclude that
2 3−1 2 3
66 0 17 63 0 47
6 7 6 7
6 7 6 7
6 7 =6 7
63 2 07 66 4 17 .
6 7 6 7
4 5 4 5
1 0 1 4 0 4
52 Chapter 2

2 3
6 0 1 −1 7
6 7
6 7
6 7
257 Example Use Gauß-Jordan reduction to find the inverse of the matrix A = 6 4 −3 4 7 . Also, find A2001 .
6 7
6 7
4 5
3 −3 4

Solution: Operating on the augmented matrix

2 3 2 3
60 1 −1 1 0 07 60 1 −1 1 0 0 7
6 7 6 7
6 7 R2 −R3 → R2 6 7
6 7 6 7
64 −3 4 0 1 07 61 0 0 0 1 −17
6 7 6 7
4 5 4 5
3 −3 4 0 0 1 3 −3 4 0 0 1
2 3
60 1 −1 1 0 0 7
6 7
R3 −3R2 → R3 6 7
6 7
61 0 0 0 1 −17
6 7
4 5
0 −3 4 0 −3 4
2 3
60 1 −1 1 0 0 7
6 7
R3 +3R1 → R3 6 7
6
61 0 0 0 1 −17
7
6 7
4 5
0 0 1 3 −3 4
2 3
60 1 0 4 −3 4 7
6 7
R1 +R3 → R1 6 7
6 7
61 0 0 0 1 −17
6 7
4 5
0 0 1 3 −3 4
2 3
61 0 0 0 1 −17
6 7
R1 ↔ R2 6 7
6 7
60 1 0 4 −3 4 7.
6 7
4 5
0 0 1 3 −3 4

Thus we deduce that

2 3
60 1 −17
6 7
6 7
A −1
=6
64 −3
7
4 7 = A.
6 7
4 5
3 −3 4

From A−1 = A we deduce A2 = In . Hence A2000 = (A2 )1000 = I1000

n = In and A2001 = A(A2000 ) = AIn = A.
Rank and Invertibility 53

258 Example Find the inverse of the triangular matrix A ∈ Mn (R),

2 3
61 1 1 ··· 17
6 7
6 7
6 7
60 1 1 ··· 17
6 7
6 7
6 7
A = 60 0 1 ··· 17 .
6 7
6 7
6. .. .. .. 7
6 .. . . ··· .7
6 7
6 7
4 5
0 0 0 ··· 1

Solution: Form the augmented matrix

2 3
61 1 1 ··· 1 1 0 0 ··· 07
6 7
6 7
6 7
60 1 1 ··· 1 0 1 0 ··· 07
6 7
6 7
6 7
60 0 1 ··· 1 0 0 1 ··· 07 ,
6 7
6 7
6. .. .. .. .. .. .. .. 7
6 .. . . ··· . . . . ··· .7
6 7
6 7
4 5
0 0 0 ··· 1 0 0 0 ··· 1

and perform Rk − Rk+1 → Rk successively for k = 1, 2, . . . , n − 1, obtaining

2 3
61 0 0 ··· 0 1 −1 0 ··· 07
6 7
6 7
6 7
60 1 0 ··· 0 0 1 −1 ··· 07
6 7
6 7
6 7
60 0 1 ··· 0 0 0 1 ··· 07 ,
6 7
6 7
6. .. .. .. .. .. .. .. 7
6 .. . . ··· . . . . ··· .7
6 7
6 7
4 5
0 0 0 ··· 1 0 0 0 ··· 1

whence 2 3
61 −1 0 ··· 07
6 7
6 7
6 7
60 1 −1 · · · 07
6 7
6 7
6 7
A−1 = 60 0 1 ··· 07 ,
6 7
6 7
6. .. .. .. 7
6 .. . . ··· .7
6 7
6 7
4 5
0 0 0 ··· 1

that is, the inverse of A has 1’s on the diagonal and −1’s on the superdiagonal.

259 Theorem Let A ∈ Mn (F) be a triangular matrix such that a11 a22 · · · ann 6= 0F . Then A is invertible.

Proof: Since the entry akk 6= 0F we multiply the k-th row by a−1kk and then proceed to subtract
the appropriate multiples of the preceding k − 1 rows at each stage. ❑
54 Chapter 2

260 Example (Putnam Exam, 1969) Let A and B be matrices of size 3 × 2 and 2 × 3 respectively. Suppose that
their product AB is given by
2 3
6 8 2 −27
6 7
6 7
AB = 6
6 2 5 4 7
7.
6 7
4 5
−2 4 5

Demonstrate that the product BA is given by

2 3
69 07
BA = 6
4
7.
5
0 9

Solution: Observe that

2 32 3 2 3
6 8 2 −27 6 8 2 −27 6 72 18 −187
6 76 7 6 7
6 76 7 6 7
(AB)2 = 6
6 2 5 4 7 6
76 2 5 4 7 6
7 = 6 18 45 36 7
7 = 9AB.
6 76 7 6 7
4 54 5 4 5
−2 4 5 −2 4 5 −18 36 45

Performing R3 + R2 → R3 , R1 − 4R2 → R2 , and 2R3 + R1 → R3 in succession we see that

2 3 2 3 2 3 2 3
6 8 2 −27 60 −18 −187 60 −18 0 7 60 −18 0 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 2 5 4 7 62 5 4 7 62 5 −17 60 5 −17 ,
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
−2 4 5 0 0 0 0 0 0 0 0 0


and so rank (AB) = 2. This entails that rank (AB)2 = 2. Now, since BA is a 2 × 2 matrix, rank (BA) ≤ 2.
Also

2 = rank (AB)2 = rank (ABAB) ≤ rank (ABA) ≤ rank (BA) ,
and we must conclude that rank (BA) = 2. This means that BA is invertible and so

(AB)2 = 9AB =⇒ A(BA − 9I2 )B = 03

=⇒ BA(BA − 9I2 )BA = B03 A

=⇒ BA(BA − 9I2 )BA = 02

=⇒ (BA)−1 BA(BA − 9I2 )BA(BA)−1 = (BA)−1 02 (BA)−1

=⇒ BA − 9I2 = 02

261 Problem Find the inverse of the matrix

2 3
61 2 37
6 7
6 7
62 3 17 ∈ M3 (Z7 ).
6 7
4 5
3 1 2
Rank and Invertibility 55

262 Problem Let (A, B) ∈ M3 (R) be given by 267 Problem Prove that for the n × n (n > 1) matrix
2 3−1 2 3
2 3 2 3
60 1 1 ... 17 62 − n 1 1 ... 1 7
−17 6 7 6 7
6a b c7 6 0 0 6 7 6 7
6 7 6 7 61 0 1 ... 17 62 − n 1 1 ... 1 7
6 7 6 7 6 7 6 7
A = 61 0 07 , B=6 0 −1 a7 . 6 7 6 7
6 7 6 7 6 7 1 6 7
4 5 4 5 61 1 0 ... 17 = 6 2−n 7
6 7 n−1 6 1 1 ... 1 7
0 1 0 −1 a b 6 7 6 7
6 .. .. .. .. 7 6 .. .. .. .. 7
6. . . ... .7 6 . . . ... . 7
6 7 6 7
4 5 4 5
Find B−1 and prove that AT = BAB−1 . 1 1 1 ... 0 1 1 1 ... 2−n

263 Problem Find all the values of the parameter a for which 268 Problem Prove that the n × n (n > 1) matrix
the matrix B given below is not invertible. 2 3
61 + a 1 1 ... 1 7
2 3 6 7
6 7
6 1 1+a 1 ... 1 7
6 −1 a+2 2 7 6 7
6 7 6 7
6 7 6 7
B=6 0 a 1 7 6 1 1 1+a ... 1 7
6 7 6 7
4 5 6 7
6 .. .. .. .. 7
2 1 a 6 . . . ... . 7
6 7
4 5
1 1 1 ... 1+a
264 Problem Find the inverse of the triangular matrix has inverse
2 3
2 3
61 − n − a 1 1 ... 1 7
6 7
6a 2a 3a7
6 7
6 7 61 − n − a 7
6 7 6 1 1 ... 1 7
60 b 2b7 ∈ M3 (R) 6 7
6 7 1 6 7
4 5 − 6 1 1 1−n−a ... 1 7
a(n + a) 6 7
0 0 c 6 7
6 .. .. .. .. 7
6 . . . ... . 7
6 7
4 5
assuming that abc 6= 0.
1 1 1 ... 1−n−a

265 Problem Under what conditions is the matrix 269 Problem Prove that
2 3
2 3 1 3 5 7 ··· (2n − 1)7
6
6 7
6b a 07 6 7
6 7 6(2n − 1) 1 3 5 ··· (2n − 3)7
6 7 6 7
6c 0 a7 6 7
6 7 6 7
4 5 6(2n − 3) (2n − 1) 1 3 ··· (2n − 5)7
6 7
0 c b 6 7
6 .. .. .. .. .. .. 7
6 . . . . . . 7
6 7
4 5
invertible? Find the inverse under these conditions. 3 5 7 9 ··· 1

has inverse
266 Problem Find the inverse of the matrix 2 3
2 2
62 − n 2+n 2 2 ··· 2 7
2 3 6 7
6 7
6 2 2 − n2 2 + n2 7
61 + a 1 1 7 6 2 ··· 2 7
6 7 6 7
6 7 1 6 7
6 1 1+b 1 7 6 2 − n2 2 + n2 ··· 7.
6 7 2n3 6 2 2 2 7
4 5 6 7
6 .. .. .. .. .. .. 7
1 1 1+c 6 . . . . . . 7
6 7
4 5
2 + n2 2 2 2 ··· 2 − n2
56 Chapter 2

1 1 1
270 Problem Prove that the n × n (n > 1) matrix where s = 1 + a1
+ a2
+··· + an
.
2 3


61 + a1 1 1 ... 1 7 271 Problem Let A ∈ M5 (R). Shew that if rank A2 < 5,
6 7
6 7 then rank (A) < 5.
6 1 1 + a2 1 ... 1 7
6 7
6 7
6 7
6 1 1 1 + a3 ... 1 7 272 Problem Let A ∈ M3,2 (R) and B ∈ M2,3 (R) be matri-
6 7 2 3
6 7
6 .. .. .. .. 7 6 0 −1 −17
6 . . . ... . 7 6 7
6 7 6 7
4 5 ces such that AB = 6−1
6 0 −17
7
. Prove that BA = I2 .
1 1 1 ... 1 + an
4 5
has inverse 1 1 2
2 3
1 − a1 s 1 1 1
6 ...
6 a21 a1 a2 a1 a3 a1 an 7 7
6 1 1−a s 7
6 2 1 1 7
6a a a22
... 7
6 2 1 a2 a3 a2 an 7
16 7
− 6 1 1 1 − a3 s 1 7,
s66 a3 a1 a3 a2 a23
... 7
a3 a1 n 7
6 7
6 .. .. .. .. 7
6 . . . ... . 7
6 7
4 5
1 1 1 1 − an s
...
an a1 an a2 an a3 a2n
Chapter 3
Linear Equations

3.1 Definitions
We can write a system of m linear equations in n variables over a field F

a11 x1 + a12 x2 + a13 x3 + · · · + a1n xn = y1 ,

a21 x1 + a22 x2 + a23 x3 + · · · + a2n xn = y2 ,

..
.
am1 x1 + am2 x2 + am3 x3 + · · · + amn xn = ym ,
in matrix form as 2 32 3 2 3
6 a11 a12 ··· a1n 7 6 x1 7 6 y1 7
6 76 7 6 7
6 76 7 6 7
6a a22 ··· a2n 7 6 x2 7 6 y2 7
7 6 7 6
6 21 7
6 76 7 = 6 7. (3.1)
6 . .. .. .. 7 6 .. 7 6 .. 7
7 6 7 6
6 . 7
6 . . . . 76 . 7 6 . 7
6 76 7 6 7
4 54 5 4 5
am1 am2 ··· amn xn ym

We write the above matrix relation in the abbreviated form

AX = Y, (3.2)

where A is the matrix of coefficients, X is the matrix of variables and Y is the matrix of constants. Most often we
will dispense with the matrix of variables X and will simply write the augmented matrix of the system as
2 3
6 a11 a12 ··· a1n y1 7
6 7
6 7
6a a22 ··· a2n y2 7
6 21 7
[A|Y] = 6
6 .
7. (3.3)
6 . .. .. .. .. 7
7
6 . . . . . 7
6 7
4 5
am1 am2 ··· amn ym

273 Definition Let AX = Y be as in 3.1. If Y = 0m×1 , then the system is called homogeneous, otherwise it is called
inhomogeneous. The set
{X ∈ Mn×1 (F) : AX = 0m×1}
is called the kernel or nullspace of A and it is denoted by ker (A).

57
58 Chapter 3

☞ Observe that we always have 0 n×1 ∈ ker (A) ∈ Mm×n (F).

274 Definition A system of linear equations is consistent if it has a solution. If the system does not have a solution
then we say that it is inconsistent.

275 Definition If a row of a matrix is non-zero, we call the first non-zero entry of this row a pivot for this row.

276 Definition A matrix M ∈ Mm×n (F) is a row-echelon matrix if

 All the zero rows of M, if any, are at the bottom of M.

 For any two consecutive rows Ri and Ri+1 , either Ri+1 is all 0F ’s or the pivot of Ri+1 is immediately to the
right of the pivot of Ri .

The variables accompanying these pivots are called the leading variables. Those variables which are not leading
variables are the free parameters.

277 Example The matrices

2 3 2 3
6 1 0 1 1 7 6 1 0 1 1 7
6 7 6 7
6 7 6 7
6 0 0 2 2 7 6 0 0 0 1 7
6 7 6 7
6 7, 6 7,
6 7 6 7
6 7 6 7
6 0 0 0 3 7 6 0 0 0 0 7
6 7 6 7
4 5 4 5
0 0 0 0 0 0 0 0

are in row-echelon form, with the pivots circled, but the matrices
2 3 2 3
61 0 1 17 61 0 1 17
6 7 6 7
6 7 6 7
60 27 60 07
6 0 1 7 6 0 0 7
6 7, 6 7,
6 7 6 7
6 7 6 7
60 0 1 17 60 0 0 17
6 7 6 7
4 5 4 5
0 0 0 0 0 0 0 0

are not in row-echelon form.

☞ Observe that given a matrix A ∈ M m×n (F), by following Gauß-Jordan reduction à la Theorem
234, we can find a matrix P ∈ GLm (F) such that PA = B is in row-echelon form.

278 Example Solve the system of linear equations

2 32 3 2 3
61 1 1 1 76x7 6−37
6 76 7 6 7
6 76 7 6 7
60 2 1 0 7 6 y 7 6−17
7 6 7 6
6 7
6 76 7 = 6 7.
6 76 7 6 7
6 76 7 6 7
60 0 1 −17 6 z 7 6 4 7
6 76 7 6 7
4 54 5 4 5
0 0 0 2 w −6
Definitions 59

Solution: Observe that the matrix of coefficients is already in row-echelon form. Clearly every variable is a leading
variable, and by back substitution
6
2w = −6 =⇒ w = − = −3,
2

z − w = 4 =⇒ z = 4 + w = 4 − 3 = 1,

1 1
2y + z = −1 =⇒ y = − − z = −1,
2 2

x + y + z + w = −3 =⇒ x = −3 − y − z − w = 0.

The (unique) solution is thus

2 3 2 3
6x7 6 0 7
6 7 6 7
6 7 6 7
6 y 7 6−17
6 7 6 7
6 7=6 7.
6 7 6 7
6 7 6 7
6z7 6 1 7
6 7 6 7
4 5 4 5
w −3

279 Example Solve the system of linear equations

2 3
2 3 2 3
6x7
6 7
1 7 6 7 6−37
61 1 1
6 7 6y7 6 7
6 76 7 6 7
6 76 7 = 6 7
60 2 1 0 7 6 7 6−17 .
6 76 7 6 7
4 56z7 4 5
6 7
0 0 1 −1 4 5 4
w

Solution: The system is already in row-echelon form, and we see that x, y, z are leading variables while w is a free
parameter. We put w = t. Using back substitution, and operating from the bottom up, we find

z − w = 4 =⇒ z = 4 + w = 4 + t,

1 1 1 1 5 1
2y + z = −1 =⇒ y = − − z=− −2− t=− − t,
2 2 2 2 2 2

5 1 9 3
x + y + z + w = −3 =⇒ x = −3 − y − z − w = −3 + + t − 4 − t − t = − − t.
2 2 2 2
The solution is thus
2 3 2 3
9 3
6x7 6− 2 − 2 t 7
6 7 6 7
6 7 6 7
6 y 7 6− 5 − 1 t 7
6 7 6 2 2 7
6 7=6 7 , t ∈ R.
6 7 6 7
6 7 6 7
6z7 6 4 + t 7
6 7 6 7
4 5 4 5
w t
60 Chapter 3

280 Example Solve the system of linear equations

2 3
6x7
2 36 7 2 3
6 7
6 7
17 6 y 7 6−37
61 1 1
6 76 7 = 6 7.
4 56 7 4 5
6 7
0 2 1 0 6z7 −1
6 7
4 5
w

Solution: We see that x, y are leading variables, while z, w are free parameters. We put z = s, w = t. Operating
from the bottom up, we find
1 1 1 1
2y + z = −1 =⇒ y = − − z = − − s,
2 2 2 2
5 3
x + y + z + w = −3 =⇒ x = −3 − y − z − w = − − s − t.
2 2
The solution is thus 2 3 2 3
5 3
6x7 6 − 2 − 2 s − t7
6 7 6 7
6 7 6 7
6y7 6 − 12 − 1 7
6 7 6 2
s 7
6 7=6 7 , (s, t) ∈ R2 .
6 7 6 7
6 7 6 7
6z7 6 s 7
6 7 6 7
4 5 4 5
w t

281 Example Find all the solutions of the system

x + 2y + 2z = 0,

y + 2z = 1,
working in Z3 .

Solution: The augmented matrix of the system is

2 3
61 2 2 07
6 7.
4 5
0 1 2 1

The system is already in row-echelon form and x, y are leading variables while z is a free parameter. We find

y = 1 − 2z = 1 + 1z,

and
x = −2y − 2z = 1 + 2z.
Thus 2 3 2 3
6 x 7 61 + 2z7
6 7 6 7
6 7 6 7
6 7=6 7
6y7 61 + 1z7 , z ∈ Z3 .
6 7 6 7
4 5 4 5
z z
Existence of Solutions 61

Letting z = 0, 1, 2 successively, we find the three solutions

2 3 2 3
6 x 7 617
6 7 6 7
6 7 6 7
6 7 = 6 7,
6y7 617
6 7 6 7
4 5 4 5
z 0
2 3 2 3
6 x7 607
6 7 6 7
6 7 6 7
6 7 = 6 7,
6y7 627
6 7 6 7
4 5 4 5
z 1
and 2 3 2 3
6 x7 627
6 7 6 7
6 7 6 7
6 7 = 6 7.
6y7 607
6 7 6 7
4 5 4 5
z 2
282 Problem Find all the solutions in Z3 of the system find all solutions of the system

x + y + z + w = 0, 1x + 2y + 3z = 5;

2x + 3y + 1z = 6;
2y + w = 2.
3x + 1y + 2z = 0.

283 Problem In Z7 , given that

2 3−1 2 3
61 2 37 64 2 07
6 7 6 7
6 7 6 7
62 3 17 = 62 0 47 ,
6 7 6 7
4 5 4 5
3 1 2 0 4 2

3.2 Existence of Solutions

We now answer the question of deciding when a system of linear equations is solvable.

284 Lemma Let A ∈ Mm×n (F) be in row-echelon form, and let X ∈ Mn×1 (F) be a matrix of variables. The
homogeneous system AX = 0m×1 of m linear equations in n variables has (i) a unique solution if m = n, (ii)
multiple solutions if m < n.

Proof: If m = n then A is a square triangular matrix whose diagonal elements are different
from 0F . As such, it is invertible by virtue of Theorem 259. Thus
AX = 0n×1 =⇒ X = A−1 0n×1 = 0n×1
so there is only the unique solution X = 0n×1 , called the trivial solution.

If m < n then there are n − m free variables. Letting these variables run through the elements of
the field, we obtain multiple solutions. Thus if the field has infinitely many elements, we obtain
infinitely many solutions, and if the field has k elements, we obtain kn−m solutions. Observe that
in this case there is always a non-trivial solution.
❑
62 Chapter 3

285 Theorem Let A ∈ Mm×n (F), and let X ∈ Mn×1 (F) be a matrix of variables. The homogeneous system
AX = 0m×1 of m linear equations in n variables always has a non-trivial solution if m < n.

Proof: We can find a matrix P ∈ GLm (F) such that B = PA is in row-echelon form. Now
AX = 0m×1 ⇐⇒ PAX = 0m×1 ⇐⇒ BX = 0m×1 .
That is, the systems AX = 0m×1 and BX = 0m×1 have the same set of solutions. But by Lemma
284 there is a non-trivial solution. ❑

286 Theorem (Kronecker-Capelli) Let A ∈ Mm×n (F), Y ∈ Mm×1 (F) be constant matrices and X ∈ Mn×1 (F)
be a matrix of variables. The matrix equation AX = Y is solvable if and only if
rank (A) = rank ([A|Y]) .

Proof: Assume first that AX = Y,

2 3
6 x1 7
6 7
6 7
6x 7
6 27
X=6 7
6 . 7.
6 . 7
6 . 7
6 7
4 5
xn
Let the columns of [A|X] be denoted by Ci , 1 ≤ i ≤ n. Observe that that [A|X] ∈ Mm×(n+1) (F)
and that the (n + 1)-th column of [A|X] is
2 3
6 x1 a11 + x2 a12 + · · · + xn a1n 7
6 7
6 7
6x a + x a + ··· + x a 7 n
6 1 21 2 22 n 2n 7 X
Cn+1 = AX = 6
6
7=
7 xi C i .
6 .. 7 i=1
6 . 7
6 7
4 5
x1 an1 + x2 an2 + · · · + xn ann
Pn
By performing Cn+1 − j=1 xj Cj → Cn+1 on [A|Y] = [A|AX] we obtain [A|0n×1 ]. Thus rank ([A|Y]) =
rank ([A|0n×1 ]) = rank (A).

Now assume that r = rank (A) = rank ([A|Y]). This means that adding an extra column to A does
not change the rank, and hence, by a sequence column operations [A|Y] is equivalent to [A|0n×1 ].
Observe that none of these operations is a permutation of the columns, since the first n columns of
[A|Y] and [A|0n×1 ] are the same. This means that Y can be obtained from the columns Ci , 1 ≤ i ≤ n
of A by means of transvections and dilatations. But then
n
X
Y = xi C i .
i=1

The solutions sought is thus

2 3
6 x1 7
6 7
6 7
6x 7
6 27
X=6 7
6 . 7.
6 . 7
6 . 7
6 7
4 5
xn
❑
Examples of Linear Systems 63

287 Problem Let A ∈ Mn×p (F), B ∈ Mn×q (F) and put rank (C) ⇐⇒ ∃P ∈ Mp (q) such that B = AP.
C = [A B] ∈ Mn×(p+q) (F) Prove that rank (A) =

3.3 Examples of Linear Systems

288 Example Use row reduction to solve the system

x + 2y + 3z + 4w = 8

x + 2y + 4z + 7w = 12

2x + 4y + 6z + 8w = 16

Solutions: Form the expanded matrix of coefficients and apply row operations to obtain
2 3 2 3
61 2 3 4 8 7 61 2 3 4 87
6 7 6 7
6 7 R3 −2R1 → R3 6 7
6
61 2 4 7 127
7
6
60 0 1 3 47
7.
6 7 R2 −R1 → R2 6 7
4 5 4 5
2 4 6 8 16 0 0 0 0 0

The matrix is now in row-echelon form. The variables x and z are the pivots, so w and y are free. Setting
w = s, y = t we have
z = 4 − 3s,
x = 8 − 4w − 3z − 2y = 8 − 4s − 3(4 − 3s) − 2t = −4 + 5s − 2t.
Hence the solution is given by
2 3 2 3
6 x 7 6−4 + 5s − 2t7
6 7 6 7
6 7 6 7
6y7 6 t 7
6 7 6 7
6 7=6 7.
6 7 6 7
6 7 6 7
6z7 6 4 − 3s 7
6 7 6 7
4 5 4 5
w s

289 Example Find α ∈ R such that the system

x + y − z = 1,

2x + 3y + αz = 3,
x + αy + 3z = 2,
posses (i) no solution, (ii) infinitely many solutions, (iii) a unique solution.

Solution: The augmented matrix of the system is

2 3
61 1 −1 17
6 7
6 7
6 7
62 3 α 37 .
6 7
4 5
1 α 3 2
64 Chapter 3

By performing R2 − 2R1 → R2 and R3 − R1 → R3 we obtain

2 3
61 1 −1 17
6 7
6 7
6 7
60 1 α+2 17 .
6 7
4 5
0 α−1 4 1

By performing R3 − (α − 1)R2 → R3 on this last matrix we obtain

2 3
61 1 −1 1 7
6 7
6 7
6 7.
60 1 α+2 1 7
6 7
4 5
0 0 (α − 2)(α + 3) α − 2

If α = −3, we obtain no solution. If α = 2, there is an infinity of solutions

2 3 2 3
6 x 7 6 5t 7
6 7 6 7
6 7 6 7
6 7=6 7 t ∈ R.
6y7 61 − 4t7 ,
6 7 6 7
4 5 4 5
z t

If α 6= 2 and α 6= 3, there is a unique solution

2 3 2 3

6 x7 6 1 7
6 7 6 7
6 7 6 1 7
6
7
6 7=6 7.
6y7 6
6 7 6α + 37
4 5 4 1 7 5
z
α+3

290 Example Solve the system

2 32 3 2 3
66 0 17 6 x 7 61 7
6 76 7 6 7
6 76 7 6 7
6 76 7 6 7
63 2 07 6y7 = 607 ,
6 76 7 6 7
4 54 5 4 5
1 0 1 z 2

for (x, y, z) ∈ (Z7 )3 .

Solution: Performing operations on the augmented matrix we have

2 3 2 3
66 0 1 17 61 0 1 27
6 7 6 7
6 7 R1 ↔ R3 6 7
6 7 6 7
63 2 0 07 63 2 0 07
6 7 6 7
4 5 4 5
1 0 1 2 6 0 1 1
2 3
61 0 1 27
6 7
R3 −6R1 → R3 6 7
6 7
60 2 4 17
R2 −3R1 → R2 6 7
4 5
0 0 2 3
Examples of Linear Systems 65

This gives
2z = 3 =⇒ z = 5,
2y = 1 − 4z = 2 =⇒ y = 1,
x = 2 − z = 4.
The solution is thus
(x, y, z) = (4, 1, 5).

291 Problem Find the general solution to the system 295 Problem Determine the values of the parameter m for
2 32 3 2 3 which the system

61 1 1 1 17 6a 7 6 1 7
6 76 7 6 7 x + y + (1 − m)z = m+2
6 76 7 6 7
61 17 6 7 6 7
6 0 1 0 7 6b7 6−17
6 76 7 6 7 (1 + m)x − y + 2z = 0
6 76 7 6 7
62 1 2 1 27 6 7=6 7
6 7 6c7 6 0 7
6 76 7 6 7 2x − my + 3z = m+2
6 76 7 6 7
64 2 4 2 47 6d7 6 0 7
6 76 7 6 7
4 54 5 4 5 is solvable.
1 0 0 0 1 f 0

or shew that there is no solution. 296 Problem Determine the values of the parameter m for
which the system

292 Problem Find all solutions of the system

2 32 3 2 3 x + y + z + t = 4a

61 1 1 1 17 6a7 637 x − y − z + t = 4b
6 76 7 6 7
6 76 7 6 7
61 1 1 1 27 6 7 6 7
6 7 6b7 647 −x − y + z + t = 4c
6 76 7 6 7
6 76 7 6 7
61 1 1 3 37 6 7 = 6 7,
6 7 6 c 7 677 x − y + z − t = 4d
6 76 7 6 7
6 76 7 6 7
61 1 4 4 47 6d7 667
6 76 7 6 7 is solvable.
4 54 5 4 5
1 2 3 4 5 f 9
297 Problem It is known that the system
if any.
ay + bx = c;

293 Problem Study the system

cx + az = b;
x + 2my + z = 4m; bz + cy = a

2mx + y + z = 2; possesses a unique solution. What conditions must

2 (a, b, c) ∈ R3 fulfill in this case? Find this unique solution.
x + y + 2mz = 2m ,
with real parameter m. You must determine, with proof, for
298 Problem Find strictly positive real numbers x, y, z such
which m this system has (i) no solution, (ii) exactly one so-
that
lution, and (iii) infinitely many solutions.
x 3 y 2 z6 = 1
294 Problem Study the following system of linear equations
with parameter a. x4 y5 z12 = 2

(2a − 1)x + ay − (a + 1)z = 1, x 2 y 2 z5 = 3.

ax + y − 2z = 1,
299 Problem (Leningrad Mathematical Olympiad, 1987, Grade 5)
2x + (3 − a)y + (2a − 6)z = 1. The numbers 1, 2, . . . , 16 are arranged in a 4 × 4 matrix A
as shewn below. We may add 1 to all the numbers of any row
You must determine for which a there is: (i) no solution, (ii) or subtract 1 from all numbers of any column. Using only the
a unique solution, (iii) infinitely many solutions. allowed operations, how can we obtain AT ?
66 Chapter 3

2 3 300 Problem (International Mathematics Olympiad, 1963)

Find all solutions x1 , x2 , x3 , x4 , x5 of the system
61 2 3 47
6 7 x5 + x2 = yx1 ;
6 7
65 6 7 87
6 7
A=6 7 x1 + x3 = yx2 ;
6 7
69 10 11 127 x2 + x4 = yx3 ;
6 7
4 5
x3 + x5 = yx4 ;
13 14 15 16
x4 + x1 = yx5 ,
where y is a parameter.
Chapter 4
R2, R3 and Rn

4.1 Points and Bi-points in R2

0 1
Ba1 C
R2 is the set of all points A = B C
 A with real number coordinates on the plane, as in figure 4.1. We use the
a2
0 1
B0C
notation O = B C
 A to denote the origin.
0
0 1
y
Ba1 C
A=B
 A
C
a2
b

bO x

Figure 4.1: Rectangular coordinates in R2 .

0 1 0 1
Ba1 C Bb1 C
Given A = B

C ∈ R2 and B = B C ∈ R2 we define their addition as
A  A
a2 b2

0 1 0 1 0 1
Ba1 C Bb1 C Ba1 + b1 C
A+B=B

C+B C= B
A  A 
C.
A (4.1)
a2 b2 a2 + b2

67
68 Chapter 4

Similarly, we define the scalar multiplication of a point of R2 by the scalar α ∈ R as

0 1 0 1
Ba1 C Bαa1 C
αA = α B

C=B
A 
C.
A (4.2)
a2 αa2

☞ Throughout this chapter, unless otherwise noted, we will use the convention that a point
A ∈ R2 will have its coordinates named after its letter, thus
0 1
Ba1 C
A=B
 A.
C
a2

301 Definition Consider the points A ∈ R2 , B ∈ R2 . By the bi-point starting at A and ending at B, denoted by
[A, B], we mean the directed line segment from A to B. We define
0 1
B0C
[A, A] = O = B C
 A.
0

☞ The bi-point [A, B] can be thus interpreted as an arrow starting at A and finishing, with the
arrow tip, at B. We say that A is the tail of the bi-point [A, B] and that B is its head. Some
authors use the terminology “fixed vector” instead of “bi-point.”

302 Definition Let A 6= B be points on the plane and let L be the line passing
 through
 A and B. The direction
of the bi-point [A, B] is the direction of the line L, that is, the angle θ ∈ − π ;
2 2
π
that the line L makes with the
horizontal. See figure 4.2.

303 Definition Let A, B lie on line L, and let C, D lie on line L ′ . If L||L ′ then we say that [A, B] has the same
direction as [C, D]. We say that the bi-points [A, B] and [C, D] have the same sense if they have the same direction
and if both their heads lie on the same half-plane made by the line joining their tails. They have opposite sense
if they have the same direction and if both their heads lie on alternative half-planes made by the line joining their
tails. See figures 4.3 and 4.4 .

B B B
D
A A A
bθ C C

Figure 4.2: Direction of Figure 4.3: Bi-points Figure 4.4: Bi-points

a bi-point with the same sense. with opposite sense.
Points and Bi-points in R2 69

☞ Bi-point [B, A] has the opposite sense of [A, B] and so we write

[B, A] = −[A, B].

304 Definition Let A 6= B. The Euclidean length or norm of bi-point [A, B] is simply the distance between A and
B and it is denoted by
q
||[A, B]|| = (a1 − b1 )2 + (a2 − b2 )2 .

We define
||[A, A]|| = ||O|| = 0.

A bi-point is said to have unit length if it has norm 1.

☞ A bi-point is completely determined by three things: (i) its norm, (ii) its direction, and (iii)
its sense.

305 Definition (Chasles’ Rule) Two bi-points are said to be contiguous if one has as tail the head of the other. In
such case we define the sum of contiguous bi-points [A, B] and [B, C] by Chasles’ Rule

[A, B] + [B, C] = [A, C].

See figure 4.5.

306 Definition (Scalar Multiplication of Bi-points) Let λ ∈ R \ {0} and A 6= B. We define

0[A, B] = O

and
λ[A, A] = O.

We define λ[A, B] as follows.

1. λ[A, B] has the direction of [A, B].

2. λ[A, B] has the sense of [A, B] if λ > 0 and sense opposite [A, B] if λ < 0.

3. λ[A, B] has norm |λ|||[A, B]|| which is a contraction of [A, B] if 0 < |λ| < 1 or a dilatation of [A, B] if |λ| > 1.

See figure 4.6 for some examples.

−2[A, B]

B [A, B]
A
C
1
[A, B]
2

Figure 4.5: Chasles’ Rule. Figure 4.6: Scalar multiplication of bi-points.

70 Chapter 4

4.2 Vectors in R2
307 Definition (Midpoint) Let A, B be points in R2 . We define the midpoint of the bi-point [A, B] as
0 1
a1 +b1
A+B B 2 C
=B

C.
A
2 a2 +b2
2

308 Definition (Equipollence) Two bi-points [X, Y] and [A, B] are said to be equipollent written [X, Y] ∼ [A, B] if
the midpoints of the bi-points [X, B] and [Y, A] coincide, that is,

X+B Y +A
[X, Y] ∼ [A, B] ⇔ = .
2 2
See figure 4.7.

Geometrically, equipollence means that the quadrilateral XYBA is a parallelogram. Thus the bi-points [X, Y] and
[A, B] have the same norm, sense, and direction.
Y

−
X ||
|| B
−

A
Figure 4.7: Equipollent bi-points.

309 Lemma Two bi-points [X, Y] and [A, B] are equipollent if and only if
0 1 0 1
By1 − x1 C Bb1 − a1 C
B C=B C.
 A  A
y2 − x2 b2 − a2

Proof: This is immediate, since

0 1 0 1
a1 +y1
B 2 C b1 +x1
B 2 C
B C B C
[X, Y] ∼ [A, B] ⇐⇒ B C= A
 A
a2 +y2 b 2 +x2
2 2
0 1 0 1
By1 − x1 C Bb1 − a1 C
⇐⇒ B C=B C,
 A  A
y2 − x2 b2 − a2

as desired. ❑

☞ From Lemma 309, equipollent bi-points have the same norm, the same direction, and the
same sense.

310 Theorem Equipollence is an equivalence relation.

Vectors in R2 71

0 1
By1 − x1 C
Proof: Write [X, Y] ∼ [A, B] if [X, Y] if equipollent to [A, B]. Now [X, Y] ∼ [X, Y] since B

C=
A
y2 − x2
0 1
By1 − x1 C
B C and so the relation is reflexive. Also
 A
y2 − x2
0 1 0 1
By1 − x1 C Bb1 − a1 C
[X, Y] ∼ [A, B] ⇐⇒ B C=B C
 A  A
y2 − x2 b2 − a2
0 1 0 1
Bb1 − a1 C By1 − x1 C
⇐⇒ B C=B C
 A  A
b2 − a2 y2 − x2

⇐⇒ [A, B] ∼ [X, Y],

and the relation is symmetric. Finally

0 1 0 1
By1 − x1 C Bb1 − a1 C
[X, Y] ∼ [A, B] ∧ [A, B] ∼ [U, V] ⇐⇒ B C=B C
 A  A
y2 − x2 b2 − a2
0 1 0 1
Bb1 − a1 C B v1 − u1 C
∧B

C=B
A 
C
A
b2 − a2 v2 − au2
0 1 0 1
By1 − x1 C Bv1 − u1 C
⇐⇒ B C=B C
 A  A
y2 − x2 v2 − u2

⇐⇒ [X, Y] ∼ [U, V],

and the relation is transitive. ❑

311 Definition (Vectors on the Plane) The equivalence class in which the bi-point [X, Y] falls is called the vector
−
−→
(or free vector) from X to Y, and is denoted by XY . Thus we write
2 3

→ 6 y1 − x1 7
[X, Y] ∈ XY = 6 7.
−
−
4 5
y2 − x2

If we desire to talk about a vector without mentioning a bi-point representative, we write, say, v, thus denoting
vectors with boldface lowercase letters. If it is necessary to mention the coordinates of v we will write
2 3
6v1 7
v=6
4
7.
5
v2
72 Chapter 4

☞ For any point X on the plane, we have −XX

−→
= 0, the zero vector. If [X, Y] ∈ v then [Y, X] ∈ −v.
0 1 2 3
Bp1 C → 6p1 7
312 Definition (Position Vector) For any particular point P = B C ∈ R2 we may form the vector OP = 6 7.
−
−
 A 4 5
p2 p2
−
−→ −
−→
We call OP the position vector of P and we use boldface lowercase letters to denote the equality OP = p.

0 1 0 1
B−1C B3C
313 Example The vector into which the bi-point with tail at A = B

C and head at B = B C falls is
A  A
2 4

2 3 2 3

→ 63 − (−1)7 647
AB = 6 7 = 6 7.
−
−
4 5 4 5
4−2 2

314 Example The bi-points [A, B] and [X, Y] with

0 1 0 1
B−1C B3C
A=B

C,B = B C,
A  A
2 4

0 1 0 1
B3C B7C
X=B C B C
 A,Y =  A
7 9

represent the same vector

2 3 2 3 2 3

→ 63 − (−1)7 647 67 − 37 − →
AB = 6 7=6 7=6 7 = XY.
−
− −
4 5 4 5 4 5
4−2 2 9−7
0 1 0 1
B−1 + nC B3 + nC
In fact, if S = B

C,T = B
A 
C then the infinite number of bi-points [S, T ] are representatives of of the
A
2+m 4+m
−
−→ − −→ −→
vectors AB = XY = ST .

−
−→
Given two vectors u, v we define their sum u + v as follows. Find a bi-point representative AB ∈ u and a
−
−→
contiguous bi-point representative BC ∈ v. Then by Chasles’ Rule
−
−→ − −→ −−→
u + v = AB + BC = AC.

Again, by virtue of Chasles’ Rule we then have

−
−→ −−→ −−→ −−→ −−→
AB = AO + OB = −OA + OB = b − a (4.3)

Similarly we define scalar multiplication of a vector by scaling one of its bi-point representatives.We define the
norm of a vector v ∈ R2 to be the norm of any of its bi-point representatives.
Vectors in R2 73

2 3 2 3
6u1 7 6v1 7
Componentwise we may see that given vectors u = 6
4
7, v = 6 7, and a scalar λ ∈ R then their sum and
5 4 5
u2 v2
scalar multiplication take the form
2 3 2 3 2 3
6u1 7 6v1 7 6λu1 7
u+v =6
4
7+6 7,
5 4 5 λu = 6
4
7.
5
u2 v2 λu2

−2u

B u
u v
A
C
u+v
1
u
2

Figure 4.8: Addition of Vectors. Figure 4.9: Scalar multiplication of vectors.

−
−→ −−→ −
−→ −−→
315 Example Diagonals are drawn in a rectangle ABCD. If AB = x and AC = y, then BC = y − x, CD = −x,
−−→ −−→
DA = x − y, and BD = y − 2x.

316 Definition (Parallel Vectors) Two vectors u and v are said to be parallel if there is a scalar λ such that u = λv.
If u is parallel to v we write u||v. We denote by Rv = {αv : α ∈ R}, the set of all vectors parallel to v.

☞ 0 is parallel to every vector.

2 3
6u1 7 q
317 Definition If u = 6
4
7, then we define its norm as ||u|| = u2 + u2 . The distance between two vectors u and
5 1 2
u2
v is dhu, vi = ||u − v||.

318 Example Let a ∈ R, a > 0 and let v 6= 0. Find a vector with norm a and parallel to v.

v
Solution: Observe that has norm 1 as
||v||
 
 v  ||v||
 
 ||v||  = ||v|| = 1.

v v
Hence the vector a has norm a and it is in the direction of v. One may also take −a .
||v|| ||v||

−−→ −−→ −−→ 1−−→

319 Example If M is the midpoint of the bi-point [X, Y] then XM = MY from where XM = 2
XY. Moreover, if T
74 Chapter 4

is any point, by Chasles’ Rule

−→ − → −−→ −−→ −−→ −−→
TX + TY = TM + MX + TM + MY
−−→ −−→ −−→
= 2TM − XM + MY
−−→
= 2TM.

320 Example Let △ABC be a triangle on the plane. Prove that the line joining the midpoints of two sides of the
triangle is parallel to the third side and measures half its length.

−
−→ −−−−−→
−
Solution: Let the midpoints of [A, B] and [A, C] be MC and MB , respectively. We shew that BC = 2MC MB . We
−−−−→ − →
− −−−−→ −−→
have 2AMC = AB and 2AMB = AC. Thus

−
−→ −
−→ −−→
BC = BA + AC
−
−→ −−→
= −AB + AC
−−−−→ −−−−→
= −2AMC + 2AMB
−−−−→ −−−−→
= 2MC A + 2AMB
−−−−→ −−−−→
= 2(MC A + AMB )
−−−−−→
−
= 2MC MB ,

as we wanted to shew.

321 Example In △ABC, let MC be the midpoint of side AB. Shew that

−−−−→ 1 −−→ −−→

CMC = CA + CB .
2

−−−−→ − →
−−−
Solution: Since AMC = MC B, we have

−−→ −−→ −−−−→ −−−−→ −−−−→ − →

−−−
CA + CB = CMC + MC A + CMC + MC B
−−−−→ −−−−→ − →
−−−
= 2CMC − AMC + MC B
−−−−→
= 2CMC ,

which yields the desired result.

322 Theorem (Section Formula) Let APB be a straight line and λ and µ be real numbers such that

||[A, P]|| λ
= .
||[P, B]|| µ
−−→ −−→ −
−→
With a = OA, b = OB, and p = OP, then
λb + µa
p= . (4.4)
λ+µ
Vectors in R2 75

Proof: Using Chasles’ Rule for vectors,

−
−→ −−→ −−→
AB = AO + OB = −a + b,
−
−→ −−→ − −→
AP = AO + OP = −a + p.
Also, using Chasles’ Rule for bi-points,

[A, P]µ = λ([P, B]) = λ([P, A] + [A, B]) = λ(−[A, P] + [A, B]),

whence
λ −
−→ λ −
−→ λ
[A, P] = [A, B] =⇒ AP = AB =⇒ p − a = (b − a).
λ+µ λ+µ λ+µ
On combining these formulæ

(λ + µ)(p − a) = λ(b − a) =⇒ (λ + µ)p = λb + µa,

from where the result follows. ❑

a a a

c d
b c c b
b

Figure 4.10: [A]. Problem 328. Figure 4.11: [B]. Problem 328. Figure 4.12: [C]. Problem 328.

a a b c

c d c d a d

f e
b b

Figure 4.13: [D]. Problem 328. Figure 4.14: [E]. Problem 328. Figure 4.15: [F]. Problem 328.

323 Problem
2 3 Let2a be a3real number. Find the distance be- 326 Problem For which values of a will the vectors
2 3 2 3
61 7 61 − a7
tween 4 5 and 4 5. 6a+17 6 2a + 5 7
a 1 a=4 5, b=4 5
2
a −1 a2 − 4a + 3

324 Problem Find all scalars λ for which ||λv|| = 1

, where will be parallel?
2 3 2

6 1 7 327 Problem In △ABC let the midpoints of [A, B] and [A, C]

v = 4 5. −−−−→ −−−−→
−1 be MC and MB , respectively. Put MC B = x, MB C =
−−→ −−→ −−→ −
−−−−− →
y, and CA = z. Express [A] AB + BC + MC MB , [B]
−−−−→ − −−−−−→ − →
−−− −−→ −−−−→ − −−− →
AMC + MC MB + MB C, [C] AC + MC A − BMB in terms
−
−→ −
−→ of x, y, and z.
325 Problem Given a pentagon ABCDE, find AB + BC +
−−→ − → −
− −→
CD + DE + EA.
76 Chapter 4

328 Problem A circle is divided into three, four equal, or six 332 Problem Let A, B be two points on the plane. Construct
equal parts (figures 4.10 through 4.15). Find the sum of the two points I and J such that
vectors. Assume that the divisions start or stop at the centre
of the circle, as suggested in the figures.
−→ →
− −→ 1−→
IA = −3IB, JA = − JB,
3
and then demonstrate that for any arbitrary point M on the
329 Problem Diagonals are drawn in a square (figures ??
plane
through ??). Find the vectorial sum a + b + c. Assume that − →
−− −−→ −−→
the diagonals either start, stop, or pass through the centre of MA + 3MB = 4MI
the square, as suggested by the figures. and
−
−−→ −−→ −−→
3MA + MB = 4MJ.

333 Problem You find an ancient treasure map in your great-

grandfather’s sea-chest. The sketch indicates that from the
gallows you should walk to the oak tree, turn right 90◦ and
330 Problem Prove that the mid-points of the sides of a skew
walk a like distance, putting and x at the point where you
quadrilateral form the vertices of a parallelogram.
stop; then go back to the gallows, walk to the pine tree, turn
left 90◦ , walk the same distance, mark point Y . Then you
331 Problem ABCD is a parallelogram. E is the midpoint of will find the treasure at the midpoint of the segment XY . So
[B, C] and F is the midpoint of [D, C]. Prove that you charter a sailing vessel and go to the remote south-seas
island. On arrival, you readily locate the oak and pine trees,
−−→ −−→ −
−→ but unfortunately, the gallows was struck by lightning, burned
AC + BD = 2BC.
to dust and dispersed to the winds. No trace of it remains.
What do you do?

4.3 Dot Product in R2

334 Definition Let (a, b) ∈ (R2 )2 . The dot product a•b of a and b is defined by
2 3 2 3
6a1 7 6b1 7
a•b = 6
4
7 • 6 7 = a1 b1 + a2 b2 .
5 4 5
a2 b2

The following properties of the dot product are easy to deduce from the definition.
DP1 Bilinearity
(x + y)•z = x•z + y•z, x•(y + z) = x•y + x•z (4.5)

DP2 Scalar Homogeneity

(αx)•y = x•(αy) = α(x•y), α ∈ R. (4.6)

DP3 Commutativity
x•y = y•x (4.7)

DP4
x•x ≥ 0 (4.8)

DP5
x•x = 0 ⇔ x = 0 (4.9)

DP6 √
||x|| = x•x (4.10)

335 Example If we put 2 3 2 3

617 60 7
i=6 7 6 7
4 5, j = 4 5,
0 1
Dot Product in R2 77

2 3
6a1 7
then we can write any vector a = 6
4
7 as a sum
5
a2

a = a1 i + a2 j.

The vectors 2 3 2 3
617 60 7
i=6 7 6 7
4 5, j = 4 5,
0 1

satisfy i•j = 0, and ||i|| = ||j|| = 1.

\
336 Definition Given vectors a and b, we define the angle between them, denoted by (a, b), as the angle between
any two contiguous bi-point representatives of a and b.

337 Theorem
\
a•b = ||a||||b|| cos (a, b).

Proof: Using Al-Kashi’s Law of Cosines on the length of the vectors, we have

\
||b − a||2 = ||a||2 + ||b||2 − 2||a||||b|| cos (a, b)

\
⇔ (b − a)•(b − a) = ||a||2 + ||b||2 − 2||a||||b|| cos (a, b)

\
⇔ b•b − 2a•b + a•a = ||a||2 + ||b||2 − 2||a||||b|| cos (a, b)

\
⇔ ||b||2 − 2a•b + ||b||2 = ||a||2 + ||b||2 − 2||a||||b|| cos (a, b)

\
⇔ a•b = ||a||||b|| cos (a, b),

as we wanted to shew. ❑
\
Putting (a, b) = π
in Theorem 337 we obtain the following corollary.
2

338 Corollary Two vectors in R2 are perpendicular if and only if their dot product is 0.

b−a
a

Figure 4.19: Theorem 337.

339 Definition Two vectors are said to be orthogonal if they are perpendicular. If a is orthogonal to b, we write
a ⊥ b.
78 Chapter 4

340 Definition If a ⊥ b and ||a|| = ||b|| = 1 we say that a and b are orthonormal.

☞ It follows that the vector 0 is simultaneously parallel and perpendicular to any vector!
341 Definition Let a ∈ R2 be fixed. Then the orthogonal space to a is defined and denoted by

a⊥ = {x ∈ R2 : x ⊥ a}.

Since | cos θ| ≤ 1 we also have

342 Corollary (Cauchy-Bunyakovsky-Schwarz Inequality)

|a•b| ≤ ||a||||b||.

343 Corollary (Triangle Inequality)

||a + b|| ≤ ||a|| + ||b||.

Proof:

||a + b||2 = (a + b)•(a + b)

= a•a + 2a•b + b•b

≤ ||a||2 + 2||a||||b|| + ||b||2

= (||a|| + ||b||)2 ,

from where the desired result follows. ❑

344 Corollary (Pythagorean Theorem) If a ⊥ b then

2 2
||a + b||2 = ||a|| + ||b|| .

Proof: Since a•b = 0, we have

||a + b||2 = (a + b)•(a + b)

= a•a + 2a•b + b•b

= a•a + 0 + b•b

= ||a||2 + ||b||2 ,

from where the desired result follows. ❑

345 Definition The projection of t onto v (or the v-component of t) is the vector

\ 1
projtv = (cos (t, v))||t|| v,
||v||

\
where (v, t) ∈ [0; π] is the convex angle between v and t read in the positive sense.
Dot Product in R2 79

☞ Given two vectors t and vector v 6= 0, find bi-point representatives of them having a common
tail and join them together at their tails. The projection of t onto v is the “shadow” of t in the
direction of v. To obtain projtv we prolong v if necessary and drop a perpendicular line to it from
the head of t. The projection is the portion between the common tails of the vectors and the point
where this perpendicular meets t. See figure 4.20.

b b

Figure 4.20: Vector Projections.

346 Corollary Let a 6= 0. Then

\ 1 x•a
projxa = cos (x, a)||x|| a= 2
a.
||a|| ||a||

347 Theorem Let a ∈ R2 \ {0}. Then any x ∈ R2 can be decomposed as

x = u + v,

where u ∈ Ra and v ∈ a⊥ .

Proof: We know that projxa is parallel to a, so we take u = projxa . This means that we must
then take v = x − projxa . We must demonstrate that v is indeed perpendicular to a. But this is
clear, as
a•v = a•x − a•projxa
x a •
= a•x − a• ||a|| 2a

= a•x − x•a

= 0,

completing the proof. ❑

348 Corollary Let v ⊥ w be non-zero vectors in R2 . Then any vector a ∈ R2 has a unique representation as a
linear combination of v, w ,
a = sv + tw, (s, t) ∈ R2 .

Proof: By Theorem 347, there exists a decomposition

a = sv + s ′ v ′ ,

where v ′ is orthogonal to v. But then v ′ ||w and hence there exists α ∈ R with v ′ = αw. Taking
t = s ′ α we achieve the decomposition
a = sv + tw.
To prove uniqueness, assume
sv + tw = a = pv + qw.
Then (s − p)v = (q − t)w. We must have s = p and q = t since otherwise v would be parallel to
w. This completes the proof. ❑
80 Chapter 4

349 Corollary Let p, q be non-zero, non-parallel vectors in R2 . Then any vector a ∈ R2 has a unique representation
as a linear combination of p, q ,
a = lp + mq, (l, m) ∈ R2 .

Proof: Consider z = q − projqp . Clearly p ⊥ z and so by Corollary 348, there exists unique
(s, t) ∈ R2 such that

a = sp + tz

= sp − tprojqp + tq
 •

q p
= s − t ||p|| 2 p + tq,
q p •
establishing the result upon choosing l = s − t ||p|| 2 and m = t. ❑

2 3 2 3
617 617
350 Example Let p = 6
4 5
7, q = 6 7. Write p as the sum of two vectors, one parallel to q and the other perpendicular
4 5
1 2
to q.

Solution: We use Theorem 347. We know that projp

q is parallel to q, and we find
2 3
3
p•q 657 3
projp
q = 2
q= q=6
4 5.
7
||q|| 5 6
5

We also compute 2 3 2 3
3 2
61 − 57 6 5 7
p − projp 6 7=6 7.
q = 4 5 4 5
6
1− 5
− 51

Observe that 2 3 2 3
3 2
657 6 5 7 6 6
6 7•6 7=
4 5 4 5 25 − 25 = 0,
6 1
5
−5

and the desired decomposition is 2 3 2 3 2 3

3 2
617 6 657 5 7
6 7=6 7+6 7.
4 5 4 5 4 5
6
1 5
− 15

b
C
b
B b H

b
A

Figure 4.21: Orthocentre.

Lines on the Plane 81

351 Example Prove that the altitudes of a triangle △ABC on the plane are concurrent. This point is called the
orthocentre of the triangle.

−−→ −−→ −−→

Solution: Put a = OA, b = OB, c = OC. First observe that for any x, we have, upon expanding,

(x − a)•(b − c) + (x − b)•(c − a) + (x − c)•(a − b) = 0. (4.11)

Let H be the point of intersection of the altitude from A and the altitude from B. Then
−−→ −
−→ −−→ −−→ −−→ −−→ −−→
0 = AH•CB = (OH − OA)•(OB − OC) = (OH − a)•(b − c), (4.12)

and
−−→ −−→ −−→ −−→ −−→ −−→ −−→
0 = BH•AC = (OH − OB)•(OC − OA) = (OH − b)•(c − a). (4.13)
−−→
Putting x = OH in (4.11) and subtracting from it (4.12) and (4.13), we gather that
−−→ −−→ −
−→
0 = (OH − c)•(a − b) = CH•AB,

which gives the result.

2 3
358 Problem Let a, b be fixed vectors in R2 . Prove that if
6 a 7
352 Problem Determine the value of a so that 4 5 be
∀v ∈ R2 , v•a = v•b,
1−a
2 3 then a = b.
6 1 7
perpendicular to 4 5 .
−1 359 Problem Let (a, b) ∈ (R2 )2 . Prove that

||a + b||2 + ||a − b||2 = 2||a||2 + 2||b||2 .

353 Problem Demonstrate that
(b + c = 0) ∧ (||a|| = ||b||) ⇐⇒ (a − b)•(a − c) = 0. 360 Problem Let u, v be vectors in R2 . Prove the polarisa-
tion identity:
2 3 2 3 2 3
1

4
6 7 −1
6 7 62 7 u•v = ||u + v||2 − ||u − v||2 .
354 Problem Let p = 4 5, r = 4 5, s = 4 5. Write p as 4
5 1 1
the sum of two vectors, one parallel to r and the other parallel 361 Problem Let x, a be non-zero vectors in R2 . Prove that
to s.
projax
355 Problem Prove that proj = αa,
a
||a||2 = (a•i)2 + (a•j)2 .

with 0 ≤ α ≤ 1.
356 Problem Let a 6= 0 6= b be vectors in R2 such that
a•b = 0. Prove that
362 Problem Let (λ, a) ∈ R×R2 be fixed. Solve the equation
αa + βb = 0 =⇒ α = β = 0.
a•x = λ
357 Problem Let (x, y) ∈ (R2 )2 with ||x|| = 3
2
||y||. Shew
that 2x + 3y and 2x − 3y are perpendicular. for x ∈ R2 .

4.4 Lines on the Plane

363 Definition Three points A, B, and C are collinear if they lie on the same line.

−
−→ −−→
It is clear that the points A, B, and C are collinear if and only if AB is parallel to AC. Thus we have the following
definition.
82 Chapter 4

364 0
Definition
1 The parametric equation with parameter t ∈ R of the straight line passing through the point

Bp1 C
P=B C
 A in the direction of the vector v 6= 0 is
p2
2 3
6 x − p1 7
6 7 = tv.
4 5
y − p2
2 3
6x7
If r = 6 7
4 5, then the equation of the line can be written in the form
y

r − p = tv. (4.14)

The Cartesian equation of a line is an equation of the form ax + by = c, where a2 + b2 6= 0. We write (AB) for
the line passing through the points A and B.

365 Theorem Let v 6= 0 and let n ⊥ v. An alternative form for the equation of the line r − p = tv is

(r − p)•n = 0.
2 3
6a7
Moreover, the vector 6 7
4 5 is perpendicular to the line with Cartesian equation ax + by = c.
b

Proof: The first part follows at once by observing that v•n = 0 and taking dot products to both
sides of 4.14. For the second part observe that at least one of a and b is 6= 0. First assume that
a 6= 0. Then we can put y = t and x = − ba
c
t+ a and the parametric equation of this line is
2 3 2 3
c b
6x − a7 6− a 7
6 7 = t6 7,
4 5 4 5
y 1

and we have 2 3 2 3
b
6− a 7 6a7 b
6 7 • 6 7 = − · a + b = 0.
4 5 4 5 a
1 b

Similarly if b 6= 0 we can put x = t and y = − a

b
t+ c
b
and the parametric equation of this line is
2 3 2 3
6 x 7 6 1 7
6 7 = t6 7,
4 5 4 5
c
y− b
−a
b

and we have 2 3 2 3
6 1 7 6a7 a
6 7•6 7 = a − · b = 0,
4 5 4 5 b
−a
b
b

proving the theorem in this case. ❑

Lines on the Plane 83

2 3
√ a

☞ The vector 664 a2 +b2 7

7
5 has norm 1 and is orthogonal to the line ax + by = c.
√ b
a2 +b2

0 1 2 3
B2C 6−47
366 Example The equation of the line passing through A = B
 A
C and in the direction of v = 6
4
7 is
5
3 5

2 3 2 3
6x − 27 6−47
6 7 = λ6 7.
4 5 4 5
y−3 5

0 1 0 1
B−1C B−2C
367 Example Find the equation of the line passing through A = B

C and B = B
A 
C.
A
1 3

Solution: The direction of this line is that of

2 3 2 3

→ 6−2 − (−1)7 6−17

AB = 6 7=6 7.
−
−
4 5 4 5
3−1 2

The equation is thus

2 3 2 3
6x + 17 6−17
6 7 = λ6 7 , λ ∈ R.
4 5 4 5
y−1 2

368 Example Suppose that (m, b) ∈ R2 . Write the Cartesian equation of the line y = mx + b in parametric form.

Solution: Here is a way. Put x = t. Then y = mt + b and so the desired parametric form is
2 3 2 3
6 x 7 617
6 7 = t6 7.
4 5 4 5
y−b m

369 Example Let (m1 , m2 , b1 , b2 ) ∈ R4 , m1 m2 6= 0. Consider the lines L1 : y = m1 x + b1 and L2 : y =

m2 x + b2 . By translating this problem in the language of vectors in R2 , shew that L1 ⊥ L2 if and only if
m1 m2 = −1.
Solution: The parametric equations of the lines are
2 3 2 3 2 3 2 3
6 x 7 6 1 7 6 x 7 6 1 7
L1 : 6
4
7 = s6
5 4
7,
5 L2 : 6
4
7 = t6
5 4
7.
5
y − b1 m1 y − b2 m2
84 Chapter 4

2 3 2 3
6 1 7 6 1 7
Put v = 6
4
7 and w = 6
5 4
7. Since the lines are perpendicular we must have v•w = 0, which yields
5
m1 m2

0 = v•w = 1(1) + m1 (m2 ) =⇒ m1 m2 = −1.

370 Theorem (Distance Between a Point and a Line) Let (r − a)•n = 0 be a line passing through the point A
and perpendicular to vector n. If B is not a point on the line, then the distance from B to the line is

|(a − b)•n|
.
||n||

If the line has Cartesian equation ax + by = c, then this distance is

|ab1 + bb2 − c|
√ .
a2 + b2

−
−−→
Proof: Let R0 be the point on the line that is nearest to B. Then BR0 = r0 − b is orthogonal to
the line, and the distance we seek is
 
 (r − b)•n  |(r − b)•n|
 0  0
||projrn0 −b || =  n = .
 ||n||2  ||n||

Since R0 is on the line, r0 •n = a•n, and so

|r0 •n − b•n| |a•n − b•n| |(a − b)•n|

||projrn0 −b || = = = ,
||n||| ||n|| ||n||

as we wanted to shew.

If the line has Cartesian equation ax + by = c, then at least one of a and b is 6= 0. Let us suppose
a 6= 0, as the argument when a = 0 and b 6= 0 is similar. Then ax + by = c is equivalent to
02 3 2 31 2 3
c
B6 x 7 6 7C 6a7
B6 7 − 6 a 7C • 6 7 = 0.
4 5 4 5A 4 5
y 0 b
2 3 2 3 0 1
c
6a76a7 Bb1 C √
We use the result obtained above with a = 6 7 6 7 B C
4 5, n = 4 5, and B =  A. Then ||n|| = a + b
2 2

0 b b2
and 2 3 2 3
 
 c −b 
6 a 1 7 6 a 7
|(a − b)•n| = 64
7 • 6 7 = |c − ab1 − bb2 | = |ab1 + bb2 − c|,
5 4 5
 
 −b2 b 

giving the result. ❑

371 Example Recall that the medians of △ABC are lines joining the vertices of △ABC with the midpoints of the
side opposite the vertex. Prove that the medians of a triangle are concurrent, that is, that they pass through a
common point.

☞ This point of concurrency is called, alternatively, the isobarycentre, centroid, or centre of gravity
of the triangle.
Lines on the Plane 85

Solution: Let MA , MB , and MC denote the midpoints of the lines opposite

2 A,3 B, and C, respectively. The equation

−−−−→ 6x7
of the line passing through A and in the direction of AMA is (with r = 6
4 5)
7
y

−−→ −−−−→
r = OA + rAMA .
− →
−−−
Similarly, the equation of the line passing through B and in the direction of BMB is
−−→ −−−−→
r = OB + sBMA .
−−→ −−−−→
These two lines must intersect at a point G inside the triangle. We will shew that GC is parallel to CMC , which
means that the three points G, C, MC are collinear.
Now, ∃(r0 , s0 ) ∈ R2 such that
−−→ −−−−→ −−→ −−→ − →
−−−
OA + r0 AMA = OG = OB + s0 BMB ,

that is −−−−→ − → −−→ −−→

−−−
r0 AMA − s0 BMB = OB − OA,
or −
−→ − −−−→ −
−→ −−−−→ −
−→
r0 (AB + BMA ) − s0 (BA + AMB ) = AB.
−−−→ −
− −→ −−−−→ −−→
Since MA and MB are the midpoints of [B, C] and [C, A] respectively, we have 2BMA = BC and 2AMB = AC =
−→ −
− →
−
AB + BC. The relationship becomes

−
−→ 1−→
− −
−→ 1−→
− 1−→
− −
−→
r0 (AB + BC) − s0 (−AB + AB + BC) = AB,
2 2 2
s0 −
−→ r0 s0 −
−→
(r0 + − 1)AB = (− + )BC.
2 2 2
We must have
s0
r0 + − 1 = 0,
2
r0 s0
− + = 0,
2 2
−
−→ −
−→
since otherwise the vectors AB and BC would be parallel, and the triangle would be degenerate. Solving, we find
−
−→ −
− −−→ −−→ −−→ −−−−→ −−→ −−−−→
s0 = r0 = 23 . Thus we have OA + 23 AMA = OG, or AG = 23 AMA , and similarly, BG = 23 BMB .

−−→ −−−−→ −−→ −−−−→ −→

− −−→
From AG = 23 AMA , we deduce AG = 2GMA . Since MA is the midpoint of [B, C], we have GB + GC =
−−−−→ −−→
2GMA = AG, which is equivalent to
−−→ −−→ −−→
GA + GB + GC = 0.
−−→ − −→ −−−−→
As MC is the midpoint of [A, B] we have GA + GB = 2GMC . Thus
−−→ −−→ −−→ −−−−→ −−→
0 = GA + GB + GC = 2GMC + GC.
−−→ −−−−→
This means that GC = −2GMC , that is, that they are parallel, and so the points G, C and MC all lie on the same
line. This achieves the desired result.
☞ The centroid of △ABC satisfies thus
−−→ −−→ −−→
GA + GB + GC = 0,

and divides the medians on the ratio 2 : 1, reckoning from a vertex.

86 Chapter 4

372 Problem Find the angle between the lines 2x − y = 1 and ➊ Let E and F be points such that
x − 3y = 1.
−
−→ 1 −−→ −→ 3 −−→
AE = AC and AF = AC.
4 4
373 Problem Find the equation of the line passing through
0 1 2 3 Demonstrate that the lines (BE) and (DF) are parallel.
B 1 C 627 ➋ Let I be the midpoint of [A, D] and J be the midpoint
 A and in a direction perpendicular to 4 5. of [B, C]. Demonstrate that the lines (AB) and (IJ) are
−1 1 parallel. What type of quadrilateral is IEJF?

374 Problem △ABC has centroid G, and △A ′ B ′ C ′ satisfies 377 Problem ABCD is a parallelogram; point I is the mid-
→
− −→
point of [A, B]. Point E is defined by the relation IE = 13 ID.
→ −−→ − → Prove that  
−
−− −−
AA ′ + BB ′ + CC ′ = 0. −→ 1 −
− −→ −−→
AE = AB + AD
3
Prove that G is also the centroid of △A ′ B ′ C ′ . and prove that the points A, C, E are collinear.

375 Problem Let ABCD be a trapezoid, with bases [A, B] −−→ −−→ −−→
378 Problem Put OA = a, OB = b, OC = c. Prove that
and [C, D]. The lines (AC) and (BD) meet at E and the A, B, C are collinear if and only if there exist real numbers
lines (AD) and (BC) meet at F. Prove that the line (EF) α, β, γ, not all zero, such that
passes through the midpoints of [A, B] and [C, D] by proving
the following steps. αa + βb + γc = 0, α + β + γ = 0.
➊ Let I be the midpoint of [A, B] and let J be the point of
intersection of the lines (FI) and (DC). Prove that J is 379 Problem Prove Desargues’ Theorem: If △ABC and
the midpoint of [C, D]. Deduce that F, I, J are collinear. △A ′ B ′ C ′ (not necessarily in the same plane) are so posi-
➋ Prove that E, I, J are collinear. tioned that (AA ′ ), (BB ′ ), (CC ′ ) all pass through the same
point V and if (BC) and (B ′ C ′ ) meet at L, (CA) and (C ′ A ′ )
meet at M, and (AB) and (A ′ B ′ ) meet at N, then L, M, N
376 Problem Let ABCD be a parallelogram. are collinear.

4.5 Vectors in R3
We now extend the notions studied for R2 to R3 . The rectangular coordinate form of a vector in R3 is
2 3
6a1 7
6 7
6 7
a=6 7
6a2 7 .
6 7
4 5
a3

In particular, if
2 3 2 3 2 3
617 60 7 607
6 7 6 7 6 7
6 7 6 7 6 7
i = 607 , j = 617 , k = 6
6 7 6 7 7
607
6 7 6 7 6 7
4 5 4 5 4 5
0 0 1
2 3
6a1 7
6 7
6 7
then we can write any vector a = 6 7
6a2 7 as a sum
6 7
4 5
a3

a = a1 i + a2 j + a3 k.
Vectors in R3 87

2 3 2 3
6a1 7 6b1 7
6 7 6 7
6 7 6 7
Given a = 6 7 and b = 6 7, their dot product is
6a2 7 6b2 7
6 7 6 7
4 5 4 5
a3 b3

a•b = a1 b1 + a2 b2 + a3 b3 ,

and q
||a|| = a21 + a22 + a23 .
We also have
i•j = j•k = k•i = 0,
and
||i|| = ||j|| = ||k|| = 1.

380 Definition A system of unit vectors i, j, k is right-handed if the shortest-route rotation which brings i to coincide
with j is performed in a counter-clockwise manner. It is left-handed if the rotation is done in a clockwise manner.

To study points in space we must first agree on the orientation that we will give our coordinate system. We will use,
unless otherwise noted, a right-handed orientation, as in figure 4.22.
k k

j j

i i

Figure 4.22: Right-handed system. Figure 4.23: Left-handed system.

☞ The analogues of the Cauchy-Bunyakovsky-Schwarz and the Triangle Inequality also hold in
R3 .

We now define the (standard) cross (wedge) product in R3 as a product satisfying the following properties.

381 Definition Let (x, y, z, α) ∈ R3 × R3 × R3 × R. The wedge product × : R3 × R3 → R3 is a closed binary

operation satisfying
CP1 Anti-commutativity:
x×y = −(y×x) (4.15)

CP2 Bilinearity:
(x + z)×y = x×y + z×y, x×(z + y) = x×z + x×y (4.16)

CP3 Scalar homogeneity:

(αx)×y = x×(αy) = α(x×y) (4.17)

CP4
x×x = 0 (4.18)

CP5 Right-hand Rule:

i×j = k, j×k = i, k×i = j (4.19)
88 Chapter 4

2 3 2 3
6x1 7 6y1 7
6 7 6 7
6 7 6 7
382 Theorem Let x = 6 7 and y = 6 7 be vectors in R3 . Then
6x2 7 6y2 7
6 7 6 7
4 5 4 5
x3 y3

x×y = (x2 y3 − x3 y2 )i + (x3 y1 − x1 y3 )j + (x1 y2 − x2 y1 )k.

Proof: Since i×i = j×j = k×k = 0 we have

(x1 i + x2 j + x3 k)×(y1 i + y2 j + y3 k) = x1 y2 i×j + x1 y3 i×k

+x2 y1 j×i + x2 y3 j×k

+x3 y1 k×i + x3 y2 k×j

= x1 y2 k − x1 y3 j − x2 y1 k

+x2 y3 i + x3 y1 j − x3 y2 i,

from where the theorem follows. ❑

383 Example Find

2 3 2 3
6 1 7 6 07
6 7 6 7
6 7 6 7
6 7 6 7
6 0 7 × 6 17 .
6 7 6 7
4 5 4 5
−3 2

Solution: We have

(i − 3k)×(j + 2k) = i×j + 2i×k − 3k×j − 6k×k

= k − 2j − 3i + 60

= −3i − 2j + k.

Hence
2 3 2 3 2 3
6 1 7 607 6−37
6 7 6 7 6 7
6 7 6 7 6 7
6 7×6 7 = 6 7
6 0 7 617 6−27 .
6 7 6 7 6 7
4 5 4 5 4 5
−3 2 1

384 Theorem The cross product vector x×y is simultaneously perpendicular to x and y.
Vectors in R3 89

Proof: We will only check the first assertion, the second verification is analogous.

x•(x×y) = (x1 i + x2 j + x3 k)•((x2 y3 − x3 y2 )i

+(x3 y1 − x1 y3 )j + (x1 y2 − x2 y1 )k)

= x1 x2 y3 − x1 x3 y2 + x2 x3 y1 − x2 x1 y3 + x3 x1 y2 − x3 x2 y1

= 0,

completing the proof. ❑

385 Theorem a×(b×c) = (a•c)b − (a•b)c.

Proof:

a×(b×c) = (a1 i + a2 j + a3 k)×((b2 c3 − b3 c2 )i+

+(b3 c1 − b1 c3 )j + (b1 c2 − b2 c1 )k)

= a1 (b3 c1 − b1 c3 )k − a1 (b1 c2 − b2 c1 )j

−a2 (b2 c3 − b3 c2 )k + a2 (b1 c2 − b2 c1 )i

+a3 (b2 c3 − b3 c2 )j − a3 (b3 c1 − b1 c3 )i

= (a1 c1 + a2 c2 + a3 c3 )(b1 i + b2 j + b3 i)

+(−a1 b1 − a2 b2 − a3 b3 )(c1 i + c2 j + c3 i)

= (a•c)b − (a•b)c,

completing the proof. ❑

386 Theorem (Jacobi’s Identity)

a×(b×c) + b×(c×a) + c×(a×b) = 0.

Proof: From Theorem 385 we have

a×(b×c) = (a•c)b − (a•b)c,

b×(c×a) = (b•a)c − (b•c)a,

c×(a×b) = (c•b)a − (c•a)b,

and adding yields the result. ❑

\
387 Theorem Let (x, y) ∈ [0; π[ be the convex angle between two vectors x and y. Then

\
||x×y|| = ||x||||y|| sin (x, y).
90 Chapter 4

Proof: We have

||x×y||2 = (x2 y3 − x3 y2 )2 + (x3 y1 − x1 y3 )2 + (x1 y2 − x2 y1 )2

= x22 y23 − 2x2 y3 x3 y2 + x23 y22 + x23 y21 − 2x3 y1 x1 y3 +

+x21 y23 + x21 y22 − 2x1 y2 x2 y1 + x22 y21

= (x21 + x22 + x23 )(y21 + y22 + y23 ) − (x1 y1 + x2 y2 + x3 y3 )2

= ||x||2 ||y||2 − (x•y)2

\
= ||x||2 ||y||2 − ||x||2 ||y||2 cos2 (x, y)

\
= ||x||2 ||y||2 sin2 (x, y),

whence the theorem follows. The Theorem is illustrated in Figure 4.24. Geometrically it means
that the area of the parallelogram generated by joining x and y at their heads is ||x×y||. ❑
x×
y

x y

Figure 4.24: Theorem 387.

The following corollaries are now obvious.

388 Corollary Two non-zero vectors x, y satisfy x×y = 0 if and only if they are parallel.

389 Corollary (Lagrange’s Identity)

2 2
||x×y||2 = ||x|| ||y|| − (x•y)2 .

390 Example Let x ∈ R3 , ||x|| = 1. Find

||x×i||2 + ||x×j||2 + ||x×k||2 .

Solution: By Lagrange’s Identity,

2 2
||x×i||2 = ||x|| ||i|| − (x•i)2 = 1 − (x•i)2 ,
2 2
||x×k||2 = ||x|| ||j|| − (x•j)2 = 1 − (x•j)2 ,
2 2
||x×j||2 = ||x|| ||k|| − (x•k)2 = 1 − (x•k)2 ,
2
and since (x•i)2 + (x•j)2 + (x•k)2 = ||x|| = 1, the desired sum equals 3 − 1 = 2.
−
−→
391 Problem Consider a tetrahedron ABCS. [A] Find AB + 392 Problem Find a vector simultaneously perpendicular to
−
−→ − −→ −−→ − −→ − → −
− →
−
BC + CS. [B] Find AC + CS + SA + AB.
Planes and Lines in R3 91

2 3 2 3
395 Problem Prove that x×x = 0 follows from the anti-
617 617 commutativity of the cross product.
6 7 6 7
6 7 6 7
617 and 617 and having norm 3.
6 7 6 7
4 5 4 5 396 Problem Expand the product (a − b)×(a + b).
1 0

397 Problem The vectors a, b are constant vectors. Solve the

393 Problem Find the area of the triangle whose vertices are equation a×(x×b) = b×(x×a).
0 1 0 1 0 1
B0C B0C B1C
B C B C B C 398 Problem The vectors a, b, c are constant vectors. Solve
B C B C B C
at P = B0C, Q = B1C, and R = B0C. the system of equations
B C B C B C
 A  A  A
1 0 0 2x + y×a = b, 3y + x×a = c,

394 Problem Prove or disprove! The cross product is associa- 399 Problem Prove that there do not exist three unit vectors
2π
tive. in R3 such that the angle between any two of them be > .
3

4.6 Planes and Lines in R3

400 Definition If bi-point representatives of a family of vectors in R3 lie on the same plane, we will say that the
vectors are coplanar or parallel to the plane.

401 Lemma Let v, w in R3 be non-parallel vectors. Then every vector u of the form

u = av + bw,

((a, b) ∈ R2 arbitrary) is coplanar with both v and w. Conversely, any vector t coplanar with both v and w can
be uniquely expressed in the form
t = pv + qw.

Proof: This follows at once from Corollary 349, since the operations occur on a plane, which
can be identified with R2 . ❑

A plane is determined by three non-collinear points. Suppose that A, B, and C are non-collinear points on the
0 1
BxC
B C
B C →
same plane and that R = B C −
−
ByC is another arbitrary point on this plane. Since A, B, and C are non-collinear, AB
B C
 A
z
−−→ −
−→
and AC, which are coplanar, are non-parallel. Since AR also lies on the plane, we have by Lemma 401, that there
exist real numbers p, q with
−→
− −
−→ −−→
AR = pAB + qAC.
By Chasles’ Rule,
−−→ −−→ −−→ −−→ −−→ −−→
OR = OA + p(OB − OA) + q(OC − OA),
−−→ −−→
is the equation of a plane containing the three non-collinear points A, B, and C. By letting r = OR, a = OA, etc.,
we deduce that
r − a = p(b − a) + q(c − a).
Thus we have the following definition.
92 Chapter 4

402 Definition The parametric equation of a plane containing the point A, and parallel to the vectors u and v is
given by
r − a = pu + qv.

Componentwise this takes the form

x − a1 = pu1 + qv1 ,

y − a2 = pu2 + qv2 ,

z − a3 = pu3 + qv3 .

The Cartesian equation of a plane is an equation of the form ax + by + cz = d with (a, b, c, d) ∈ R4 and
a2 + b2 + c2 6= 0.

403 Example Find both the parametric equation and the Cartesian equation of the plane parallel to the vectors
2 3 2 3 0 1
617 617 B 0 C
6 7 6 7 B C
6 7 6 7 B C
6 7 and 6 7 and passing through the point B C
617 617 B−1C.
6 7 6 7 B C
4 5 4 5  A
1 0 2

Solution: The desired parametric equation is

2 3 2 3 2 3
6 x 7 617 617
6 7 6 7 6 7
6 7 6 7 6 7
6 7 = s6 7 + t6 7.
6y + 17 617 617
6 7 6 7 6 7
4 5 4 5 4 5
z−2 1 0

This gives s = z − 2, t = y + 1 − s = y + 1 − z + 2 = y − z + 3 and x = s + t = z − 2 + y − z + 3 = y + 1.

Hence the Cartesian equation is x − y = 1.

404 Theorem Let u and v be non-parallel vectors and let r − a = pu + qv be the equation of the plane containing
A an parallel to the vectors u and v. If n is simultaneously perpendicular to u and v then

(r − a)•n = 0.

2 3
6a7
6 7
6 7
Moreover, the vector 6 7
6b7 is normal to the plane with Cartesian equation ax + by + cz = d.
6 7
4 5
c

Proof: The first part is clear, as u•n = 0 = v•n. For the second part, recall that at least one of
a, b, c is non-zero. Let us assume a 6= 0. The argument is similar if one of the other letters is
non-zero and a = 0. In this case we can see that

d b c
x= − y − z.
a a a
Planes and Lines in R3 93

Put y = s and z = t. Then

2 3 2 3 2 3
d b c
6x − a7 6− a 7 6− a 7
6 7 6 7 6 7
6 7 6 7 6 7
6 7 = s6 7+t6 7
6 y 7 6 1 7 6 0 7
6 7 6 7 6 7
4 5 4 5 4 5
z 0 1

is a parametric equation for the plane. ❑

405 Example Find once again, by appealing to Theorem 404, the Cartesian equation of the plane parallel to the
2 3 2 3 0 1
617 61 7 B 0 C
6 7 6 7 B C
6 7 6 7 B C
vectors 6 7 and 6 7 and passing through the point B
617 61 7
C
B−1C.
6 7 6 7 B C
4 5 4 5  A
1 0 2

2 3 2 3 2 3
617 617 6−17
6 7 6 7 6 7
6 7 6 7 6 7
Solution: The vector 617 × 617 = 6 1 7
6 7 6 7 6
7 is normal to the plane. The plane has thus equation
6 7 6 7 6 7
4 5 4 5 4 5
1 0 0

2 3 2 3
6 x 7 6−17
6 7 6 7
6 7 6 7
6 7•6 7
6y + 17 6 1 7 = 0 =⇒ −x + y + 1 = 0 =⇒ x − y = 1,
6 7 6 7
4 5 4 5
z−2 0

as obtained before.

406 Theorem (Distance Between a Point and a Plane) Let (r − a)•n = 0 be a plane passing through the point
A and perpendicular to vector n. If B is not a point on the plane, then the distance from B to the plane is

|(a − b)•n|
.
||n||

−
−−→
Proof: Let R0 be the point on the plane that is nearest to B. Then BR0 = r0 − b is orthogonal
to the plane, and the distance we seek is
 
 (r − b)•n  |(r − b)•n|
 0  0
||projrn0 −b || =  n = .
 ||n||
2  ||n||

Since R0 is on the plane, r0 •n = a•n, and so

|r0 •n − b•n| |a•n − b•n| |(a − b)•n|

||projrn0 −b || = = = ,
||n||| ||n|| ||n||

as we wanted to shew. ❑

☞ Given three planes in space, they may (i) be parallel (which allows for some of them to
coincide), (ii) two may be parallel and the third intersect each of the other two at a line, (iii)
intersect at a line, (iv) intersect at a point.
94 Chapter 4

407 Definition The equation of a line passing through A ∈ R3 in the direction of v 6= 0 is given by

r − a = tv, t ∈ R.

−−→ −−→ −−→

408 Theorem Put OA = a, OB = b, and OC = c. Points (A, B, C) ∈ (R3 )3 are collinear if and only if

a×b + b×c + c×a = 0.

−
−→ −−→
Proof: If the points A, B, C are collinear, then AB is parallel to AC and by Corollary 388, we
must have
(c − a)×(b − a) = 0.
Rearranging, gives
c×b − c×a − a×b = 0.
Further rearranging completes the proof. ❑

409 Theorem (Distance Between a Point and a Line) Let L : r = a + λv, v 6= 0, be a line and let B be a point
not on L. Then the distance from B to L is given by

||(a − b)×v||
.
||v||

Proof: If R0 —with position vector r0 —is the point on L that is at shortest distance from B then
−
−−→
BR0 is perpendicular to the line, and so
−
−−→ −
−−→ π −
−−→
||BR0 ×v|| = ||BR0 ||||v|| sin = ||BR0 ||||v||.
2
− →
 −−
The distance we must compute is BR0  = ||r0 − b||, which is then given by

−
−−→
||BR0 ×v|| ||(r0 − b)×v||
||r0 − b|| = = .
||v|| ||v||

Now, since R0 is on the line ∃t0 ∈ R such that r0 = a + t0 v. Hence

(r0 − b)×v = (a − b)×v,

giving
||(a − b)×v||
||r0 − b|| = ,
||v||
proving the theorem. ❑

☞ Given two lines in space, one of the following three situations might arise: (i) the lines
intersect at a point, (ii) the lines are parallel, (iii) the lines are skew (one over the other, without
intersecting).

410 Problem Find the equation of the plane passing through (1, −1, −1) and containing the line x = 2y = 3z.
the points (a, 0, a), (−a, 1, 0), and (0, 1, 2a) in R3 .
413 Problem Find the equation of the plane perpendicular to
411 Problem Find the equation of plane containing the point the line ax = by = cz, abc 6= 0 and passing through the
(1, 1, 1) and perpendicular to the line x = 1 + t, y = −2t, z = point (1, 1, 1) in R3 .
1 − t.
414 Problem Find the equation of the line perpendicular to
412 Problem Find the equation of plane containing the point the plane ax + a2 y + a3 z = 0, a =
6 0 and passing through
Rn 95

the point (0, 0, 1). 417 Problem Points a, b, c in R3 are collinear and it is known
that a×c = i − 2j and a×b = 2k − 3i. Find b×c.
415 Problem The two planes
x − y − z = 1, x − z = −1, 418 Problem Find the equation of the plane which is equidis-
0 1 0 1
intersect at a line. Write the equation of this line in the form
2 3 B3C B 1 C
B C B C
B C B C
6x7 tant of the points B2C and B−1C.
6 7 B C B C
6 7  A  A
6y7 = a + tv, t ∈ R.
6 7 1 1
4 5
z
419 Problem (Putnam Exam, 1980) Let S be the solid in
416 Problem Find the equation of the plane passing through three-dimensional space consisting of all points (x, y, z) sat-
0 1 0 1 2 3 isfying the following system of six conditions:
B 1 C B2C 6x7
B C B C 6 7 x ≥ 0, y ≥ 0, z ≥ 0,
B C B C 6 7
the points B 0 C, B1C and parallel to the line 6y7 =
B C B C 6 7
 A  A 4 5
−1 1 z x + y + z ≤ 11,
2 3 2 3
2x + 4y + 3z ≤ 36,
6−17 617
6 7 6 7
6 7 6 7 2x + 3z ≤ 24.
6−27 + t 607 ..
6 7 6 7
4 5 4 5 Determine the number of vertices and the number of edges of
0 1 S.

4.7 Rn
As a generalisation of R2 and R3 we define Rn as the set of n-tuples
2 3 

 



 6 x1 7 



 6 7 


 6 7 

 6 7
6 x 2 7 

6 7 : xi ∈ R .
 6 . 7 
6 7
6 .. 7

 

 


 6 7 

4 5
 


 x 

n

The dot product of two vectors in Rn is defined as

2 3 2 3
6 x1 7 6 y1 7
6 7 6 7
6 7 6 7
6x 7 6y 7
6 27 6 27
x•y = 6 7•6 7
6 . 7 6 . 7 = x1 y1 + x2 y2 + · · · + xn yn .
6 . 7 6 . 7
6 . 7 6 . 7
6 7 6 7
4 5 4 5
xn yn

The norm of a vector in Rn is given by √

||x|| = x•x.
As in the case of R2 and R3 we have

420 Theorem (Cauchy-Bunyakovsky-Schwarz Inequality) Given (x, y) ∈ (Rn )2 the following inequality holds
|x•y| ≤ ||x||||y||.
96 Chapter 4

n
X n
X n
X
Proof: Put a = x2k , b = xk yk , and c = y2k . Consider
k=1 k=1 k=1

n
X n
X n
X n
X
f(t) = (txk − yk )2 = t2 x2k − 2t xk yk + y2k = at2 + bt + c.
k=1 k=1 k=1 k=1

This is a quadratic polynomial which is non-negative for all real t, so it must have complex roots.
Its discriminant b2 − 4ac must be non-positive, from where we gather
!2 ! !
n
X n
X n
X
4 xk yk ≤4 x2k y2k .
k=1 k=1 k=1

This gives
|x•y|2 ≤ ||x||2 ||y||2
from where we deduce the result. ❑

421 Example Assume that ak , bk , ck , k = 1, . . . , n, are positive real numbers. Shew that
!4 ! ! !2
n
X n
X n
X n
X
ak bk ck ≤ a4k b4k c2k .
k=1 k=1 k=1 k=1

Pn
Solution: Using CBS on k=1 (ak bk )ck once we obtain
!1/2 !1/2
n
X n
X n
X
ak bk ck ≤ a2k b2k c2k .
k=1 k=1 k=1
Pn
1/2
Using CBS again on k=1 a2k b2k we obtain

Pn Pn
1/2 Pn
1/2
k=1 ak bk ck ≤ k=1 a2k b2k k=1 c2k
Pn
1/4 Pn
1/4 Pn
1/2
≤ k=1 a4k k=1 b4k k=1 c2k ,

which gives the required inequality.

422 Theorem (Triangle Inequality) Given (x, y) ∈ (Rn )2 the following inequality holds
||x + y|| ≤ ||x|| + ||y||.

Proof: We have
||a + b||2 = (a + b)•(a + b)

= a•a + 2a•b + b•b

≤ ||a||2 + 2||a||||b|| + ||b||2

= (||a|| + ||b||)2 ,

from where the desired result follows.

❑
We now consider a generalisation of the Euclidean norm. Given p > 1 and x ∈ Rn we put
!1/p
n
X
p
||x||p = |xk | (4.20)
k=1
Rn 97

Clearly
||x||p ≥ 0 (4.21)
||x||p = 0 ⇔ x = 0 (4.22)
||αx||p = |α|||x||p , α ∈ R (4.23)
We now prove analogues of the Cauchy-Bunyakovsky-Schwarz and the Triangle Inequality for ||·||p . For this we need
the following lemma.

1 1
423 Lemma (Young’s Inequality) Let p > 1 and put + = 1. Then for (a, b) ∈ ([0; +∞[)2 we have
p q

ap bq
ab ≤ + .
p q

Proof: Let 0 < k < 1, and consider the function

[0; +∞[ → R
f: .
k
x 7→ x − k(x − 1)

Then 0 = f ′ (x) = kxk−1 − k ⇔ x = 1. Since f ′′ (x) = k(k − 1)xk−2 < 0 for 0 < k < 1, x ≥ 0, x = 1
1
is a maximum point. Hence f(x) ≤ f(1) for x ≥ 0, that is xk ≤ 1 + k(x − 1). Letting k = and
p
p
a
x = q we deduce
b  
a 1 ap
≤ 1+ −1 .
bq/p p bq
Rearranging gives
ap b1+p/q−p b1+p/q
ab ≤ b1+p/q + −
p p
from where we obtain the inequality. ❑

The promised generalisation of the Cauchy-Bunyakovsky-Schwarz Inequality is given in the following theorem.

424 Theorem (Hölder Inequality) Given (x, y) ∈ (Rn )2 the following inequality holds

|x•y| ≤ ||x||p ||y||q .

Proof: If ||x||p = 0 or ||y||q = 0 there is nothing to prove, so assume otherwise. From the Young
Inequality we have
|xk | |yk | |xk |p |yk |q
≤ p + q .
||x||p ||y||q ||x||p p ||y||q q
Adding, we deduce

Pn |xk | |yk | 1 Pn 1 Pn
k=1 ≤ p k=1 |xk |p + q k=1 |yk |q
||x||p ||y||q ||x||p p ||y||q q
||x||p p ||y||q q
= +
||x||p p p ||y||q q q
1 1
= +
p q
= 1.
98 Chapter 4

This gives
n
X
|xk yk | ≤ ||x||p ||y||q .
k=1

The result follows by observing that

 
Xn  n
X
 
 xk yk  ≤ |xk yk | ≤ ||x||p ||y||q .
 
k=1 k=1

As a generalisation of the Triangle Inequality we have

425 Theorem (Minkowski Inequality) Let p ∈]1; +∞[. Given (x, y) ∈ (Rn )2 the following inequality holds

||x + y||p ≤ ||x||p + ||y||p .

Proof: From the triangle inequality for real numbers 1.6

|xk + yk |p = |xk + yk ||xk + yk |p−1 ≤ (|xk | + |yk |) |xk + yk |p−1 .

Adding
n
X n
X n
X
|xk + yk |p ≤ |xk ||xk + yk |p−1 + |yk ||xk + yk |p−1 . (4.24)
k=1 k=1 k=1

By the Hölder Inequality

Pn Pn
1/p Pn
1/q
k=1 |xk ||xk + yk |p−1 ≤ k=1 |xk |p k=1 |xk + yk |(p−1)q
Pn
1/p Pn
1/q
= |xk |p |xk + yk |p (4.25)
k=1 k=1

p/q
= ||x||p ||x + y||p

In the same manner we deduce

n
X p/q
|yk ||xk + yk |p−1 ≤ ||y||p ||x + y||p . (4.26)
k=1

Hence (4.24) gives

n
X
||x + y||p
p = |xk + yk |p ≤ ||x||p ||x + y||p/q
p + ||y||p ||x + y||p/q
p ,
k=1

from where we deduce the result. ❑

Pn P n
426 Problem Prove Lagrange’s identity: i=1 ai = 0. Prove that 1≤i<j≤n ai • aj = − .
2
P
2 P
P

1≤j≤n a j bj = 1≤j≤n a2j 1≤j≤n b2j 428 Problem Let ak > 0. Use the CBS Inequality to shew
P that ! !
− 1≤k<j≤n (ak bj − aj bk )2 Xn Xn
2 1
ak ≥ n2 .
a2k
and then deduce the CBS Inequality in Rn . k=1 k=1

427 Problem Let ai ∈ Rn for 1 ≤ i ≤ n be unit vectors with 429 Problem Let ak ≥ 0, 1 ≤ k ≤ n be arbitrary. Prove
Rn 99

that !2
n n
X n(n + 1)(2n + 1) X a2k
ak ≤ .
6 k2
k=1 k=1
Chapter 5
Vector Spaces

5.1 Vector Spaces

430 Definition A vector space hV, +, ·, Fi over a field hF, +, i is a non-empty set V whose elements are called
vectors, possessing two operations + (vector addition), and · (scalar multiplication) which satisfy the following
axioms.
∀(a, b, c) ∈ V 3 , ∀(α, β) ∈ F2 ,

VS1 Closure under vector addition :

a + b ∈ V, (5.1)

VS2 Closure under scalar multiplication

αa ∈ V, (5.2)

VS3 Commutativity
a+b=b+a (5.3)

VS4 Associativity
(a + b) + c = a + (b + c) (5.4)

VS5 Existence of an additive identity

∃ 0∈ V : a+0=a+0=a (5.5)

VS6 Existence of additive inverses

∃ − a ∈ V : a + (−a) = (−a) + a = 0 (5.6)

VS7 Distributive Law

α(a + b) = αa + αb (5.7)

VS8 Distributive Law

(α + β)a = αa + βa (5.8)

VS9
1F a = a (5.9)

VS10
(αβ)a = α(βa) (5.10)

100
Vector Spaces 101

431 Example If n is a positive integer, then hFn , +, ·, Fi is a vector space by defining

(a1 , a2 , . . . , an ) + (b1 , b2 , . . . , bn ) = (a1 + b1 , a2 + b2 , . . . , an + bn ),

λ(a1 , a2 , . . . , an ) = (λa1 , λa2 , . . . , λan ).

In particular, hZ22 , +, ·, Z2 i is a vector space with only four elements and we have seen the two-dimensional and
tridimensional spaces hR2 , +, ·, Ri and hR3 , +, ·, Ri.

432 Example hMm×n (F), +, ·, Fi is a vector space under matrix addition and scalar multiplication of matrices.

433 Example If
F[x] = {a0 + a1 x + a2 x + · · · + an xn : ai ∈ F, n ∈ N}
denotes the set of polynomials with coefficients in a field hF, +, i then hF[x], +, ·, Fi is a vector space, under
polynomial addition and scalar multiplication of a polynomial.

434 Example If
Fn [x] = {a0 + a1 x + a2 x + · · · + ak xk : ai ∈ F, n ∈ N, k ≤ n}
denotes the set of polynomials with coefficients in a field hF, +, i and degree at most n, then hFn [x], +, ·, Fi is a
vector space, under polynomial addition and scalar multiplication of a polynomial.

435 Example Let k ∈ N and let Ck (R[a;b] ) denote the set of k-fold continuously differentiable real-valued functions
defined on the interval [a; b]. Then Ck (R[a;b] ) is a vector space under addition of functions and multiplication of a
function by a scalar.

436 Example Let p ∈]1; +∞[. Consider the set of sequences {an }∞
n=0 , an ∈ C,
 ∞

X
p ∞ p
l = {an }n=0 : |an | < +∞ .
n=0

Then lp is a vector space by defining addition as termwise addition of sequences and scalar multiplication as termwise
multiplication:
{an }∞ ∞ ∞
n=0 + {bn }n=0 = {(an + bn )}n=0 ,

λ{an }∞ ∞
n=0 = {λan }n=0 , λ ∈ C.

All the axioms of a vector space follow trivially from the fact that we are adding complex P∞numbers, except that
p
we
P∞ must prove that in l there is closure under addition and scalar multiplication. Since n=0 |an |p < +∞ =⇒
p
n=0 |λan | < +∞ closure under scalar multiplication follows easily. To prove closure under addition, observe that
if z ∈ C then |z| ∈ R+ and so by the Minkowski Inequality Theorem 425 we have

P 1/p P 1/p P 1/p

N N N
n=0 |an + bn |p ≤ n=0 |an |p + n=0 |bn |p
(5.11)
P∞

p 1/p
P∞

p 1/p
≤ n=0 |an | + n=0 |bn | .

This in turn implies that the series on the left in (5.11) converges, and so we may take the limit as N → +∞
obtaining
∞
!1/p ∞
!1/p ∞
!1/p
X X X
p p p
|an + bn | ≤ |an | + |bn | . (5.12)
n=0 n=0 n=0
p p
Now (5.12) implies that the sum of two sequences in l is also in l , which demonstrates closure under addition.

437 Example The set √ √

V = {a + b 2 + c 3 : (a, b, c) ∈ Q3 }
102 Chapter 5

with addition defined as

√ √ √ √ √ √
(a + b 2 + c 3) + (a ′ + b ′ 2 + c ′ 3) = (a + a ′ ) + (b + b ′ ) 2 + (c + c ′ ) 3,

and scalar multiplication defined as

√ √ √ √
λ(a + b 2 + c 3) = (λa) + (λb) 2 + (λc) 3,

constitutes a vector space over Q.

438 Theorem In any vector space hV, +, ·, Fi,

∀ α ∈ F, α0 = 0.

Proof: We have
α0 = α(0 + 0) = α0 + α0.
Hence
α0 − α0 = α0,
or
0 = α0,
proving the theorem. ❑

439 Theorem In any vector space hV, +, ·, Fi,

∀ v ∈ V, 0F v = 0.

Proof: We have
0F v = (0F + 0F )v = 0F v + 0F v.
Therefore
0F v − 0F v = 0F v,
or
0 = 0F v,
proving the theorem. ❑

440 Theorem In any vector space hV, +, ·, Fi, α ∈ F, v ∈ V,

αv = 0 =⇒ α = 0F ∨ v = 0.

Proof: Assume that α 6= 0F . Then α possesses a multiplicative inverse α−1 such that α−1 α = 1F .
Thus
αv = 0 =⇒ α−1 αv = α−1 0.
By Theorem 439, α−1 0 = 0. Hence
α−1 αv = 0.
Since by Axiom 5.9, we have α−1 αv = 1F v = v, and so we conclude that v = 0. ❑

441 Theorem In any vector space hV, +, ·, Fi,

∀α ∈ F, ∀ v ∈ V, (−α)v = α(−v) = −(αv).

Vector Subspaces 103

Proof: We have
0F v = (α + (−α))v = αv + (−α)v,
whence
−(αv) + 0F v = (−α)v,
that is
−(αv) = (−α)v.
Similarly,
0 = α(v − v) = αv + α(−v),
whence
−(αv) + 0 = α(−v),
that is
−(αv) = α(−v),
proving the theorem. ❑

442 Problem Is R2 with vector addition and scalar multipli- is a vector space over F if vector addition is defined as
cation defined as a ⊕ b = ab, (a, b) ∈ (R+ )2 and scalar multiplication is
2 3 2 3 2 3 2 3 2 3 defined as α ⊗ a = aα , (α, a) ∈ (R, R+ ).
6x1 7 6y1 7 6x1 + y1 7 6x1 7 6λx1 7
4 5+4 5=4 5, λ4 5 = 4 5
445 Problem Let C denote the complex numbers and R denote
x2 y2 x2 + y2 x2 0
the real numbers. Is C a vector space over R under ordinary
a vector space? addition and multiplication? Is R a vector space over C?

443 Problem Demonstrate that the commutativity axiom 5.3 446 Problem Construct a vector space with exactly 8 ele-
is redundant. ments.

444 Problem Let V = R+ =]0; +∞[, the positive real num- 447 Problem Construct a vector space with exactly 9 ele-
bers and F = R, the real numbers. Demonstrate that V ments.

5.2 Vector Subspaces

448 Definition Let hV, +, ·, Fi be a vector space. A non-empty subset U ⊆ V which is also a vector space under
the inherited operations of V is called a vector subspace of V.

449 Example Trivially, X1 = {0} and X2 = V are vector subspaces of V.

450 Theorem Let hV, +, ·, Fi be a vector space. Then U ⊆ V, U 6= ∅ is a subspace of V if and only if ∀α ∈ F
and ∀(a, b) ∈ U2 it is verified that
a + αb ∈ U.

Proof: Observe that U inherits commutativity, associativity and the distributive laws from V.
Thus a non-empty U ⊆ V is a vector subspace of V if (i) U is closed under scalar multiplication,
that is, if α ∈ F and v ∈ U, then αv ∈ U; (ii) U is closed under vector addition, that is, if
(u, v) ∈ U2 , then u + v ∈ U. Observe that (i) gives the existence of inverses in U, for take
α = −1F and so v ∈ U =⇒ −v ∈ U. This coupled with (ii) gives the existence of the zero-vector,
for 0 = v − v ∈ U. Thus we need to prove that if a non-empty subset of V satisfies the property
stated in the Theorem then it is closed under scalar multiplication and vector addition, and vice-
versa, if a non-empty subset of V is closed under scalar multiplication and vector addition, then
it satisfies the property stated in the Theorem. But this is trivial. ❑

451 Example Shew that X = {A ∈ Mn (F) : tr (A) = 0F } is a subspace of Mn (F).

104 Chapter 5

Solution: Take A, B ∈ X, α ∈ R. Then

tr (A + αB) = tr (A) + αtr (B) = 0F + α(0F ) = 0F .

Hence A + αB ∈ X, meaning that X is a subspace of Mn (F).

452 Example Let U ∈ Mn (F) be an arbitrary but fixed. Shew that

CU = {A ∈ Mn (F) : AU = UA}

is a subspace of Mn (F).

Solution: Take (A, B) ∈ (CU )2 . Then AU = UA and BU = UB. Now

(A + αB)U = AU + αBU = UA + αUB = U(A + αB),

meaning that A + αB ∈ CU . Hence CU is a subspace of Mn (F). CU is called the commutator of U.

453 Theorem Let X ⊆ V, Y ⊆ V be vector subspaces of a vector space hV, +, ·, Fi. Then their intersection X ∩ Y
is also a vector subspace of V.

Proof: Let α ∈ F and (a, b) ∈ (X ∩ Y)2 . Then clearly (a, b) ∈ X and (a, b) ∈ Y. Since X is a
vector subspace, a + αb ∈ X and since Y is a vector subspace, a + αb ∈ Y. Thus

a + αb ∈ X ∩ Y

and so X ∩ Y is a vector subspace of V by virtue of Theorem 450. ❑

☞ We we will soon see that the only vector subspaces of hR , +, ·, Ri are the set containing the
2

zero-vector, any line through the origin, and R itself. The only vector subspaces of hR3 , +, ·, Ri
2

are the set containing the zero-vector, any line through the origin, any plane containing the origin
and R3 itself.

454 Problem Prove that 456 Problem Let a ∈ Rn be a fixed vector. Demonstrate that
2 3 





 X = {x ∈ Rn : a•x = 0}

 6a7 



 6 7 

 is a subspace of Rn .

 6 7 

 6b7
 

6 7 4
X = 6 7 ∈ R : a − b − 3d = 0

 6 7 
 457 Problem Let a ∈ R3 be a fixed vector. Demonstrate that
6 c 7
 


 6 7 
 X = {x ∈ R3 : a×x = 0}


 4 5 



 

 d 
is a subspace of R3 .

is a vector subspace of R4 .
458 Problem Let A ∈ Mm×n (F) be a fixed matrix. Demon-
strate that
455 Problem Prove that
2 3  S = {X ∈ Mn×1 (F) : AX = 0m×1 }

 


 6 a 7 
 is a subspace of Mn×1 (F).
 

 6 7 



 6 7 



 62a − 3b7 


 6 7 
 459 Problem Prove that the set X ⊆ Mn (F) of upper trian-
6
 7 

6 7 gular matrices is a subspace of Mn (F).
X = 6 5b 7 : a, b ∈ R

 6 7 


 6 7 


 6 7 


 6 a + 2b 7 
 460 Problem Prove that the set X ⊆ Mn (F) of symmetric


 6 7 

 matrices is a subspace of Mn (F).


 4 5 



 

a
461 Problem Prove that the set X ⊆ Mn (F) of skew-
is a vector subspace of R5 . symmetric matrices is a subspace of Mn (F).
Linear Independence 105

2 3 
462 Problem Prove that the following subsets are not sub- 
 

 a b7 
spaces of the given vector space. Here you must say which of 6 2 2
➌ 4 5 : (a, b) ∈ R , a + b = 0 ⊆ M2×2 (R)
the axioms for a vector space fail. 
 

 0 0 

2 3  463 Problem Let hV, +, ·, Fi be a vector space, and let U1 ⊆


 a 


 6 7 
 V and U2 ⊆ V be vector subspaces. Prove that if U1 ∪ U2
 
6 7
 
 is a vector subspace of V, then either U1 ⊆ U2 or U2 ⊆ U1 .
6 7 2 2
➊ 6b7 : a, b ∈ R, a + b = 1 ⊆ R3

 6 7 



 4 5 



 0 
 464 Problem Let V a vector space over a field F. If F is in-
finite, show that V is not the set-theoretic union of a finite
number of proper subspaces.
2 3 

 
 465 Problem Give an example of a finite vector space V over

6 7a 


 6 7 
 a finite field F such that
 
6 7
➋ 6b7 : a, b ∈ R , ab = 0 ⊆ R3
2
6 7
 
 V = V1 ∪ V2 ∪ V3 ,
4 5






 0 

where the Vk are proper subspaces.

5.3 Linear Independence

466 Definition Let (λ1 , λ2 , · · · , λn ) ∈ Fn . Then the vectorial sum
n
X
λj aj
j=1

is said to be a linear combination of the vectors ai ∈ V, 1 ≤ i ≤ n.

2 3
6a b7
467 Example Any matrix 6
4
7 ∈ M2 (R) can be written as a linear combination of the matrices
5
c d
2 3 2 3 2 3 2 3
61 07 60 17 60 0 7 60 07
6 7, 6 7, 6 7, 6 7,
4 5 4 5 4 5 4 5
0 0 0 0 1 0 0 1

for 2 3 2 3 2 3 2 3 2 3
6a b7 61 07 60 17 60 07 60 07
6 7 = a6 7 +b6 7 + c6 7 + d6 7.
4 5 4 5 4 5 4 5 4 5
c d 0 0 0 0 1 0 0 1

468 Example Any polynomial of degree at most 2, say a + bx + cx2 ∈ R2 [x] can be written as a linear combination
of 1, x − 1, and x2 − x + 2, for
a + bx + cx2 = (a − c)(1) + (b + c)(x − 1) + c(x2 − x + 2).

Generalising the notion of two parallel vectors, we have

469 Definition The vectors ai ∈ V, 1 ≤ i ≤ n, are linearly dependent or tied if

n
X
∃(λ1 , λ2 , · · · , λn ) ∈ Fn \ {0} such that λj aj = 0,
j=1

that is, if there is a non-trivial linear combination of them adding to the zero vector.
106 Chapter 5

470 Definition The vectors ai ∈ V, 1 ≤ i ≤ n, are linearly independent or free if they are not linearly dependent.
That is, if (λ1 , λ2 , · · · , λn ) ∈ Fn then
n
X
λj aj = 0 =⇒ λ1 = λ2 = · · · = λn = 0F .
j=1

☞ A family of vectors is linearly independent if and only if the only linear combination of them
giving the zero-vector is the trivial linear combination.

471 Example 2 3 2 3 2 3 

 1 



 6 7 647 677
6 7 6 7 6 7 
 
6 7 6 7 6 7
6 7,6 7,6 7
 627 657 687

 6 7 6 7 6 7

 4 5 4 5 4 5


 3 
6 9 

is a tied family of vectors in R3 , since

2 3 2 3 2 3 2 3
617 647 67 7 60 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
(1) 627 + (−2) 657 + (1) 6
6 7 6 7 7 6 7
68 7 = 60 7 .
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
3 6 9 0

472 Example Let u, v be linearly independent vectors in some vector space over a field F with characteristic different
from 2. Shew that the two new vectors x = u − v and y = u + v are also linearly independent.

Solution: Assume that a(u − v) + b(u + v) = 0. Then

(a + b)u + (a − b)v = 0.

Since u, v are linearly independent, the above coefficients must be 0, that is, a + b = 0F and a − b = 0F . But
this gives 2a = 2b = 0F , which implies a = b = 0F , if the characteristic of the field is not 2. This proves the linear
independence of u − v and u + v.

473 Theorem Let A ∈ Mm×n (F). Then the columns of A are linearly independent if and only the only solution
to the system AX = 0m is the trivial solution.

Proof: Let A1 , . . . , An be the columns of A. Since

x1 A1 + x2 A2 + · · · + xn An = AX,

the result follows. ❑

474 Theorem Any family

{0, u1 , u2 , . . . , uk }
containing the zero-vector is linearly dependent.

Proof: This follows at once by observing that

1F 0 + 0F u1 + 0F u2 + . . . , +0F uk = 0

is a non-trivial linear combination of these vectors equalling the zero-vector. ❑

Spanning Sets 107

475 Problem Shew that 481 Problem Let {v1 , v2 , v3 , v4 } be a linearly independent
2 3 2 3 2 3 family of vectors. Prove that the family

 



 617 617 617 

6 7 6 7 6 7
  {v1 + v2 , v2 + v3 , v3 + v4 , v4 + v1 }
6 7 6 7 6 7
6 0 7 61 7 6 1 7
, ,

 6 7 6 7 6 7 
4 5 4 5 4 5

  is not linearly independent.
 

 0 0 1 

forms a free family of vectors in R3 . 482 Problem Let {v1 , v2 , v3 } be linearly independent vectors
in R5 . Are the vectors

476 Problem Prove that the set b1 = 3v1 + 2v2 + 4v3 ,

2 3 2 3 2 3 2 3

 

 

 617 6 1 7 6 1 7 617  b2 = v1 + 4v2 + 2v3 ,

 6 7 6 7
6 7 6 7 6 7 6 76 7 6 7


 

617 6 1 7 6−17 617
  b3 = 9v1 + 4v2 + 3v3 ,
6 7 6 7 6 7 6 7
6 7,6 7,6 7,6 7

 6 7 6 7 6 7 6 7

 617 6−17 6 1 7 607 
 b4 = v1 + 2v2 + 5v3 ,

 6 7 6 7 6 7 6 7 


 4 5 4 5 4 5 4 5

 linearly independent? Prove or disprove!

 
 1 −1 −1 1 

√
4 483 Problem Is the family {1, 2} linearly independent over
2 3 independent set of vectors in R and shew that
is a linearly
Q?
617
6 7 √
6 7
627 484 Problem Is the family {1, 2} linearly independent over
6 7 R?
X = 6 7 can be written as a linear combination of these
6 7
617
6 7
4 5 485 Problem Consider the vector space
1 √ √
vectors. V = {a + b 2 + c 3 : (a, b, c) ∈ Q3 }.
√ √
477 Problem Let (a, b) ∈ (R3 )2 and assume that a•b = 0 1. Shew that {1, 2, 3} are linearly independent over Q.
and that a and b are linearly independent. Prove that
a, b, a×b are linearly independent. 2. Express
1 2
√ + √
478 Problem Let ai ∈ Rn , 1 ≤ i ≤ k (k ≤ n) be k non-zero 1− 2 12 − 2
vectors such that ai •aj = 0 for i =
6 j. Prove that these k √ √
as a linear combination of {1, 2, 3}.
vectors are linearly independent.

479 Problem Let (u, v) ∈ (Rn )2 . Prove that |u•v| = ||u||||v||

if and only if u and v are linearly dependent. 486 Problem Let f, g, h belong to C∞ (RR ) (the space of in-
finitely continuously differentiable real-valued functions de-
fined on the real line) and be given by
480 Problem Prove that
2 3 2 3 2 3 2 3

  f(x) = ex , g(x) = e2x , h(x) = e3x .
 1 0
6 7 61 0 7 60 17 6 0 17
4 5,4 5,4 5,4 5

 
 Shew that f, g, h are linearly independent over R.
 0 1 0 −1 1 0 −1 0 
2 3
487 Problem Let f, g, h belong to C∞ (RR ) be given by
61 17
is a linearly independent family over R. Write 4 5 as a
f(x) = cos2 x, g(x) = sin2 x, h(x) = cos 2x.
1 1
linear combination of these matrices. Shew that f, g, h are linearly dependent over R.

5.4 Spanning Sets

488 Definition A family {u1 , u2 , . . . , uk , . . . , } ⊆ V is said to span or generate V if every v ∈ V can be written
as a linear combination of the uj ’s.
108 Chapter 5

489 Theorem If {u1 , u2 , . . . , uk , . . . , } ⊆ V spans V, then any superset

{w, u1 , u2 , . . . , uk , . . . , } ⊆ V

also spans V.

Proof: This follows at once from

l
X l
X
λi ui = 0F w + λi ui .
i=1 i=1

490 Example The family of vectors  2 3 2 3 2 3


 



 617 60 7 607

 6 7 6 7 6 7
6 7 6 7 6 7
6 7 6 7 6 7
i = 607 , j = 617 , k = 607

 6 7 6 7 6 7

 4 5 4 5 4 5

  
 0 0 1 
2 3
6a7
6 7
6 7
spans R since given 6
3 7 3
6b7 ∈ R we may write
6 7
4 5
c

2 3
6a7
6 7
6 7
6 7 = ai + bj + ck.
6 b7
6 7
4 5
c

491 Example Prove that the family of vectors

 2 3 2 3 2 3

 



 61 7 617 617

 6 7 6 7 6 7
6 7 6 7 6 7
t1 = 6 7 , t2 = 6 7 , t3 = 6 7
 60 7 617 617

 6 7 6 7 6 7

 4 5 4 5 4 5

 
0 0 1 

spans R3 .

Solution: This follows from the identity

2 3 2 3 2 3 2 3
6a7 617 61 7 617
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 7 = (a − b) 6 7 + (b − c) 6 7 + c 6 7 = (a − b)t1 + (b − c)t2 + ct3 .
6 b7 607 617 61 7
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
c 0 0 1
Spanning Sets 109

492 Example Since 2 3 2 3 2 3 2 3 2 3

6a b7 61 07 60 17 60 07 60 07
6 7 = a6 7 + b6 7 +c6 7 +d6 7
4 5 4 5 4 5 4 5 4 5
c d 0 0 0 0 1 0 0 1
2 3 2 3 2 3 2 3
61 07 60 17 60 07 60 07
the set of matrices 6
4
7, 6
5 4
7, 6
5 4
7, 6
5 4
7 is a spanning set for M2 (R)
5
0 0 0 0 1 0 0 1

493 Example The set

{1, x, x2 , x3 , . . . , xn , . . .}
spans R[x], the set of polynomials with real coefficients and indeterminate x.

494 Definition The span of a family of vectors {u1 , u2 , . . . , uk , . . . , } is the set of all finite linear combinations
obtained from the ui ’s. We denote the span of {u1 , u2 , . . . , uk , . . . , } by
span (u1 , u2 , . . . , uk , . . . , ) .

495 Theorem Let hV, +, ·, Fi be a vector space. Then

span (u1 , u2 , . . . , uk , . . . , ) ⊆ V
is a vector subspace of V.

Proof: Let α ∈ F and let

l
X l
X
x= ak uk , y = bk uk ,
k=1 k=1
be in span (u1 , u2 , . . . , uk , . . . , ) (some of the coefficients might be 0F ). Then
l
X
x + αy = (ak + αbk )uk ∈ span (u1 , u2 , . . . , uk , . . . , ) ,
k=1

and so span (u1 , u2 , . . . , uk , . . . , ) is a subspace of V. ❑

496 Corollary span (u1 , u2 , . . . , uk , . . . , ) ⊆ V is the smallest vector subspace of V (in the sense of set inclusion)
containing the set
{u1 , u2 , . . . , uk , . . . , }.

Proof: If W ⊆ V is a vector subspace of V containing the set

{u1 , u2 , . . . , uk , . . . , }
then it contains every finite linear combination of them, and hence, it contains span (u1 , u2 , . . . , uk , . . . , ) .
❑
02 3 2 3 2 31
B61 07 60 0 7 6 0 1 7C
497 Example If A ∈ M2 (R), A ∈ span B
4
6 7,6
5 4
7,6
5 4
7C then A has the form
5A
0 0 0 1 1 0
2 3 2 3 2 3 2 3
61 07 60 07 60 17 6a c7
a6
4
7 + b6
5 4
7 +c6
5 4
7=6
5 4
7,
5
0 0 0 1 1 0 c b

i.e., this family spans the set of all symmetric 2 × 2 matrices over R.
110 Chapter 5

498 Theorem Let V be a vector space over a field F and let (v, w) ∈ V 2 , γ ∈ F \ {0F }. Then

span (v, w) = span (v, γw) .

Proof: The equality

av + bw = av + (bγ−1 )(γw),

proves the statement. ❑

499 Theorem Let V be a vector space over a field F and let (v, w) ∈ V 2 , γ ∈ F. Then

span (v, w) = span (w, v + γw) .

Proof: This follows from the equality

av + bw = a(v + γw) + (b − aγ)w.

500 Problem Let R3 [x] denote the set of polynomials with de- 503 Problem Prove that
gree at most 3 and real coefficients. Prove that the set 02 3 2 3 2 3 2 31
B61 07 61 0 7 60 17 6 0 1 7C
2 3
{1, 1 + x, (1 + x) , (1 + x) }
span 4 5,4 5,4 5,4 5A = M2 (R).
spans R3 [x]. 0 1 0 −1 1 0 −1 0
2 3 02 3 2 31
6 1 7 B6 1 7 6 0 7C 504 Problem For the vectors in R3 ,
6 7 B6 7 6 7C
6 7 B6 7 6 7C 2 3 2 3 2 3 2 3
501 Problem Shew that 6 1 7 ∈
6 span B6 0 7 , 6 1 7C.
6 7 B6 7 6 7C 61 7 617 61 7 637
4 5 4 5 4 5A 6 7 6 7 6 7 6 7
−1 −1 −1 6 7 6 7 6 7 6 7
a = 627 , b = 637 , c = 617 , d = 687 ,
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
02 3 2 3 2 31 1 2 0 5
B61 0 7 60 07 6 0 1 7C
502 Problem What is span 4 5,4 5,4 5A? prove that
0 0 0 1 −1 0 span (a, b) = span (c, d) .

5.5 Bases

505 Definition A family {u1 , u2 , . . . , uk , . . .} ⊆ V is said to be a basis of V if (i) they are linearly independent,
(ii) they span V.
Bases 111

506 Example The family 2 3

6 0F 7
6 7
6 . 7
6 . 7
6 . 7
6 7
6 7
6 7
6 0F 7
6 7
6 7
6 7
ei = 6 1 F 7 ,
6 7
6 7
6 7
60 7
6 F7
6 7
6 . 7
6 .. 7
6 7
6 7
4 5
0F

where there is a 1F on the i-th slot and 0F ’s on the other n − 1 positions, is a basis for Fn .

507 Theorem Let hV, +, ·, Fi be a vector space and let

U = {u1 , u2 , . . . , uk , . . .} ⊆ V

be a family of linearly independent vectors in V which is maximal in the sense that if U ′ is any other family of
vectors of V properly containing U then U ′ is a dependent family. Then U forms a basis for V.

Proof: Since U is a linearly independent family, we need only to prove that it spans V. Take
v ∈ V. If v ∈ U then there is nothing to prove, so assume that v ∈ V \ U. Consider the set
U ′ = U ∪ {v}. This set properly contains U, and so, by assumption, it forms a dependent family.
There exists scalars α0 , α1 , . . . , αn such that

α0 v + α1 u1 + · · · + αn un = 0.

Now, α0 6= 0F , otherwise the ui would be linearly dependent. Hence α−1

0 exists and we have

v = −α−1
0 (α1 u1 + · · · + αn un ),

and so the ui span V. ❑

☞ From Theorem 507 it follows that to shew that a vector space has a basis it is enough to shew
that it has a maximal linearly independent set of vectors. Such a proof requires something called
Zörn’s Lemma, and it is beyond our scope. We dodge the whole business by taking as an axiom
that every vector space possesses a basis.

508 Theorem (Steinitz Replacement Theorem) Let hV, +, ·, Fi be a vector space and let U = {u1 , u2 , . . .} ⊆ V.
Let W = {w1 , w2 , . . . , wk } be an independent family of vectors in span (U). Then there exist k of the ui ’s, say
{u1 , u2 , . . . , uk } which may be replaced by the wi ’s in such a way that

span (w1 , w2 , . . . , wk , uk+1 , . . .) = span (U) .

Proof: We prove this by induction on k. If k = 1, then

w1 = α1 u1 + α2 u2 + · · · + αn un

for some n and scalars αi . There is an αi 6= 0F , since otherwise w1 = 0 contrary to the assumption
that the wi are linearly independent. After reordering, we may assume that α1 6= 0F . Hence

u1 = α−1
1 (w1 − (α2 u2 + · · · + αn un )),
112 Chapter 5

and so u1 ∈ span (w1 , u2 , . . . , ) and

span (w1 , u2 , . . . , ) = span (u1 , u2 , . . . , ) .

Assume now that the theorem is true for any set of fewer than k independent vectors. We may
thus assume that that {u1 , . . .} has more than k − 1 vectors and that

span (w1 , w2 , . . . , wk−1 , uk , , . . .) = span (U) .

Since wk ∈ U we have

wk = β1 w1 + β2 w2 + · · · + βk−1 wk−1 + γk uk + γk+1uk+1 + γm um .

If all the γi = 0F , then the {w1 , w2 , . . . , wk } would be linearly dependent, contrary to assumption.
Thus there is a γi 6= 0F , and after reordering, we may assume that γk 6= 0F . We have therefore

uk = γ−1
k (wk − (β1 w1 + β2 w2 + · · · + βk−1 wk−1 + γk+1 uk+1 + γm um )).

But this means that

span (w1 , w2 , . . . , wk , uk+1 , . . .) = span (U) .
This finishes the proof. ❑

509 Corollary Let {w1 , w2 , . . . , wn } be an independent family of vectors with V = span (w1 , w2 , . . . , wn ). If we
also have V = span (u1 , u2 , . . . , uν ), then
1. n ≤ ν,
2. n = ν if and only if the {y1 , y2 , . . . , yν } are a linearly independent family.
3. Any basis for V has exactly n elements.

Proof:
1. In the Steinitz Replacement Theorem 508 replace the first n ui ’s by the wi ’s and n ≤ ν
follows.
2. If {u1 , u2 , . . . , uν } are a linearly independent family, then we may interchange the rôle of the
wi and ui obtaining ν ≤ n. Conversely, if ν = n and if the ui are dependent, we could
express some ui as a linear combination of the remaining ν − 1 vectors, and thus we would
have shewn that some ν − 1 vectors span V. From (1) in this corollary we would conclude
that n ≤ ν − 1, contradicting n = ν.
3. This follows from the definition of what a basis is and from (2) of this corollary.
❑

510 Definition The dimension of a vector space hV, +, ·, Fi is the number of elements of any of its bases, and we
denote it by dim V.

511 Theorem Any linearly independent family of vectors

{x1 , x2 , . . . , xk }

in a vector space V can be completed into a family

{x1 , x2 , . . . , xk , yk+1 , yk+2 , . . .}

so that this latter family become a basis for V.

Proof: Take any basis {u1 , . . . , uk , uk+1 , . . . , } and use Steinitz Replacement Theorem 508. ❑
Bases 113

512 Corollary If U ⊆ V is a vector subspace of a finite dimensional vector space V then dim U ≤ dim V.

Proof: Since any basis of U can be extended to a basis of V, it follows that the number of
elements of the basis of U is at most as large as that for V. ❑

513 Example Find a basis and the dimension of the space generated by the set of symmetric matrices in Mn (R).

Solution: Let Eij ∈ Mn (R) be the n × n matrix with a 1 on the ij-th position and 0’s everywhere else. For

n(n − 1)
1 ≤ i < j ≤ n, consider the n 2
= matrices Aij = Eij + Eji . The Aij have a 1 on the ij-th and ji-th
2
position and 0’s everywhere else. They, together with the n matrices Eii , 1 ≤ i ≤ n constitute a basis for the space
of symmetric matrices. The dimension of this space is thus

n(n − 1) n(n + 1)
+n= .
2 2

514 Theorem Let {u1 , . . . , un } be vectors in Rn . Then the u’s form a basis if and only if the n × n matrix A
formed by taking the u’s as the columns of A is invertible.

Proof: Since we have the right number of vectors, it is enough to prove that the u’s are linearly
2 3
6 x1 7
6 7
6 7
6x 7
6 27
independent. But if X = 6 7
6 . 7, then
6 . 7
6 . 7
6 7
4 5
xn

x1 u1 + · · · + xn un = AX.

If A is invertible, then AX = 0n =⇒ X = A−1 0n = 0n , meaning that x1 = x2 = · · · xn = 0, so the

u’s are linearly independent.

Conversely, assume that the u’s are linearly independent. Then the equation AX = 0n has a unique
solution. Let r = rank (A) and let (P, Q) ∈ (GLn (R))2 be matrices such that A = P−1 Dn,n,r Q−1 ,
where Dn,n,r is the Hermite normal form of A. Thus

AX = 0n =⇒ P−1 Dn,n,r Q−1 X = 0n =⇒ Dn,n,r Q−1 X = 0n .

2 3
6 y1 7
6 7
6 7
6y 7
6 27
Put Q X = 6
−1 7
6 . 7. Then
6 . 7
6 . 7
6 7
4 5
yn

Dn,n,r Q−1 X = 0n =⇒ y1 e1 + · · · + yr er = 0n ,
where ej is the n-dimensional column vector with a 1 on the j-th slot and 0’s everywhere else. If
r < n then yr+1 , . . . , yn can be taken arbitrarily and so there would not be a unique solution, a
contradiction. Hence r = n and A is invertible. ❑
114 Chapter 5

515 Problem In problem 455 we saw that 519 Problem Prove that the dimension of the vector subspace
2 3  n(n + 1)
of lower triangular n × n matrices is and find a

 
 2

 6 a 7 
 basis for this space.
 

 6 7 

6
 7 



 6 2a − 3b 7 



 6 7 
 520 Problem Find a basis and the dimension of
6
 7 

6 7 0 2 3 2 3 2 31
X = 6 5b 7 : a, b ∈ R
6 7
 

 6 7 

 B 61 7 61 7 6 2 7C

 6 7 
 B 6 7 6 7 6 7C

 6 a + 2b 7 
 B 6 7 6 7 6 7C


 6 7 

 B 61 7 61 7 6 2 7C

 4 5 
 B 6 7 6 7 6 7C

 
 X = span Bv1 = 6 7 , v2 = 6 7 , v 3 = 6 7C .
 a  B 6 7 6 7 6 7C
B 61 7 61 7 6 2 7C
B 6 7 6 7 6 7C
 4 5 4 5 4 5A
is a vector subspace of R5 . Find a basis for this subspace.
1 0 1

516 Problem Let {v1 , v2 , v3 , v4 , v5 } be a basis for a vector

space V over a field F. Prove that 521 Problem Find a basis and the dimension of
0 2 3 2 3 2 31
{v1 + v2 , v2 + v3 , v3 + v4 , v4 + v5 , v5 + v1 }
B 61 7 61 7 6 2 7C
B 6 7 6 7 6 7C
is also a basis for V. B 6 7 6 7 6 7C
B 61 7 61 7 6 2 7C
B 6 7 6 7 6 7C
X = span Bv1 = 6 7 , v2 = 6 7 , v 3 = 6 7C .
B 6 7 6 7 6 7C
517 Problem Find a basis for the solution space of the system B 61 7 61 7 6 2 7C
B 6 7 6 7 6 7C
of n + 1 linear equations of 2n unknowns  4 5 4 5 4 5A
1 0 2
x1 + x2 + · · · + xn = 0,

x2 + x3 + · · · + xn+1 = 0, 522 Problem Find a basis and the dimension of

...... 0 2 3 2 3 2 3 2
...
B 61 07 61 0 7 60 17 61
xn+1 + xn+2 + · · · + x2n = 0. X = span v1 = 4 5, v2 = 4 5, v3 = 4 5 , v4 = 4
0 1 2 0 2 0 0
518 Problem Prove that the set
523 Problem Let (a, b) ∈ R3 × R3 be fixed. Solve the equa-
X = {(a, b, c, d)|b + 2c = 0} ⊆ R4
tion
is a vector subspace of R4 . Find its dimension and a basis for a×x = b,
X. for x.

5.6 Coordinates
524 Theorem Let {v1 , v2 , . . . , vn } be a basis for a vector space V. Then any v ∈ V has a unique representation
v = a1 v1 + a2 v2 + · · · + an vn .

Proof: Let
v = b1 v1 + b2 v2 + · · · + bn vn
be another representation of v. Then
0 = (a1 − b1 )v1 + (a2 − b2 )v2 + · · · + (an − bn )vn .
Since {v1 , v2 , . . . , vn } forms a basis for V, they are a linearly independent family. Thus we must
have
a1 − b1 = a2 − b2 = · · · = an − bn = 0F ,
that is
a1 = b1 ; a2 = b2 ; · · · ; an = bn ,
proving uniqueness. ❑
Coordinates 115

525 Definition An ordered basis {v1 , v2 , . . . , vn } of a vector space V is a basis where the order of the vk has been
fixed. Given an ordered basis {v1 , v2 , . . . , vn } of a vector space V, Theorem 524 ensures that there are unique
(a1 , a2 , . . . , an ) ∈ Fn such that
v = a1 v1 + a2 v2 + · · · + an vn .

The ak ’s are called the coordinates of the vector v.

2 3
61 7
6 7
6 7
526 Example The standard ordered basis for R is S = {i, j, k}. The vector 6
3 7 3
627 ∈ R for example, has coordinates
6 7
4 5
3
2 3
617
6 7
6 7
(1, 2, 3)S . If the order of the basis were changed to the ordered basis S1 = {i, k, j}, then 6 7 3
627 ∈ R would have
6 7
4 5
3
coordinates (1, 3, 2)S1 .

☞ Usually, when we give a coordinate representation for a vector v ∈ R , we assume that we n

are using the standard basis.

2 3
617
6 7
6 7
527 Example Consider the vector 6 7 3
627 ∈ R (given in standard representation). Since
6 7
4 5
3

2 3 2 3 2 3 2 3
617 617 617 6 17
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 7 = −1 6 7 − 1 6 7 + 3 6 7 ,
627 607 617 6 17
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
3 0 0 1

 2 3 2 3 2 3 2 3

 



 617 617 617
 617
6
 7 6 7 6 7
 6 7
6 7 6 7 6 7 6 7
under the ordered basis B1 = 6 7,6 7,6 7 ,
607 617 617
6 7 has coordinates (−1, −1, 3)B . We write
627 1

 6 7 6 7 6 7 6 7

 4 5 4 5 4 5
 4 5

 

 0 0 1  3

2 3 2 3
617 6−17
6 7 6 7
6 7 6 7
6 7=6 7 .
627 6−17
6 7 6 7
4 5 4 5
3 3
B1
116 Chapter 5

528 Example The vectors of

2 3 2 3 

 
617 617 
B1 = 6 7,6 7
4 5 4 5

 
 1 2 

are non-parallel, and so form a basis for R2 . So do the vectors

2 3 2 3

 
627 6 1 7 
6 7 6
B2 = 4 5 , 4 5 . 7

 

 1 −1 

2 3
63 7
Find the coordinates of 6
4 5
7 in the base B2 .
4
B1

Solution: We are seeking x, y such that

2 3 2 3 2 3 2 3 2 32 3 2 32 3
617 617 627 6 1 7 61 17 637 62 1 7 6x7
36
4 5
7 +46 7 = x6 7 +y6
4 5 4 5 4
7 =⇒ 6
5 4
76 7 = 6
54 5 4
76 7
54 5 .
1 2 1 −1 1 2 4 1 −1 y
B2

Thus
2 3 2 3−1 2 32 3
6x7 62 1 7 61 17 637
6 7 = 6 7 6 76 7
4 5 4 5 4 54 5
y 1 −1 1 2 4
B2
2 32 32 3
1 1
63 3 7 61 1 7 63 7
= 6
4
76
54
76 7
54 5
1
3
− 23 1 2 4
2 32 3
2
6 3
1 7 63 7
= 6
4
76 7
54 5
− 13 −1 4
2 3
6 6 7
= 6
4
7
5 .
−5
B2

Let us check by expressing both vectors in the standard basis of R2 :

2 3 2 3 2 3 2 3
637 617 617 6 7 7
6 7 = 36 7 6 7 6 7
4 5 4 5 +44 5 = 4 5,
4 1 2 11
B1

2 3 2 3 2 3 2 3
6 6 7 627 6 1 7 677
6 7 = 66 7 −56 7 = 6 7.
4 5 4 5 4 5 4 5
−5 1 −1 11
B2
Coordinates 117

In general let us consider bases B1 , B2 for the same vector space V. We want to convert XB1 to YB2 . We
let A be the matrix formed with the column vectors of B1 in the given order an B be the matrix formed with the
column vectors of B2 in the given order. Both A and B are invertible matrices since the B’s are bases, in view of
Theorem 514. Then we must have

AXB1 = BYB2 =⇒ YB2 = B−1 AXB1 .

Also,

XB1 = A−1 BYB2 .

This prompts the following definition.

529 Definition Let B1 = {u1 , u2 , . . . , un } and B2 = {v1 , v2 , . . . , vn } be two ordered bases for a vector space V.
Let A ∈ Mn (F) be the matrix having the u’s as its columns and let B ∈ Mn (F) be the matrix having the v’s as
its columns. The matrix P = B−1 A is called the transition matrix from B1 to B2 and the matrix P−1 = A−1 B is
called the transition matrix from B2 to B1 .

530 Example Consider the bases of R3

 2 3 2 3 2 3

 



 617 617 617 

6 7 6 7 6
 6 7 6 7 6 7
 7 
6 7 6 7 6 7
B1 = 617 , 617 , 607 ,

 6 7 6 7 6 7 
 
 4 5 4 5 4 5

 

 1 0 0 

2 3 2 3 2 3

 



 6 1 7 6 1 7 627 

6 7 6
6 7 6 7 6 7
 7 6 7 
6 7 6 7 6 7
B2 = 6 1 7 , 6−17 , 607 .

 6 7 6 7 6 7 
 
4 5 4 5 4 5

 

 −1 0 0 

Find
2 3
the transition matrix from B1 to B2 and also the transition matrix from B2 to B1 . Also find the coordinates

61 7
6 7
6 7
of 6 7
62 7 in terms of B2 .
6 7
4 5
3
B1

Solution: Let
2 3 2 3
61 1 17 6 1 1 27
6 7 6 7
6 7 6 7
A=6
61 1 7, B = 6
07 6 1 −1 07
7.
6 7 6 7
4 5 4 5
1 0 0 −1 0 0
118 Chapter 5

The transition matrix from B1 to B2 is

P = B−1 A
2 3−1 2 3
6 1 1 27 61 1 17
6 7 6 7
6 7 6 7
= 6 7 6 7
6 1 −1 07 61 1 07
6 7 6 7
4 5 4 5
−1 0 0 1 0 0
2 32 3
60 0 −17 61 1 17
6 76 7
6 76 7
= 6 76 7
60 −1 −17 61 1 07
6 76 7
4 54 5
1 1
2 2
1 1 0 0
2 3
6−1 0 0 7
6 7
6 7
= 6 7
6−2 −1 −07 .
6 7
4 5
1
2 1 2

The transition matrix from B2 to B1 is

2 3−1 2 3
6−1 0 07 6−1 0 07
6 7 6 7
6 7 6 7
P−1 =6
6−2 −1
7
07 =6
6 2 −1
7
07 .
6 7 6 7
4 5 4 5
1
2 1 2
0 2 2

Now,
2 32 3 2 3
6−1 0 0 7 617 6−17
6 76 7 6 7
6 76 7 6 7
YB2 =6
6−2 −1
76 7
0 7 627 = 6−47
6
7 .
6 76 7 6 7
4 54 5 4 5
1 11
2 1 2
3 2
B1 B2
3
As a check, observe that in the standard basis for R
2 3 2 3 2 3 2 3 2 3
6 17 61 7 61 7 61 7 66 7
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
6 7 = 1 61 7 + 2 61 7 + 3 6
6 7 6 7 7 6 7
6 27 60 7 = 63 7 ,
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
3 1 0 0 1
B1
2 3 2 3 2 3 2 3 2 3
6−17 6 1 7 66 7 6 1 7 627
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 11 6 7 6 7
6 7 = −1 6 7 6 7 6 7 6 7
6−47 6 1 7 − 4 6−17 + 2 607 = 637 .
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
11
2
−1 0 0 1
B2
Coordinates 119

531 Problem 1. Prove that the following vectors are lin- 532 Problem Consider the matrix
early independent in R4 2 3
2 3 2 3 2 3 2 3 6a 1 1 17
6 7
6 7
61 7 6 1 7 6 1 7 6 1 7 60 1 0 17
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 A(a) = 6 7.
61 7 6 1 7 6−17 6−17 6 7
6 7 6 7 6 7 6 7 61 0 a 17
a1 = 6 7 , a2 = 6 7 , a3 = 6 7 , a4 = 6 7 . 6 7
6 7 6 7 6 7 6 7 4 5
61 7 6−17 6 1 7 6−17
6 7 6 7 6 7 6 7 1 1 1 1
4 5 4 5 4 5 4 5
1 −1 −1 1
➊ Determine all a for which A(a) is not invertible.
➋ Find the inverse of A(a) when A(a) is invertible.
2 3
➌ Find the transition matrix from the basis
61 7 2 3 2 3 2 3 2 3
6 7
6 7
62 7 617 617 617 617
6 7 6 7 6 7 6 7 6 7
2. Find the coordinates of 6 7 under the ordered basis 6 7 6 7 6 7 6 7
6 7 617 617 617 607
61 7 6 7 6 7 6 7 6 7
6 7 B1 = 6 7 , 6 7 , 6 7 , 6 7
4 5 6 7 6 7 6 7 6 7
1 617 617 607 607
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
(a1 , a2 , a3 , a4 ).
1 0 0 0
2 3
61 7 to the basis
6 7 2 3 2 3 2 3 2 3
6 7
62 7
6 7 6a7 617 6 1 7 617
3. Find the coordinates of 6 7 under the ordered basis
6 7 6 7 6 7 6 7 6 7
61 7 6 7 6 7 6 7 6 7
6 7 6 0 7 617 6 0 7 617
4 5 6 7 6 7 6 7 6 7
B2 = 6 7 , 6 7 , 6 7 , 6 7 .
1 6 7 6 7 6 7 6 7
6 1 7 607 6a7 617
6 7 6 7 6 7 6 7
(a1 , a3 , a2 , a4 ). 4 5 4 5 4 5 4 5
1 1 1 1
Chapter 6
Linear Transformations

6.1 Linear Transformations

533 Definition Let hV, +, ·, Fi and hW, +, ·, Fi be vector spaces over the same field F. A linear transformation
or homomorphism

V → W
L: ,
a 7→ L(a)

is a function which is

 Linear: L(a + b) = L(a) + L(b),

 Homogeneous: L(αa) = αL(a), for α ∈ F.

☞ It is clear that the above two conditions can be summarised conveniently into
L(a + αb) = L(a) + αL(b).

534 Example Let

Mn (R) → R
L: .
A 7→ tr (A)

Then L is linear, for if (A, B) ∈ Mn (R), then

L(A + αB) = tr (A + αB) = tr (A) + αtr (B) = L(A) + αL(B).

535 Example Let

Mn (R) → Mn (R)
L: .
T
A 7→ A

Then L is linear, for if (A, B) ∈ Mn (R), then

L(A + αB) = (A + αB)T = AT + αBT = L(A) + αL(B).

120
Linear Transformations 121

536 Example For a point (x, y) ∈ R2 , its reflexion about the y-axis is (−x, y). Prove that

R2 → R2
R:
(x, y) 7→ (−x, y)

is linear.

Solution: Let (x1 , y1 ) ∈ R2 , (x2 , y2 ) ∈ R2 , and α ∈ R. Then

R((x1 , y1 ) + α(x2 , y2 )) = R((x1 + αx2 , y1 + αy2 ))

= (−(x1 + αx2 ), y1 + αy2 )

= (−x1 , y1 ) + α(−x2 , y2 )

= R((x1 , y1 )) + αR((x2 , y2 )),

whence R is linear.

537 Example Let L : R2 → R4 be a linear transformation with

2 3 2 3
6−17 62 7
2 3 6 7 2 6 7 3
6 7 6 7
6 7 6 7
617 6 1 7 6−17 607
L4 5 = 6 7
6 7 6
7; L4 5 = 6
6 7 7
6 7.
6 7 6 7
1 6 2 7 1 62 7
6 7 6 7
4 5 4 5
3 3

2 3
65 7
Find L 6
4 5.
7
3

Solution: Since 2 3 2 3 2 3
657 617 6−17
6 7 = 46 7 − 6 7,
4 5 4 5 4 5
3 1 1

we have 2 3 2 3 2 3
6−17 627 6−67
2 3 2 3 2 6 7 6 7 6
3 7
6 7 6 7 6 7
6 7 6 7 6 7
657 617 6−17 6 1 7 607 6 4 7
L 4 5 = 4L 4 5 − L 4 5 = 4 6 7 − 6 7 = 6 7
6 7 6 7 6 7 6 7 6 7 6
7.
6 7 6 7 6 7
3 1 1 6 2 7 627 6 6 7
6 7 6 7 6 7
4 5 4 5 4 5
3 3 9

538 Theorem Let hV, +, ·, Fi and hW, +, ·, Fi be vector spaces over the same field F, and let L : V → W be a
linear transformation. Then
122 Chapter 6

 L(0V ) = 0W .

 ∀x ∈ V, L(−x) = −L(x).

Proof: We have
L(0V ) = L(0V + 0V ) = L(0V ) + L(0V ),
hence
L(0V ) − L(0V ) = L(0V ).
Since
L(0V ) − L(0V ) = 0W ,
we obtain the first result.

Now
0W = L(0V ) = L(x + (−x)) = L(x) + L(−x),
from where the second result follows. ❑

539 Problem Consider L : R3 → R3 , 541 Problem Let A ∈ GLn (R) be a fixed matrix. Prove that
2 3 2 3
6 x 7 6x − y − z7
6 7 6 7 GLn (R) → GLn (R)
6 7 6 7 L:
L 6y7 = 6x + y + z7 .
6 7 6 7
4 5 4 5 H 7→ −A−1 HA−1
z z
is a linear transformation.
Prove that L is linear.

540 Problem Let h, k be fixed vectors in R3 . Prove that 542 Problem Let V be a vector space and let S ⊆ V. The
set S is said to be convex if ∀α ∈ [0; 1], ∀x, y ∈ S,
R3 × R3 → R3 (1 − α)x + αy ∈ S, that is, for any two points in S, the
L: straight line joining them also belongs to S. Let T : V → W
(x, y) 7→ x×k + h×y be a linear transformation from the vector space V to the
vector space W. Prove that T maps convex sets into convex
is a linear transformation. sets.

6.2 Kernel and Image of a Linear Transformation

543 Definition Let hV, +, ·, Fi and hW, +, ·, Fi be vector spaces over the same field F, and

V → W
T :
v 7→ T (v)

be a linear transformation. The kernel of T is the set

ker (T ) = {v ∈ V : T (v) = 0W }.

The image of T is the set

Im (T ) = {w ∈ w : ∃v ∈ V such that T (v) = w} = T (V).

☞ Since T (0 V) = 0W by Theorem 538, we have 0V ∈ ker (T ) and 0W ∈ Im (T ).

Kernel and Image of a Linear Transformation 123

544 Theorem Let hV, +, ·, Fi and hW, +, ·, Fi be vector spaces over the same field F, and

V → W
T :
v 7→ T (v)

be a linear transformation. Then ker (T ) is a vector subspace of V and Im (T ) is a vector subspace of W.

Proof: Let (v1 , v2 ) ∈ (ker (T ))2 and α ∈ F. Then T (v1 ) = T (v2 ) = 0V . We must prove that
v1 + αv2 ∈ ker (T ), that is, that T (v1 + αv2 ) = 0W . But

T (v1 + αv2 ) = T (v1 ) + αT (v2 ) = 0V + α0V = 0V

proving that ker (T ) is a subspace of V.

Now, let (w1 , w2 ) ∈ (Im (T ))2 and α ∈ F. Then ∃(v1 , v2 ) ∈ V 2 such that T (v1 ) = w1 and
T (v2 ) = w2 . We must prove that w1 + αw2 ∈ Im (T ), that is, that ∃v such that T (v) = w1 + αw2 .
But
w1 + αw2 = T (v1 ) + αT (v2 ) = T (v1 + αv2 ),

and so we may take v = v1 + αv2 . This proves that Im (T ) is a subspace of W.

❑

545 Theorem Let hV, +, ·, Fi and hW, +, ·, Fi be vector spaces over the same field F, and

V → W
T :
v 7→ T (v)

be a linear transformation. Then T is injective if and only if ker (T ) = 0V .

Proof: Assume that T is injective. Then there is a unique x ∈ V mapping to 0W :

T (x) = 0W .

By Theorem 538, T (0V ) = 0W , i.e., a linear transformation takes the zero vector of one space to
the zero vector of the target space, and so we must have x = 0V .

Conversely, assume that ker (T ) = {0V }, and that T (x) = T (y). We must prove that x = y. But

T (x) = T (y) =⇒ T (x) − T (y) = 0W

=⇒ T (x − y) = 0W

=⇒ (x − y) ∈ ker (T )

=⇒ x − y = 0V

=⇒ x = y,

as we wanted to shew. ❑
124 Chapter 6

546 Theorem (Dimension Theorem) Let hV, +, ·, Fi and hW, +, ·, Fi be vector spaces of finite dimension over the
same field F, and

V → W
T :
v 7→ T (v)
be a linear transformation. Then
dim ker (T ) + dim Im (T ) = dim V.

Proof: Let {v1 , v2 , . . . , vk } be a basis for ker (T ). By virtue of Theorem 511, we may extend
this to a basis A = {v1 , v2 , . . . , vk , vk+1 , vk+2 , . . . , vn } of V. Here n = dim V. We will now shew
that B = {T (vk+1 ), T (vk+2 ), . . . , T (vn )} is a basis for Im (T ). We prove that (i) B spans Im (T ),
and (ii) B is a linearly independent family.

Let w ∈ Im (T ). Then ∃v ∈ V such that T (v) = w. Now since A is a basis for V we can write
n
X
v= αi vi .
i=1

Hence
n
X n
X
w = T (v) = αi T (vi ) = αi T (vi ),
i=1 i=k+1

since T (vi ) = 0V for 1 ≤ i ≤ k. Thus B spans Im (T ).

To prove the linear independence of the B assume that

!
n
X n
X
0W = βi T (vi ) = T βi vi .
i=k+1 i=k+1
Pn
This means that i=k+1 βi vi ∈ ker (T ), which is impossible unless βk+1 = βk+2 = · · · = βn = 0F .
❑

547 Corollary If dim V = dim W < +∞, then T is injective if and only if it is surjective.

Proof: Simply observe that if T is injective then dim ker (T ) = 0, and if T is surjective Im (T ) =
T (V) = W and Im (T ) = dim W. ❑

548 Example Let

M2 (R) → M2 (R)
L: .
T
A 7→ A − A

Observe that L is linear. Determine ker (L) and Im (L) .

2 3
6a b7
Solution: Put A = 6
4
7 and assume L(A) = 02 . Then
5
c d
2 3 2 3 2 3 2 3
60 07 6a c7 6a b7 6 0 17
6 7 = L(A) = 6 7−6 7 = (c − b) 6 7.
4 5 4 5 4 5 4 5
0 0 b d c d −1 0
Kernel and Image of a Linear Transformation 125

This means that c = b. Thus 2 3 


 

 a
6 b7 
ker (L) = 6 7 : (a, b, d) ∈ R3 ,
4

5 

 b d 
2 3 

 

6 0 k7 
Im (L) = 6 7:k∈R .
4

5 

 −k 0 

This means that dim ker (L) = 3, and so dim Im (L) = 4 − 3 = 1.

549 Example Consider the linear transformation L : M2 (R) → R3 [X] given by

2 3
6a b7
L6
4
7 = (a + b)X2 + (a − b)X3 .
5
c d

Determine ker (L) and Im (L).

Solution: We have
2 3
6a b7
0 = L6
4
7 = (a + b)X2 + (a − b)X3 =⇒ a + b = 0, a − b = 0, =⇒ a = b = 0.
5
c d

Thus 2 3 

 0 

6 07 
ker (L) = 6 7 : (c, d) ∈ R2 .
4

5 

 c d 

Thus dim ker (L) = 2 and hence dim Im (L) = 2. Now

(a + b)X2 + (a − b)X3 =⇒ a(X2 + X3 ) + b(X2 − X3 ).

Clearly X2 + X3 , and X2 − X3 are linearly independent and span Im (L). Thus

Im (L) = span X2 + X3 , X2 − X3 .

550 Problem In problem 539 we saw that L : R3 → R3 , satisfy

2 3 2 3
6 x 7 6x − y − z7 2 3 2 3 2 3
6 7 6 7
6 7 6 7 2 3 2 3 2 3
L 6y7 = 6x + y + z7 6 1 7 6 2 7 6 1 7
6 7 6 7 1 6 7 1 6 7 0 6 7
4 5 4 5 6 7 6 7 6 7 6 7 6 7 6 7
6 7 6 0 7 6 7 6−17 6 7 6−17
z z 6 7 6 7 6 7 6 7 6 7 6 7
L 607 = 6 7 ; L 617 = 6 7 ; L 607 = 6 7 .
6 7 6 7 6 7 6 7 6 7 6 7
4 5 6−17 4 5 6 0 7 4 5 6 1 7
is linear. Determine ker (L) and Im (L). 6 7 6 7 6 7
0 4 5 0 4 5 1 4 5
0 0 0
551 Problem Let

R3 → R4 Determine ker (L) and Im (L).

L:
a 7→ L(a)
126 Chapter 6

552 Problem It is easy to see that L : R2 → R3 , 555 Problem Let

2 3
2 3 x + 2y M2 (R) R
6 7 →
6 7 L: .
6x7 6 7
L 4 5 = 6x + 2y7 A 7→ tr (A)
6 7
y 4 5
0 Determine ker (L) and Im (L).

is linear. Determine ker (L) and Im (L).

556 Problem 1. Demonstrate that

553 Problem It is easy to see that L : R2 → R3 ,

M2 (R) → M2 (R)
2 3 L:
2 3
6x − y7 A 7→ AT + A
6 7
6x7 6 7
L 4 5 = 6x + y7
6 7 is a linear transformation.
y 4 5
0 2. Find a basis for ker (L) and find dim ker (L)
3. Find a basis for Im (L) and find dim Im (L).
is linear. Determine ker (T ) and Im (T ).

557 Problem Let V be an n-dimensional vector space, where

554 Problem It is easy to see that L : R3 → R2 , the characteristic of the underlying field is different from 2.
2 3 A linear transformation T : V → V is idempotent if T 2 = T .
2 3 Prove that if T is idempotent, then
6x 7
6 7
6 7 6x − y − z7 ➊ I − T is idempotent (I is the identity function).
L 6y7 = 4 5
6 7 ➋ I + T is invertible.
4 5 y − 2z
z
➌ ker (T ) = Im (I − T )
is linear. Determine ker (L) and Im (L).
Matrix Representation 127

6.3 Matrix Representation

Let V, W be two vector spaces over the same field F. Assume that dim V = m and {vi }i∈[1;m] is an ordered basis
for V, and that dim W = n and A = {wi }i∈[1;n] an ordered basis for W. Then
2 3
6 a11 7
6 7
6 7
6a 7
6 21 7
L(v1 ) = a11 w1 + a21 w2 + · · · + an1 wn = 6 7
6 . 7
6 . 7
6 . 7
6 7
4 5
an1
2 3A
6 a12 7
6 7
6 7
6a 7
6 22 7
L(v2 ) = a12 w1 + a22 w2 + · · · + an2 wn = 6 7
6 . 7
6 . 7 .
6 . 7
6 7
4 5
an2
A
.. .. .. .. ..
. . . . .
2 3
6 a1n 7
6 7
6 7
6a 7
6 2n 7
L(vm ) = a1m w1 + a2m w2 + · · · + anm wn = 6 7
6 . 7
6 . 7
6 . 7
6 7
4 5
anm
A

558 Definition The n × m matrix

2 3
6 a11 a12 ··· a1n 7
6 7
6 7
6a a12 ··· a2n 7
6 21 7
ML = 6
6 .
7
6 . .. .. .. 7
7
6 . . . . 7
6 7
4 5
an1 an2 ··· anm

formed by the column vectors above is called the matrix representation of the linear map L with respect to the
bases {vi }i∈[1;m] , {wi }i∈[1;n] .

559 Example Consider L : R3 → R3 ,

2 3 2 3
6 x 7 6x − y − z7
6 7 6 7
6 7 6 7
L 6y7 = 6x + y + z7
6 7 6
7.
6 7 6 7
4 5 4 5
z z

Clearly L is a linear transformation.

1. Find the matrix corresponding to L under the standard ordered basis.
128 Chapter 6

2 3 2 3 2 3
617 617 617
6 7 6 7 6 7
6 7 6 7 6 7
2. Find the matrix corresponding to L under the ordered basis 6 7 6 7 6 7
607 , 617 , 607 , for both the domain and the
6 7 6 7 6 7
4 5 4 5 4 5
0 0 1
image of L.

Solution:
2 3 2 3 2 3 2 3 2 3 2 3
61 7 617 60 7 6−17 607 6−17
6 7 6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7 6 7
1. Solution: The matrix will be a 3 × 3 matrix. We have L 607 = 617, L 617 = 6 1 7, and L 607 = 6 1 7
6 7 6 7 6 7 6 7 6 7 6
7,
6 7 6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5 4 5
0 0 0 0 1 1
whence the desired matrix is 2 3
61 −1 −17
6 7
6 7
6 7
61 1 1 7.
6 7
4 5
0 0 1

2. Call this basis A . We have

2 3 2 3 2 3 2 3 2 3 2 3
617 617 617 617 617 607
6 7 6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7 6 7
L 607 = 617 = 0 607 + 1 617 + 0 6
6 7 6 7 6 7 6 7 7 6 7
607 = 617 ,
6 7 6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5 4 5
0 0 0 0 1 0
A

2 3 2 3 2 3 2 3 2 3 2 3
617 607 617 617 617 6−27
6 7 6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7 6 7
L 617 = 627 = −2 607 + 2 617 + 0 607 = 6 2 7
6 7 6 7 6 7 6 7 6 7 6
7 ,
6 7 6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5 4 5
0 0 0 0 1 0
A

and 2 3 2 3 2 3 2 3 2 3 2 3
617 607 617 617 617 6−37
6 7 6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7 6 7
L 607 = 627 = −3 607 + 2 617 + 1 607 = 6 2 7
6 7 6 7 6 7 6 7 6 7 6
7 ,
6 7 6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5 4 5
1 1 0 0 1 1
A

whence the desired matrix is 2 3

60 −2 −37
6 7
6 7
6 7
61 2 2 7.
6 7
4 5
0 0 1

560 Example Let Rn [x] denote the set of polynomials with real coefficients with degree at most n.
Matrix Representation 129

1. Prove that

R3 [x] → R1 [x]
L:
p(x) 7→ p ′′ (x)

is a linear transformation. Here p ′′ (x) denotes the second derivative of p(x) with respect to x.

2. Find the matrix of L using the ordered bases {1, x, x2 , x3 } for R3 [x] and {1, x} for R2 [x].

3. Find the matrix of L using the ordered bases {1, x, x2 , x3 } for R3 [x] and {1, x + 2} for R1 [x].

4. Find a basis for ker (L) and find dim ker (L).

5. Find a basis for Im (L) and find dim Im (L).

Solution:

1. Let (p(x), q(x)) ∈ (R3 [x])2 and α ∈ R. Then

L(p(x) + αq(x)) = (p(x) + αq(x)) ′′ = p ′′ (x) + αq ′′ (x) = L(p(x)) + αL(q(x)),

whence L is linear.

2. We have
2 3

d 2 607
L(1) = 1 = 0 = 0(1) + 0(x) = 6 7
4 5,
dx2
0
2 3

d 2 607
L(x) = x = 0 = 0(1) + 0(x) = 6 7
4 5,
dx2
0
2 3

d2 627
L(x2 ) = x2 = 2 = 2(1) + 0(x) = 6 7
4 5,
dx2
0
2 3

d2 607
L(x3 ) = x3 = 6x = 0(1) + 6(x) = 6 7
4 5,
dx 2
6

whence the matrix representation of L under the standard basis is

2 3
60 0 2 07
6 7.
4 5
0 0 0 6
130 Chapter 6

3. We have 2 3

d2 607
L(1) = 1 = 0 = 0(1) + 0(x + 2) = 6 7,
dx2 4 5
0
2 3

d2 607
L(x) = x = 0 = 0(1) + 0(x + 2) = 6 7,
dx2 4 5
0
2 3
2
d 2 627
L(x2 ) = x = 2 = 2(1) + 0(x + 2) = 6 7,
dx2 4 5
0
2 3

d2 6−127
L(x3 ) = x3 = 6x = −12(1) + 6(x + 2) = 6
4
7,
5
dx 2
6
whence the matrix representation of L under the standard basis is
2 3
60 0 2 −127
6 7.
4 5
0 0 0 6

4. Assume that p(x) = a + bx + cx2 + dx3 ∈ ker (L). Then

0 = L(p(x)) = 2c + 6dx, ∀x ∈ R.
This means that c = d = 0. Thus a, b are free and
ker (L) = {a + bx : (a, b) ∈ R2 }.
Hence dim ker (L) = 2.
5. By the Dimension Theorem, dim Im (L) = 4 − 2 = 2. Put q(x) = α + βx + γx2 + δx3 . Then
L(q(x)) = 2γ + 6δ(x) = (2γ)(1) + (6δ)(x).
Clearly {1, x} are linearly independent and span Im (L). Hence
Im (L) = span (1, x) = R1 [x].

561 Example 1. A linear transformation T : R3 → R3 is such that

2 3 2 3
627 6 3 7
6 7 6 7
6 7 6 7
T (i) = 6 7
617 ; T (j) = 6 0 7
6
7.
6 7 6 7
4 5 4 5
1 −1

It is known that
Im (T ) = span (T (i), T (j))
and that 02 31
B6 1 7C
B6 7C
B6 7C
ker (T ) = span B6 2 7
B 6 C
7C .
B6 7C
4 5A
−1
Matrix Representation 131

Argue that there must be λ and µ such that

T (k) = λT (i) + µT (j).

2. Find λ and µ, and hence, the matrix representing T under the standard ordered basis.

Solution:
1. Since T (k) ∈ Im (T ) and Im (T ) is generated by T (i) and T (k) there must be (λ, µ) ∈ R2 with
2 3 2 3 2 3
627 6 3 7 62λ + 3µ7
6 7 6 7 6 7
6 7 6 7 6 7
T (k) = λT (i) + µT (j) = λ 6 7 +µ6
617
7=6
6 0 7 6 λ
7.
7
6 7 6 7 6 7
4 5 4 5 4 5
1 −1 λ−µ

2. The matrix of T is 2 3
62 3 2λ + 3µ7
 6 7
6 7
6 7.
T (i) T (j) T (k) = 61 0 λ 7
6 7
4 5
1 −1 λ−µ
2 3
6 1 7
6 7
6 7
Since 6 7
6 2 7 ∈ ker (T ) we must have
6 7
4 5
−1

2 32 3 2 3
62 3 2λ + 3µ7 6 1 7 607
6 76 7 6 7
6 76 7 6 7
6 76 7 6 7
61 0 λ 7 6 2 7 = 607 .
6 76 7 6 7
4 54 5 4 5
1 −1 λ−µ −1 0

Solving the resulting system of linear equations we obtain λ = 1, µ = 2. The required matrix is thus
2 3
62 3 8 7
6 7
6 7
6 7
61 0 1 7.
6 7
4 5
1 −1 −1

☞ If the linear mapping L : V → W, dim V = n, dim W = m has matrix representation

A ∈ Mm×n (F), then dim Im (L) = rank (A).

562 Problem 1. A linear transformation T : R3 → R3 has kernel the line x = y = z. If

as image the plane with equation x + y + z = 0 and as 2 3 2 3 2 3 2 3 2 3 2 3
617 6a7 627 6 3 7 617 6−17
6 7 6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7 6 7
T 617 = 6 0 7 , T 617 = 6 b 7 , T 627 = 6 2 7 .
6 7 6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5 4 5
2 1 1 −5 1 c
132 Chapter 6

2 3 2 3 2 3
Find a, b, c.
617 617 617
2. Find the matrix representation of T under the standard 6 7 6 7 6 7
6 7 6 7 6 7
basis. 2. The ordered basis for R3 is 607 , 617 , 617 and R2 has
6 7 6 7 6 7
4 5 4 5 4 5
563 Problem 1. Prove that T : R2 → R3 0 0 1
2 3 the standard ordered basis .
2 3 2 3 2 3
2 3
6 x+y 7 617 617 617
x 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
T 4 5=6 x−y 7 3. The ordered basis for R3 is 607 , 617 , 617 and the
6 7 6 7 6 7 6 7
y 4 5 4 5 4 5 4 5
2x + 3y 0 0 1
 2 3 2 3
is a linear transformation. 
 

 1 
2 6 7 617
2. Find a basis for ker (T ) and find dim ker (T ) ordered basis for R is A = 4 5 , 4 5 .

 

 0 1 
3. Find a basis for Im (T ) and find dim Im (T ).
4. Find the matrix of T under the
2ordered
3 2 bases =
3 2A3 2 2
2 3 2 3 
 
 2 3 2 3 T : R → R satisfies
565 Problem A linear transformation
  617 6 1 7 607

 


 1 1 
 6 7 6 7 6 7
  61 7 62 7
6 7 6 7 2 6 7 6 7 6 7 ker (T ) = Im (T ), and T 4 5 = 4 5. Find the matrix repre-
4 5 , 4 5 of R and B = 617 , 6 0 7 , 617

 
 
 6 7 6 7 6 7 
 2 3  4 5 4 5 4 5

 

1 3

 1 
−1 0  senting T under the standard ordered basis.
3
of R .
566 Problem Find the matrix representation for the linear
map
564 Problem Let
M2 (R) → R
3 2 L: ,
R → R
L: , A 7→ tr (A)
a 7→ L(a) under the standard basis
2 3 2 3 2 3 2 3
where 
 
2 3  1 0
6 7 60 1 7 60 07 60 07
2 3 A = 4 5,4 5,4 5,4 5
6x7 

 0 0


6 7 x + 2y 0 0 1 0 0 1 
6 7 6 7
L 6y7 = 4 5.
6 7 for M2 (R).
4 5 3x − z
z
567 Problem Let A ∈ Mn×p (R), B ∈ Mp×q (R), and
Clearly L is linear. Find a matrix representation for L if C ∈ Mq×r (R), be such that rank (B) = rank (AB). Shew
1. The bases for both R3 and R2 are both the standard that
ordered bases. rank (BC) = rank (ABC) .
Chapter 7
Determinants

7.1 Permutations
568 Definition Let S be a finite set with n ≥ 1 elements. A permutation is a bijective function τ : S → S. It is
easy to see that there are n! permutations from S onto itself.

Since we are mostly concerned with the action that τ exerts on S rather than with the particular names of the
elements of S, we will take S to be the set S = {1, 2, 3, . . . , n}. We indicate a permutation τ by means of the
following convenient diagram 2 3
6 1 2 ··· n 7
τ=6
4
7.
5
τ(1) τ(2) · · · τ(n)

569 Definition The notation Sn will denote the set of all permutations on {1, 2, 3, . . . , n}. Under this notation, the
composition of two permutations (τ, σ) ∈ S2n is
2 3 2 3
6 1 2 ··· n 7 6 1 2 ··· n 7
τ◦σ = 6
4
7◦6
5 4
7
5
τ(1) τ(2) · · · τ(n) σ(1) σ(2) · · · σ(n)
2 3 .
6 1 2 ··· n 7
= 6
4
7
5
(τ ◦ σ)(1) (τ ◦ σ)(2) · · · (τ ◦ σ)(n)

The k-fold composition of τ is

· · ◦ τ} = τk .
| ◦ ·{z
τ
k compositions

☞ We usually do away with the ◦ and write τ ◦ σ simply as τσ. This “product of permutations”
is thus simply function composition.
Given a permutation τ : S → S, since τ is bijective,
τ−1 : S → S
exists and is also a permutation. In fact if
2 3
6 1 2 ··· n 7
τ=6
4
7,
5
τ(1) τ(2) · · · τ(n)

133
134 Chapter 7

then 2 3
6τ(1) τ(2) · · · τ(n)7
τ−1 = 6
4
7.
5
1 2 ··· n

2 3

Figure 7.1: S3 are rotations and reflexions.

570 Example The set S3 has 3! = 6 elements, which are given below.
1. Id : {1, 2, 3} → {1, 2, 3} with
2 3
61 2 37
Id = 6
4
7.
5
1 2 3

2. τ1 : {1, 2, 3} → {1, 2, 3} with

2 3
61 2 37
τ1 = 6
4
7.
5
1 3 2

3. τ2 : {1, 2, 3} → {1, 2, 3} with

2 3
61 2 37
τ2 = 6
4
7.
5
3 2 1

4. τ3 : {1, 2, 3} → {1, 2, 3} with

2 3
61 2 37
τ3 = 6
4
7.
5
2 1 3

5. σ1 : {1, 2, 3} → {1, 2, 3} with

2 3
61 2 37
σ1 = 6
4
7.
5
2 3 1

6. σ2 : {1, 2, 3} → {1, 2, 3} with

2 3
61 2 37
σ2 = 6
4
7.
5
3 1 2
Permutations 135

571 Example The compositions τ1 ◦ σ1 and σ1 ◦ τ1 can be found as follows.

2 3 2 3 2 3
61 2 37 61 2 37 61 2 37
τ1 ◦ σ1 = 6
4
7◦6
5 4
7=6
5 4
7 = τ2 .
5
1 3 2 2 3 1 3 2 1

(We read from right to left 1 → 2 → 3 (“1 goes to 2, 2 goes to 3, so 1 goes to 3”), etc. Similarly
2 3 2 3 2 3
61 2 37 61 2 37 61 2 37
σ1 ◦ τ1 = 6
4
7◦6
5 4
7=6
5 4
7 = τ3 .
5
2 3 1 1 3 2 2 1 3

Observe in particular that σ1 ◦ τ1 6= τ1 ◦ σ1 . Finding all the other products we deduce the following “multiplication
table” (where the “multiplication” operation is really composition of functions).

◦ Id τ1 τ2 τ3 σ1 σ2

Id Id τ1 τ2 τ3 σ1 σ2

τ1 τ1 Id σ1 σ2 τ2 τ3

τ2 τ2 σ2 Id σ1 τ3 τ1

τ3 τ3 σ1 σ2 Id τ1 τ2

σ2 σ2 τ2 τ3 τ1 Id σ1

σ1 σ1 τ3 τ1 τ2 σ2 Id

The permutations in example 570 can be conveniently interpreted as follows. Consider an equilateral triangle
with vertices labelled 1, 2 and 3, as in figure 7.1. Each τa is a reflexion (“flipping”) about the line joining the vertex
a with the midpoint of the side opposite a. For example τ1 fixes 1 and flips 2 and 3. Observe that two successive
flips return the vertices to their original position and so (∀a ∈ {1, 2, 3})(τ2a = Id ). Similarly, σ1 is a rotation of
the vertices by an angle of 120◦ . Three successive rotations restore the vertices to their original position and so
σ31 = Id .

572 Example To find τ−1

1 take the representation of τ1 and exchange the rows:
2 3
61 3 27
τ−1
1 =6
4
7.
5
1 2 3

This is more naturally written as

2 3
61 2 37
τ−1
1 =6
4
7.
5
1 3 2

Observe that τ−1

1 = τ1 .
136 Chapter 7

573 Example To find σ−1

1 take the representation of σ1 and exchange the rows:
2 3
62 3 17
σ−1
1 =6
4
7.
5
1 2 3

This is more naturally written as

2 3
61 2 37
σ−1
1 =6
4
7.
5
3 1 2

Observe that σ−1

1 = σ2 .

7.2 Cycle Notation

We now present a shorthand notation for permutations by introducing the idea of a cycle. Consider in S9 the
permutation
2 3
61 2 3 4 5 6 7 8 97
τ=6
4
7.
5
2 1 3 6 9 7 8 4 5

We start with 1. Since 1 goes to 2 and 2 goes back to 1, we write (12). Now we continue with 3. Since 3 goes to 3,
we write (3). We continue with 4. As 4 goes 6, 6 goes to 7, 7 goes 8, and 8 goes back to 4, we write (4678). We
consider now 5 which goes to 9 and 9 goes back to 5, so we write (59). We have written τ as a product of disjoint
cycles
τ = (12)(3)(4678)(59).
This prompts the following definition.

574 Definition Let l ≥ 1 and let i1 , . . . , il ∈ {1, 2, . . . n} be distinct. We write (i1 i2 . . . il ) for the element
σ ∈ Sn such that σ(ir ) = ir+1 , 1 ≤ r < l, σ(il ) = i1 and σ(i) = i for i 6∈ {i1 , . . . , il }. We say that (i1 i2 . . . il )
is a cycle of length l. The order of a cycle is its length. Observe that if τ has order l then τl = Id .

☞ Observe that (i 2 . . . il i1 ) = (i1 . . . il ) etc., and that (1) = (2) = · · · = (n) = Id . In fact,
we have
(i1 . . . il ) = (j1 . . . jm )
if and only if (1) l = m and if (2) l > 1: ∃a such that ∀k: ik = jk+a mod l . Two cycles (i1 , . . . , il )
and (j1 , . . . , jm ) are disjoint if {i1 , . . . , il } ∩ {j1 , . . . , jm } = ∅. Disjoint cycles commute and if
τ = σ1 σ2 · · · σt is the product of disjoint cycles of length l1 , l2 , . . . , lt respectively, then τ has
order
lcm (l1 , l2 , . . . , lt ) .

575 Example A cycle decomposition for α ∈ S9 ,

2 3
61 2 3 4 5 6 7 8 97
α=6
4
7
5
1 8 7 6 2 3 4 5 9

is
(285)(3746).
The order of α is lcm (3, 4) = 12.
Cycle Notation 137

576 Example The cycle decomposition β = (123)(567) in S9 arises from the permutation
2 3
61 2 3 4 5 6 7 8 97
β=6
4
7.
5
2 3 1 4 6 7 5 8 9

Its order is lcm (3, 3) = 3.

577 Example Find a shuffle of a deck of 13 cards that requires 42 repeats to return the cards to their original order.

Solution: Here is one (of many possible ones). Observe that 7 + 6 = 13 and 7 × 6 = 42. We take the permutation

(1 2 3 4 5 6 7)(8 9 10 11 12 13)

which has order 42. This corresponds to the following shuffle: For

i ∈ {1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 12},

take the ith card to the (i + 1)th place, take the 7th card to the first position and the 13th card to the 8th position.
Query: Of all possible shuffles of 13 cards, which one takes the longest to restitute the cards to their original position?

578 Example Let a shuffle of a deck of 10 cards be made as follows: The top card is put at the bottom, the deck is
cut in half, the bottom half is placed on top of the top half, and then the resulting bottom card is put on top. How
many times must this shuffle be repeated to get the cards in the initial order? Explain.

Solution: Putting the top card at the bottom corresponds to

2 3
61 2 3 4 5 6 7 8 9 107
6 7.
4 5
2 3 4 5 6 7 8 9 10 1

Cutting this new arrangement in half and putting the lower half on top corresponds to
2 3
61 2 3 4 5 6 7 8 9 107
6 7.
4 5
7 8 9 10 1 2 3 4 5 6

Putting the bottom card of this new arrangement on top corresponds to

2 3
61 2 3 4 5 6 7 8 9 107
6 7 = (1 6)(2 7)(3 8)(4 9)(5 10).
4 5
6 7 8 9 10 1 2 3 4 5

The order of this permutation is lcm(2, 2, 2, 2, 2) = 2, so in 2 shuffles the cards are restored to their original position.
The above examples illustrate the general case, given in the following theorem.

579 Theorem Every permutation in Sn can be written as a product of disjoint cycles.

Proof: Let τ ∈ Sn , a1 ∈ {1, 2, . . . , n}. Put τk (a1 ) = ak+1 , k ≥ 0. Let a1 , a2 , . . . , as

be the longest chain with no repeats. Then we have τ(as ) = a1 . If the {a1 , a2 , . . . , as } ex-
haust {1, 2, . . . , n} then we have τ = (a1 a2 . . . as ). If not, there exist b1 ∈ {1, 2, . . . , n} \
{a1 , a2 , . . . , as }. Again, we find the longest chain of distinct b1 , b2 , . . . , bt such that τ(bk ) =
bk+1, k = 1, . . . , t − 1 and τ(bt ) = b1 . If the {a1 , a2 , . . . , as , b1 , b2 , . . . , bt } exhaust all the
138 Chapter 7

{1, 2, . . . , n} we have τ = (a1 a2 . . . as )(b1 b2 . . . bt ). If not we continue the process and

find
τ = (a1 a2 . . . as )(b1 b2 . . . bt )(c1 . . .) . . . .
This process stops because we have only n elements. ❑

580 Definition A transposition is a cycle of length 2.1

581 Example The cycle (23468) can be written as a product of transpositions as follows

(23468) = (28)(26)(24)(23).

Notice that this decomposition as the product of transpositions is not unique. Another decomposition is

(23468) = (23)(34)(46)(68).

582 Lemma Every permutation is the product of transpositions.

Proof: It is enough to observe that

(a1 a2 . . . as ) = (a1 as )(a1 as−1 ) · · · (a1 a2 )

and appeal to Theorem 579. ❑

Let σ ∈ Sn and let (i, j) ∈ {1, 2, . . . , n}2 , i 6= j. Since σ is a permutation, ∃(a, b) ∈ {1, 2, . . . , n}2 , a 6= b, such
that σ(j) − σ(i) = b − a. This means that
 
 
 Y σ(i) − σ(j) 
  = 1.
 i−j 
1≤i<j≤n 

583 Definition Let σ ∈ Sn . We define the sign sgn(σ) of σ as

Y σ(i) − σ(j)
sgn(σ) = = (−1)σ .
1≤i<j≤n
i−j

If sgn(σ) = 1, then we say that σ is an even permutation, and if sgn(σ) = −1 we say that σ is an odd permutation.

☞ Notice that in fact

sgn(σ) = (−1)I(σ) ,
where I(σ) = #{(i, j) | 1 ≤ i < j ≤ n and σ(i) > σ(j)}, i.e., I(σ) is the number of inversions that σ
effects to the identity permutation Id .

584 Example The transposition (1 2) has one inversion.

585 Lemma For any transposition (k l) we have sgn((k l)) = −1.

Proof: Let τ be transposition that exchanges k and l, and assume that k < l:
2 3
61 2 ... k−1 k k + 1 ... l−1 l l+1 ... n7
τ=6
4
7
5
1 2 ... k−1 l k + 1 ... l−1 k l+1 ... n

Let us count the number of inversions of τ:

1A cycle of length 2 should more appropriately be called a bicycle.
Cycle Notation 139

 The pairs (i, j) with i ∈ {1, 2, . . . , k − 1} ∪ {l, l + 1, . . . , n} and i < j do not suffer an inversion;
 The pair (k, j) with k < j suffers an inversion if and only if j ∈ {k + 1, k + 2, . . . , l}, making
l − k inversions;
 If i ∈ {k + 1, k + 2, . . . , l − 1} and i < j, (i, j) suffers an inversion if and only if j = l, giving
l − 1 − k inversions.

This gives a total of I(τ) = (l − k) + (l − 1 − k) = 2(l − k − 1) + 1 inversions when k < l. Since this
number is odd, we have sgn(τ) = (−1)I(τ) = −1. In general we see that the transposition (k l)
has 2|k − l| − 1 inversions. ❑

586 Theorem Let (σ, τ) ∈ S2n . Then

sgn(τσ) = sgn(τ)sgn(σ).

Proof: We have

Q (στ)(i)−(στ)(j)
sgn(στ) = 1≤i<j≤n i−j
Q  Q 
σ(τ(i))−σ(τ(j)) τ(i)−τ(j)
= 1≤i<j≤n τ(i)−τ(j)
· 1≤i<j≤n i−j
.

The second factor on this last equality is clearly sgn(τ), we must shew that the first factor is
sgn(σ). Observe now that for 1 ≤ a < b ≤ n we have

σ(a) − σ(b) σ(b) − σ(a)

= .
a−b b−a
Since σ and τ are permutations, ∃b 6= a, τ(i) = a, τ(j) = b and so στ(i) = σ(a), στ(j) = b. Thus

σ(τ(i)) − σ(τ(j)) σ(a) − σ(b)

=
τ(i) − τ(j) a−b

and so
Y σ(τ(i)) − σ(τ(j)) Y σ(a) − σ(b)
= = sgn(σ).
1≤i<j≤n
τ(i) − τ(j) 1≤a<b≤n
a−b

587 Corollary The identity permutation is even. If τ ∈ Sn , then sgn(τ) = sgn(τ−1 ).

Proof: Since there are no inversions in Id , we have sgn(Id ) = (−1)0 = 1. Since ττ−1 = Id ,
−1
we must have 1 = sgn(Id ) = sgn(ττ−1 ) = sgn(τ)sgn(τ−1 ) = (−1)τ (−1)τ by Theorem 586.
Since the values on the righthand of this last equality are ±1, we must have sgn(τ) = sgn(τ−1 ).
❑

588 Lemma We have sgn(1 2 . . . l)) = (−1)l−1 .

Proof: Simply observe that the number of inversions of (1 2 . . . l) is l − 1. ❑

589 Lemma Let (τ, (i1 . . . il ) ∈ S2n . Then

τ(i1 . . . il )τ−1 = (τ(i1 ) . . . τ(il )),

and if σ ∈ Sn is a cycle of length l then

sgn(σ) = (−1)l−1
.
140 Chapter 7

Proof: For 1 ≤ k < l we have (τ(i1 . . . il )τ−1 )(τ(ik )) = τ((i1 . . . il )(ik )) = τ(ik+1 ).
On a (τ(i1 . . . il )τ−1 )(τ(il )) = τ((i1 . . . il )(il )) = τ(i1 ). For i 6∈ {τ(i1 ) . . . τ(il )} we have
τ−1 (i) 6∈ {i1 . . . il } whence (i1 . . . il )(τ−1 (i)) = τ−1 (i) etc.

Furthermore, write σ = (i1 . . . il ). Let τ ∈ Sn be such that τ(k) = ik for 1 ≤ k ≤ l. Then

σ = τ(1 2 . . . l)τ−1 and so we must have sgn(σ) = sgn(τ)sgn((1 2 . . . l))sgn(τ−1 ), which equals
sgn((1 2 . . . l)) by virtue of Theorem 586 and Corollary 587. The result now follows by appealing
to Lemma 588 ❑

590 Corollary Let σ = σ1 σ2 · · · σr be a product of disjoint cycles, each of length l1 , . . . , lr , respectively. Then
Pr
sgn(σ) = (−1) i=1 (li −1) .
Hence, the product of two even permutations is even, the product of two odd permutations is even, and the product
of an even permutation and an odd permutation is odd.

Proof: This follows at once from Theorem 586 and Lemma 589. ❑

591 Example The cycle (4678) is an odd cycle; the cycle (1) is an even cycle; the cycle (12345) is an even cycle.

592 Corollary Every permutation can be decomposed as a product of transpositions. This decomposition is not
necessarily unique, but its parity is unique.

Proof: This follows from Theorem 579, Lemma 582, and Corollary 590. ❑

593 Example (The 15 puzzle) Consider a grid with 16 squares, as shewn in (7.1), where 15 squares are numbered
1 through 15 and the 16th slot is empty.

1 2 3 4

5 6 7 8
(7.1)
9 10 11 12

13 14 15
In this grid we may successively exchange the empty slot with any of its neighbours, as for example

1 2 3 4

5 6 7 8
. (7.2)
9 10 11 12

13 14 15

We ask whether through a series of valid moves we may arrive at the following position.

1 2 3 4

5 6 7 8
(7.3)
9 10 11 12

13 15 14
Determinants 141

Solution: Let us shew that this is impossible. Each time we move a square to the empty position, we make
transpositions on the set {1, 2, . . . , 16}. Thus at each move, the permutation is multiplied by a transposition and
hence it changes sign. Observe that the permutation corresponding to the square in (7.3) is (14 15) (the positions
14th and 15th are transposed) and hence it is an odd permutation. But we claim that the empty slot can only return
to its original position after an even permutation. To see this paint the grid as a checkerboard:

B R B R

R B R B
(7.4)
B R B R

R B R B

We see that after each move, the empty square changes from black to red, and thus after an odd number of moves
the empty slot is on a red square. Thus the empty slot cannot return to its original position in an odd number of
moves. This completes the proof.

594 Problem Decompose the permutation as a product of disjoint cycles and find its order.
2 3
61 2 3 4 5 6 7 8 97
4 5
2 3 4 1 5 8 6 7 9

7.3 Determinants
There are many ways of developing the theory of determinants. We will choose a way that will allow us to deduce
the properties of determinants with ease, but has the drawback of being computationally cumbersome. In the next
section we will shew that our way of defining determinants is equivalent to a more computationally friendly one.

It may be pertinent here to quickly review some properties of permutations. Recall that if σ ∈ Sn is a cycle of
length l, then its signum sgn(σ) = ±1 depending on the parity of l − 1. For example, in S7 ,

σ = (1 3 4 7 6)

has length 5, and the parity of 5 − 1 = 4 is even, and so we write sgn(σ) = +1. On the other hand,

τ = (1 3 4 7 6 5)

has length 6, and the parity of 6 − 1 = 5 is odd, and so we write sgn(τ) = −1.

Recall also that if (σ, τ) ∈ S2n , then

sgn(τσ) = sgn(τ)sgn(σ).

Thus from the above two examples

στ = (1 3 4 7 6)(1 3 4 7 6 5)
has signum sgn(σ)sgn(τ) = (+1)(−1) = −1. Observe in particular that for the identity permutation Id ∈ Sn
we have sgn(Id ) = +1.

595 Definition Let A ∈ Mn (F), A = [aij ] be a square matrix. The determinant of A is defined and denoted by
the sum X
det A = sgn(σ)a1σ(1) a2σ(2) · · · anσ(n) .
σ∈Sn
142 Chapter 7

☞ The determinantal sum has n! summands.

596 Example If n = 1, then S1 has only one member, Id , where Id (1) = 1. Since Id is an even permutation,
sgn(Id ) = (+1) Thus if A = (a11 ), then
det A = a11

597 Example If n = 2, then S2 has 2! = 2 members, Id and σ = (1 2). Observe that sgn(σ) = −1. Thus if
2 3
6a11 a12 7
A=6
4
7
5
a21 a22

then
det A = sgn(Id )a1Id (1) a2Id (2) + sgn(σ)a1σ(1) a2σ(2) = a11 a22 − a12 a21 .

598 Example From the above formula for 2 × 2 matrices it follows that
2 3
61 27
det A = det 6
4
7
5
3 4

= (1)(4) − (3)(2) = −2,

2 3
6−1 27
det B = det 6
4
7 (−1)(4) − (3)(2)
5
3 4

= −10,

and
2 3
60 47
det(A + B) = det 6
4
7 = (0)(8) − (6)(4) = −24.
5
6 8

Observe in particular that det(A + B) 6= det A + det B.

599 Example If n = 3, then S2 has 3! = 6 members:

Id , τ1 = (2 3), τ2 = (1 3), τ3 = (1 2), σ1 = (1 2 3), σ2 = (1 3 2).

. Observe that Id , σ1 , σ2 are even, and τ1 , τ2 , τ3 are odd. Thus if

2 3
6a11 a12 a13 7
6 7
6 7
A=6
6a21 a22 a23 7
7
6 7
4 5
a31 a32 a33
Determinants 143

then
det A = sgn(Id )a1Id (1) a2Id (2) a3Id (3) + sgn(τ1 )a1τ1 (1) a2τ1 (2) a3τ1 (3)

+sgn(τ2 )a1τ2 (1) a2τ2 (2) a3τ2 (3) + sgn(τ3 )a1τ3 (1) a2τ3 (2) a3τ3 (3)

+sgn(σ1 )a1σ1 (1) a2σ1 (2) a3σ1 (3) + sgn(σ2 )a1σ2 (1) a2σ2 (2) a3σ2 (3)

= a11 a22 a33 − a11 a23 a32 − a13 a22 a31

−a13 a21 a33 + a12 a23 a31 + a13 a21 a32 .

600 Theorem (Row-Alternancy of Determinants) Let A ∈ Mn (F), A = [aij ]. If B ∈ Mn (F), B = [bij ] is the
matrix obtained by interchanging the s-th row of A with its t-th row, then det B = − det A.

Proof: Let τ be the transposition

2 3
6 s t 7
τ=6
4
7.
5
τ(t) τ(s)

Then στ(a) = σ(a) for a ∈ {1, 2, . . . , n} \ {s, t}. Also, sgn(στ) = sgn(σ)sgn(τ) = −sgn(σ). As σ
ranges through all permutations of Sn , so does στ, hence

P
det B = σ∈Sn sgn(σ)b1σ(1) b2σ(2) · · · bsσ(s) · · · btσ(t) · · · bnσ(n)
P
= σ∈Sn sgn(σ)a1σ(1) a2σ(2) · · · ast · · · ats · · · anσ(n)
P
= − σ∈Sn sgn(στ)a1στ(1) a2στ(2) · · · asστ(s) · · · atστ(t) · · · anστ(n)
P
= − λ∈Sn sgn(λ)a1λ(1) a2λ(2) · · · anλ(n)

= − det A.

601 Corollary If A(r:k) , 1 ≤ k ≤ n denote the rows of A and σ ∈ Sn , then

2 3
6 A(r:σ(1)) 7
6 7
6 7
6A 7
6 (r:σ(2)) 7
det 6
6
7 = (sgn(σ)) det A.
7
6 .. 7
6 . 7
6 7
4 5
A(r:σ(n))

An analogous result holds for columns.

Proof: Apply the result of Theorem 600 multiple times. ❑

602 Theorem Let A ∈ Mn (F), A = [aij ]. Then

det AT = det A.
144 Chapter 7

Proof: Let C = AT . By definition

det AT = det C
P
= σ∈Sn sgn(σ)c1σ(1) c2σ(2) · · · cnσ(n)
P
= σ∈Sn sgn(σ)aσ(1)1 aσ(2)2 · · · aσ(n)n .

But the product aσ(1)1 aσ(2)2 · · · aσ(n)n also appears in det A with the same signum sgn(σ), since
the permutation 2 3
6σ(1) σ(2) · · · σ(n)7
6 7
4 5
1 2 ··· n

is the inverse of the permutation

2 3
6 1 2 ··· n 7
6 7.
4 5
σ(1) σ(2) · · · σ(n)

603 Corollary (Column-Alternancy of Determinants) Let A ∈ Mn (F), A = [aij ]. If C ∈ Mn (F), C = [cij ] is the
matrix obtained by interchanging the s-th column of A with its t-th column, then det C = − det A.

Proof: This follows upon combining Theorem 600 and Theorem 602. ❑

604 Theorem (Row Homogeneity of Determinants) Let A ∈ Mn (F), A = [aij ] and α ∈ F. If B ∈ Mn (F), B =
[bij ] is the matrix obtained by multiplying the s-th row of A by α, then

det B = α det A.

Proof: Simply observe that

sgn(σ)a1σ(1) a2σ(2) · · · αasσ(s) · · · anσ(n) = αsgn(σ)a1σ(1) a2σ(2) · · · asσ(s) · · · anσ(n) .

605 Corollary (Column Homogeneity of Determinants) If C ∈ Mn (F), C = (Cij ) is the matrix obtained by mul-
tiplying the s-th column of A by α, then
det C = α det A.

Proof: This follows upon using Theorem 602 and Theorem 604. ❑

☞ It follows from Theorem 604 and Corollary 605 that if a row (or column) of a matrix consists
of 0F s only, then the determinant of this matrix is 0F .

606 Example
2 3 2 3
6x 1 a7 61 1 a7
6 7 6 7
6 7 6 7
det 6
6x 2 1
7 = x det 6
b7 6x 1
7
b7 .
6 7 6 7
4 5 4 5
x3 1 c x2 1 c
Determinants 145

607 Corollary
det(αA) = αn det A.

Proof: Since there are n columns, we are able to pull out one factor of α from each one. ❑

608 Example Recall that a matrix A is skew-symmetric if A = −AT . Let A ∈ M2001(R) be skew-symmetric.
Prove that det A = 0.

Solution: We have
det A = det(−AT ) = (−1)2001 det AT = − det A,
and so 2 det A = 0, from where det A = 0.

609 Lemma (Row-Linearity and Column-Linearity of Determinants) Let A ∈ Mn (F), A = [aij ]. For a fixed row
s, suppose that asj = bsj + csj for each j ∈ [1; n]. Then
2 3
a11 a12 ··· a1n

6 a21 7
6 7
a22 ··· a2n

6 7
6 7
. . . .
. . . .
6 . . ··· . .
7
6 a(s−1)1 7
det 6 7
a(s−1)2 ··· a(s−1)n
6b + c 7
6 s1 s1 bs2 + cs2 ··· bsn + csn
7
6a 7
6 (s+1)1 a(s+1)2 ··· a(s+1)n
7
6 .7
4 . .
.5
.
. . .
. . ··· . .

an1 an2 ··· ann

2 3
a11 a12 ··· a1n

6 a21 7
6 7
a22 ··· a2n

6 7
6 7
. . . .
. . . .
6 . . ··· . .
7
6a(s−1)1 7
= det 6 7
a(s−1)2 ··· a(s−1)n
6 b 7
6 s1 bs2 ··· bsn
7
6a 7
6 (s+1)1 a(s+1)2 ··· a(s+1)n
7
6 .7
4 .
.5
. .
. . .
. . ··· . .

an1 an2 ··· ann

2 3
a11 a12 ··· a1n

6 a21 7
6 7
a22 ··· a2n

6 7
6 7
. . . .
. . . .
6 . . ··· . .
7
6a(s−1)1 7
+ det 6 7.
a(s−1)2 ··· a(s−1)n
6 c 7
6 s1 cs2 ··· csn
7
6a 7
6 (s+1)1 a(s+1)2 ··· a(s+1)n
7
6 .7
4 .
.
.
.
.
. ···
.
.
.
.5
.

an1 an2 ··· ann

An analogous result holds for columns.

Proof: Put 2 3
a11 a12 ··· a1n

6 a21 7
6 7
a22 ··· a2n

6 . . . . 7
6 . . . . 7
6 . . ··· . .
7
6 a(s−1)1 7
S=6 7,
a(s−1)2 ··· a(s−1)n
6b + c 7
6 s1 s1 bs2 + cs2 ··· bsn + csn
7
6a 7
6 (s+1)1 a(s+1)2 ··· a(s+1)n
7
6 .7
4 .5
. . .
. . .
. . ··· . .

an1 an2 ··· ann

146 Chapter 7

2 3
a11 a12 ··· a1n

6 a21 7
6 7
a22 ··· a2n

6 . . . .7
6 . . . .7
6 . . ··· . .
7
6a(s−1)1 7
T =6 7
a(s−1)2 ··· a(s−1)n
6 b 7
6 s1 bs2 ··· bsn
7
6a 7
6 (s+1)1 a(s+1)2 ··· a(s+1)n
7
6 .7
4 .5
. . .
. . .
. . ··· . .

an1 an2 ··· ann

and

2 3
a11 a12 ··· a1n

6 a21 7
6 7
a22 ··· a2n

6 . . . . 7
6 . . . . 7
6 . . ··· . .
7
6a(s−1)1 7
U=6 7.
a(s−1)2 ··· a(s−1)n
6 cs1 7
6 cs2 ··· csn
7
6a 7
6 (s+1)1 a(s+1)2 ··· a(s+1)n
7
6 .7
4 .
.
.
.
.
. ···
.
.
.
.5
.

an1 an2 ··· ann

Then

P
det S = σ∈S n sgn(σ)a1σ(1) a2σ(2) · · · a(s−1)σ(s−1) (bsσ(s)

+csσ(s) )a(s+1)σ(s+1) · · · anσ(n)

P
= σ∈S n sgn(σ)a1σ(1) a2σ(2) · · · a(s−1)σ(s−1) bsσ(s) a(s+1)σ(s+1) · · · anσ(n)
P
+ σ∈S n sgn(σ)a1σ(1) a2σ(2) · · · a(s−1)σ(s−1) csσ(s) a(s+1)σ(s+1) · · · anσ(n)

= det T + det U.

By applying the above argument to AT , we obtain the result for columns.

610 Lemma If two rows or two columns of A ∈ Mn (F), A = [aij ] are identical, then det A = 0F .

Proof: Suppose asj = atj for s 6= t and for all j ∈ [1; n]. In particular, this means that for any
σ ∈ Sn we have asσ(t) = atσ(t) and atσ(s) = asσ(s) . Let τ be the transposition

2 3
6 s t 7
τ=6
4
7.
5
τ(t) τ(s)

Then στ(a) = σ(a) for a ∈ {1, 2, . . . , n} \ {s, t}. Also, sgn(στ) = sgn(σ)sgn(τ) = −sgn(σ).
As σ runs through all even permutations, στ runs through all odd permutations, and viceversa.
Determinants 147

Therefore

P
detA = σ∈Sn sgn(σ)a1σ(1) a2σ(2) · · · asσ(s) · · · atσ(t) · · · anσ(n)
P
= σ∈Sn sgn(σ)a1σ(1) a2σ(2) · · · asσ(s) · · · atσ(t) · · · anσ(n)
sgn(σ)=1


+sgn(στ)a1στ(1) a2στ(2) · · · asστ(s) · · · atστ(t) · · · anστ(n)
P
= σ∈Sn sgn(σ) a1σ(1) a2σ(2) · · · asσ(s) · · · atσ(t) · · · anσ(n)
sgn(σ)=1


−a1σ(1) a2σ(2) · · · asσ(t) · · · atσ(s) · · · anσ(n)
P
= σ∈Sn sgn(σ) a1σ(1) a2σ(2) · · · asσ(s) · · · atσ(t) · · · anσ(n)
sgn(σ)=1


−a1σ(1) a2σ(2) · · · atσ(t) · · · asσ(s) · · · anσ(n)

= 0F .

Arguing on AT will yield the analogous result for the columns. ❑

611 Corollary If two rows or two columns of A ∈ Mn (F), A = [aij ] are proportional, then det A = 0F .

Proof: Suppose asj = αatj for s 6= t and for all j ∈ [1; n]. If B is the matrix obtained by pulling
out the factor α from the s-th row then det A = α det B. But now the s-th and the t-th rows in B
are identical, and so det B = 0F . Arguing on AT will yield the analogous result for the columns.

612 Example
2 3 2 3
61 a b7 61 1 b7
6 7 6 7
6 7 6 7
det 6
61 a 7 = a det 6
c7 61 1 c7
7 = 0,
6 7 6 7
4 5 4 5
1 a d 1 1 d

since on the last determinant the first two columns are identical.

613 Theorem (Multilinearity of Determinants) Let A ∈ Mn (F), A = [aij ] and α ∈ F. If X ∈ Mn (F), X = (xij ) is
the matrix obtained by the row transvection Rs +αRt → Rs then det X = det A. Similarly, if Y ∈ Mn (F), Y = (yij )
is the matrix obtained by the column transvection Cs + αCt → Cs then det Y = det A.

Proof: For the row transvection it suffices to take bsj = asj , csj = αatj for j ∈ [1; n] in Lemma
609. With the same notation as in the lemma, T = A, and so,

det X = det T + det U = det A + det U.

But U has its s-th and t-th rows proportional (s 6= t), and so by Corollary 611 det U = 0F . Hence
det X = det A. To obtain the result for column transvections it suffices now to also apply Theorem
602. ❑
148 Chapter 7

614 Example Demonstrate, without actually calculating the determinant that

2 3
62 9 97
6 7
6 7
det 6
64 6
7
87
6 7
4 5
7 4 1

is divisible by 13.

Solution: Observe that 299, 468 and 741 are all divisible by 13. Thus

2 3 2 3 2 3
62 9 97 62 9 2997 62 9 237
6 7 6 7 6 7
6 7 C3 +10C2 +100C1 → C3 6 7 6 7
det 6
64 6
7
87 = det 6
64 6
7 = 13 det 6
4687 64 6
7
367 ,
6 7 6 7 6 7
4 5 4 5 4 5
7 4 1 7 4 741 7 4 57

which shews that the determinant is divisible by 13.

615 Theorem The determinant of a triangular matrix (upper or lower) is the product of its diagonal elements.

Proof: Let A ∈ Mn (F), A = [aij ] be a triangular matrix. Observe that if σ 6= Id then

aiσ(i) aσ(i)σ2 (i) = 0F occurs in the product

a1σ(1) a2σ(2) · · · anσ(n) .

Thus

P
det A = σ∈Sn sgn(σ)a1σ(1) a2σ(2) · · · anσ(n)

= sgn(Id )a1Id (1) a2Id (2) · · · anId (n) = a11 a22 · · · ann .

616 Example The determinant of the n × n identity matrix In over a field F is

det In = 1F .

617 Example Find

2 3
61 2 37
6 7
6 7
det 6
64 5
7
67 .
6 7
4 5
7 8 9
Determinants 149

Solution: We have 2 3 2 3
61 2 37 61 0 0 7
6 7 C2 −2C1 → C2 6 7
6 7 C3 −3C1 → C3 6 7
det 6
64 5
7
67 det 6
64 −3
7
−6 7
6 7 6 7
4 5 4 5
7 8 9 7 −6 −12
2 3
61 0 07
6 7
6 7
= (−3)(−6) det 6
64 1
7
17
6 7
4 5
7 2 2

= 0,

since in this last matrix the second and third columns are identical and so Lemma 610 applies.

618 Theorem Let (A, B) ∈ (Mn (F))2 . Then

det(AB) = (det A)(det B).

Pn
Proof: Put D = AB, D = (dij ), dij = k=1 aik bkj . If A(c:k) , D(c:k) , 1 ≤ k ≤ n denote the
columns of A and D, respectively, observe that
n
X
D(c:k) = blk A(c:l) , 1 ≤ k ≤ n.
l=1

Applying Corollary 605 and Lemma 609 multiple times, we obtain

det D = det(D(c:1) , D(c:2) , . . . , D(c:n) )

Pn Pn Pn
= j1 =1 j2 =1 ··· jn =1 b1j1 b2j2 · · · bnjn

· det(A(c:j1 ) , A(c:j2 ) , . . . , A(c:jn ) ).

By Lemma 610, if any two of the A(c:jl ) are identical, the determinant on the right vanishes. So
each one of the jl is different in the non-vanishing terms and so the map

{1, 2, . . . , n} → {1, 2, . . . , n}
σ:
l 7→ jl

is a permutation. Here jl = σ(l). Therefore, for the non-vanishing

det(A(c:j1 ) , A(c:j2 ) , . . . , A(c:jn ) )

we have in view of Corollary 601,

det(A(c:j1 ) , A(c:j2 ) , . . . , A(c:jn ) ) = (sgn(σ)) det(A(c:1) , A(c:2) , . . . , A(c:n) )

= (sgn(σ)) det A.
150 Chapter 7

We deduce that

det(AB) = det D
Pn
= jn =1 b1j1 b2j2 · · · bnjn det(A(c:j1 ) , A(c:j2 ) , . . . , A(c:jn ) )
P
= (det A) σ∈Sn (sgn(σ))b1σ(1) b2σ(2) · · · bnσ(n)

= (det A)(det B),

as we wanted to shew. ❑
By applying the preceding theorem multiple times we obtain

619 Corollary If A ∈ Mn (F) and if k is a positive integer then

det Ak = (det A)k .

620 Corollary If A ∈ GLn (F) and if k is a positive integer then det A 6= 0F and
det A−k = (det A)−k .

Proof: We have AA−1 = In and so by Theorem 618 (det A)(det A−1 ) = 1F , from where the
result follows. ❑

621 Problem Let 623 Problem After the indicated column operations on a 3×3
2 3 matrix A with det A = −540, matrices A1 , A2 , . . . , A5 are
successively obtained:
6bc ca ab7
6 7
6 7 A
C 1 +3C 2 → C 1
A1
C2 ↔ C3
A2
3C 2 −C 1 → C 2
A3
C 1 −3C 2 → C 1
A4
2C 1 → C 1
A
Ω = det 6 a b c 7 . → → → → →
6 7
4 5
Determine the numerical values of det A1 , det A2 , det A3 , det A4
a2 b2 c2 and det A5 .

Without expanding either determinant, prove that

624 Problem Let A, B, C be 3 × 3 matrices with det A =
2 3
3, det B3 = −8, det C = 2. Compute (i) det ABC, (ii)
61 1 17 det 5AC, (iii) det A3 B−3 C−1 . Express your answers as frac-
6 7
6 7 tions.
Ω = det 6a2 b2 c2 7 .
6 7
4 5
625 Problem Shew that ∀A ∈ Mn (R),
a3 b3 c3
∃(X, Y ) ∈ (Mn (R))2 , (det X)(det Y ) 6= 0
622 Problem Demonstrate that such that
2 3 A = X + Y.
6a − b − c 2a 2a 7 That is, any square matrix over R can be written as a sum of
6 7
6 7 3 two matrices whose determinant is not zero.
Ω = det 6 2b b−c−a 2b 7 = (a+b+c) .
6 7
4 5
2c 2c c−a−b 626 Problem Prove or disprove! The set X = {A ∈ Mn (F) :
det A = 0F } is a vector subspace of Mn (F).

7.4 Laplace Expansion

We now develop a more computationally convenient approach to determinants.

Put X
Cij = (sgn(σ))a1σ(1) a2σ(2) · · · anσ(n) .
σ∈Sn
σ(i)=j
Laplace Expansion 151

Then
P
det A = σ∈Sn (sgn(σ))a1σ(1) a2σ(2) · · · anσ(n)
Pn P
= i=1 aij σ∈Sn (sgn(σ))a1σ(1) a2σ(2)
σ(i)=j (7.5)
· · · a(i−1)σ(i−1) a(i+1)σ(i+1) · · · anσ(n)
Pn
= i=1 aij Cij ,

is the expansion of det A along the j-th column. Similarly,

P
det A = σ∈Sn (sgn(σ))a1σ(1) a2σ(2) · · · anσ(n)
Pn P
= j=1 aij σ∈Sn (sgn(σ))a1σ(1) a2σ(2)
σ(i)=j

· · · a(i−1)σ(i−1) a(i+1)σ(i+1) · · · anσ(n)

Pn
= j=1 aij Cij ,

is the expansion of det A along the i-th row.

627 Definition Let A ∈ Mn (F), A = [aij ]. The ij-th minor Aij ∈ Mn−1 (R) is the (n − 1) × (n − 1) matrix
obtained by deleting the i-th row and the j-th column from A.

628 Example If
2 3
61 2 37
6 7
6 7
A=6
64 5
7
67
6 7
4 5
7 8 9

then, for example,

2 3 2 3 2 3 2 3 2 3
65 67 64 67 62 37 61 37 61 27
A11 = 6
4
7,
5 A12 = 6
4
7,
5 A21 = 6
4
7,
5 A22 = 6
4
7,
5 A33 = 6
4
7.
5
8 9 7 9 8 9 7 9 4 5

629 Theorem Let A ∈ Mn (F). Then

n
X n
X
det A = aij (−1)i+j det Aij = aij (−1)i+j det Aij .
i=1 j=1

Proof: It is enough to shew, in view of 7.5 that

(−1)i+j det Aij = Cij .

Now,
P
Cnn = σ∈Sn sgn(σ)a1σ(1) a2σ(2) · · · a(n−1)σ(n−1)
σ(n)=n
P
= τ∈Sn−1 sgn(τ)a1τ(1) a2τ(2) · · · a(n−1)τ(n−1)

= det Ann ,
152 Chapter 7

since the second sum shewn is the determinant of the submatrix obtained by deleting the last row
and last column from A.

To find Cij for general ij we perform some row and column interchanges to A in order to bring aij
to the nn-th position. We thus bring the i-th row to the n-th row by a series of transpositions, first
swapping the i-th and the (i + 1)-th row, then swapping the new (i + 1)-th row and the (i + 2)-th
row, and so forth until the original i-th row makes it to the n-th row. We have made thereby
n − i interchanges. To this new matrix we perform analogous interchanges to the j-th column,
thereby making n − j interchanges. We have made a total of 2n − i − j interchanges. Observe that
(−1)2n−i−j = (−1)i+j . Call the analogous quantities in the resulting matrix A ′ , Cnn
′ ′
, Ann . Then

′ ′
Cij = Cnn = det Ann = (−1)i+j det Aij ,

by virtue of Corollary 601.

❑

☞ It is irrelevant which row or column we choose to expand a determinant of a square matrix.

We always obtain the same result. The sign pattern is given by
2 3
6+ − + − ···7
6 7
6 .. 7
6− + − + . 7
6 7
6 7
6 .. 7
6 7
6+ − + − . 7
6 7
4. .. .. .. .. 5
.. . . . .

630 Example Find

2 3
61 2 37
6 7
6 7
det 6
64 5
7
67
6 7
4 5
7 8 9

by expanding along the first row.

Solution: We have
2 3 2 3 2 3
65 67 64 67 64 57
det A = 1(−1)1+1 det 6
4
7 + 2(−1)1+2 det 6
5 4
7 + 3(−1)1+3 det 6
5 4
7
5
8 9 7 9 7 8
= 1(45 − 48) − 2(36 − 42) + 3(32 − 35) = 0.

631 Example Evaluate the Vandermonde determinant

2 3
61 1 17
6 7
6 7
det 6
6a b
7
c 7.
6 7
4 5
a2 b2 c2
Laplace Expansion 153

Solution:

2 3 2 3
61 1 17 61 0 0 7
6 7 6 7
6 7 6 7
det 6
6a b
7
c7 = det 6
6a b−a
7
c−a 7
6 7 6 7
4 5 4 5
a2 b2 c2 a2 b2 − a2 c2 − a2
2 3
6 b−a c−a 7
= det 6
4
7
5
b2 − c 2 c2 − a2
2 3
6 1 1 7
= (b − a)(c − a) det 6
4
7
5
b+a c+a

= (b − a)(c − a)(c − b).

632 Example Evaluate the determinant

2 3
6 1 2 3 4 ··· 20007
6 7
6 7
6 7
6 2 1 2 3 ··· 19997
6 7
6 7
6 7
6 3 2 1 2 ··· 19987
6 7
det A = det 6 7.
6 7
6 4 3 2 1 ··· 19977
6 7
6 7
6 7
6 ··· ··· ··· ··· ··· ··· 7
6 7
6 7
4 5
2000 1999 1998 1997 · · · 1

Solution: Applying Rn − Rn+1 → Rn for 1 ≤ n ≤ 1999, the determinant becomes

2 3
6 −1 1 1 1 ··· 1 1 7
6 7
6 7
6 7
6 −1 −1 1 1 ··· 1 1 7
6 7
6 7
6 7
6 −1 −1 −1 1 ··· 1 1 7
6 7
6 7
6 7
det 6 −1 −1 −1 −1 ··· 1 1 7 .
6 7
6 7
6 7
6 ··· ··· ··· ··· ··· ··· ···7
6 7
6 7
6 7
6 −1 −1 −1 −1 ··· −1 1 7
6 7
6 7
4 5
2000 1999 1998 1997 · · · 2 1
154 Chapter 7

Applying now Cn + C2000 → Cn for 1 ≤ n ≤ 1999, we obtain

2 3
6 0 2 2 2 ··· 2 1 7
6 7
6 7
6 7
6 0 0 2 2 ··· 2 1 7
6 7
6 7
6 7
6 0 0 0 2 ··· 2 1 7
6 7
6 7
6 7
det 6 0 0 0 0 ··· 2 1 7 .
6 7
6 7
6 7
6 ··· ··· ··· ··· ··· ··· ···7
6 7
6 7
6 7
6 0 0 0 0 ··· 0 1 7
6 7
6 7
4 5
2001 2000 1999 1998 · · · 3 1

This last determinant we expand along the first column. We have

2 3
6 2 2 2 ··· 2 1 7
6 7
6 7
6 7
6 0 2 2 ··· 2 1 7
6 7
6 7
6 7
6 0 0 2 ··· 2 1 7
6 7
−2001 det 6 7 = −2001(21998).
6 7
6 0 0 0 ··· 2 1 7
6 7
6 7
6 7
6· · · ··· ··· ··· ··· ···7
6 7
6 7
4 5
0 0 0 ··· 0 1

633 Definition Let A ∈ Mn (F). The classical adjoint or adjugate of A is the n × n matrix adj (A) whose entries
are given by
[adj (A)]ij = (−1)i+j det Aji ,
where Aji is the ji-th minor of A.

634 Theorem Let A ∈ Mn (F). Then

(adj (A))A = A(adj (A)) = (det A)In .

Proof: We have
Pn
[A(adj (A))]ij = k=1 aik [adj (A)]kj
Pn
= k=1 aik (−1)i+k det Ajk .
Now, this last sum is det A if i = j by virtue of Theorem 629. If i 6= j it is 0, since then the j-th
row is identical to the i-th row and this determinant is 0F by virtue of Lemma 610. Thus on the
diagonal entries we get det A and the off-diagonal entries are 0F . This proves the theorem. ❑
The next corollary follows immediately.

635 Corollary Let A ∈ Mn (F). Then A is invertible if and only det A 6= 0F and
adj (A)
A−1 = .
det A
Laplace Expansion 155

636 Problem Find 641 Problem Use induction to shew that

2 3
2 3
61 1 1 ··· 1 17
61 2 37 6 7
6 7 6 .. 7
6 7 61 0 0 . 0 07
det 64 5 67 6 7
6 7 6 7
4 5 6 7
60 1 0 ··· 0 07
7 8 9 6 7 n+1
det 6 7 = (−1) .
6 7
60 0 1 ··· 0 0 7
by expanding along the second column. 6 7
6. 7
6. .. .. .. 7
6. . ··· . . 7
6 7
637 Problem Compute the determinant 4 5
2 3 0 0 0 ··· 1 0

6 1 0 −1 1 7
6 7
6 7
6 2 0 0 1 7 642 Problem Let
6 7 2 3
det 6 7.
6 7
6666 −3 −1 10000007 61 n n n ··· n7
6 7 6 7
4 5 6 .. 7
6n 2 n n . n7
1 0 0 1 6 7
6 7
6 7
6n n 3 n ··· n7
6 7
638 Problem If A=6 7,
6 7
2 3 6n n n 4 ··· n7
6 7
6. 7
61 1 1 17 6. .. .. .. 7
6 7 6. . . ··· . 7
6 7 6 7
6x a 0 07 4 5
6 7
det 6 7 = 0, n n n n n n
6 7
6x 0 b 07
6 7 that is, A ∈ Mn (R), A = [aij ] is a matrix such that akk = k
4 5
and aij = n when i 6= j. Find det A.
x 0 0 c

643 Problem Let n ∈ N, n >
1 be an odd
integer.

Recall that
and xabc 6= 0, prove that
the binomial coefficients n
k
satisfy n
n
= n0 = 1 and that
1 1 1 1 for 1 ≤ k ≤ n,
= + + .      
x a b c n n−1 n−1
= + .
k k−1 k
639 Problem Prove that
Prove that
2 3 2 3
n

n

n


60 a b 07 6 1 1 2
··· n−1
1
7
6 7 6 7
6 7 6


7
6a 0 b 07 6 1 1 n
··· n n 7
6 7 6 1 n−2 n−1 7
det 6 7 = 2ab(a − b). 6 7
6 7 6


7 n n
60 a 0 b7 det 6 n 1 1 ··· n n 7 = (1+(−1) ) .
6 7 6 n−1 n−3 n−2 7
4 5 6 7
6 7
1 1 1 1 6 ··· ··· ··· ··· ··· ··· 7
6 7
4


5
n n n
1 2 3
··· 1 1
640 Problem Demonstrate that
2 3
644 Problem Let A ∈ GLn (F), n > 1. Prove that
6a 0 b 07
det(adj (A)) = (det A)n−1 .
6 7
6 7
60 a 0 b7
6 7 2
det 6 7 = (ad − bc) . 645 Problem Let (A, B, S) ∈ (GLn (F))3 . Prove that
6 7
6c 0 d 07 ➊ adj (adj (A)) = (det A)n−2 A.
6 7
4 5 ➋ adj (AB) = adj (A) adj (B).


0 c 0 d ➌ adj SAS−1 = S(adj (A))S−1 .

646 Problem If A ∈ GL2 (F), , and let k be a positive integer.

Prove that det(adj · · · adj(A)) = det A.
| {z }
k
156 Chapter 7

7.5 Determinants and Linear Systems

647 Theorem Let A ∈ Mn (F). The following are all equivalent
➊ det A 6= 0F .
➋ A is invertible.
➌ There exists a unique solution X ∈ Mn×1 (F) to the equation AX = Y.
➍ If AX = 0n×1 then X = 0n×1 .

Proof: We prove the implications in sequence:

➊ =⇒ ➋: follows from Corollary 635
➋ =⇒ ➌: If A is invertible and AX = Y then X = A−1 Y is the unique solution of this equation.
➌ =⇒ ➍: follows by putting Y = 0n×1
➍ =⇒ ➊: Let R be the row echelon form of A. Since RX = 0n×1 has only X = 0n×1 as a solution,
every entry on the diagonal of R must be non-zero, R must be triangular, and hence det R 6= 0F .
Since A = PR where P is an invertible n × n matrix, we deduce that det A = det P det R 6= 0F .
❑

The contrapositive form of the implications ➊ and ➍ will be used later. Here it is for future reference.

648 Corollary Let A ∈ Mn (F). If there is X 6= 0n×1 such that AX = 0n×1 then det A = 0F .
Chapter 8
Eigenvalues and Eigenvectors

8.1 Similar Matrices

649 Definition We say that A ∈ Mn (F) is similar to B ∈ Mn (F) if there exist a matrix P ∈ GLn (F) such that

B = PAP−1 .

650 Theorem Similarity is an equivalence relation.

Proof: Let A ∈ Mn (F). Then A = In AI−1

n , so similarity is reflexive. If B = PAP
−1
(P ∈
−1 −1
GLn (F) ) then A = P BP so similarity is symmetric. Finally, if B = PAP and C = QBQ−1
(P ∈ GLn (F) , Q ∈ GLn (F)) then C = QPAP−1 Q−1 = QPA(QP)−1 and so similarity is
transitive. ❑

Since similarity is an equivalence relation, it partitions the set of n × n matrices into equivalence classes by Theorem
37.

651 Definition A matrix is said to be diagonalisable if it is similar to a diagonal matrix.

Suppose that
2 3
6 λ1 0 0 ··· 07
6 7
6 7
60 λ2 0 ··· 0 7
6 7
A=6
6 .
7.
7
6 . .. .. .. 7
6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λn

Then if K is a positive integer

2 3
K
6λ1 0 0 ··· 07
6 7
6 7
6 0 λK 0 ··· 0 7
6 2 7
AK =6
6
7.
7
6 .. .. .. .. 7
6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λK
n

157
158 Chapter 8

In particular, if B is similar to A then

2 3
K
6 λ1 0 0 ··· 07
6 7
6 7
6 λK 0 ··· 7
6 0 2 0 7 −1
BK = (PAP−1 )(PAP−1 ) · · · (PAP−1 ) = PAK P−1 = P6
6
7P ,
7
| {z } 6 .. .. .. .. 7
K factors 6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λK
n

so we have a simpler way of computing BK . Our task will now be to establish when a particular square matrix is
diagonalisable.

8.2 Eigenvalues and Eigenvectors

Let A ∈ Mn (F) be a square diagonalisable matrix. Then there exist P ∈ GLn (F) and a diagonal matrix D ∈
Mn (F) such that P−1 AP = D, whence AP = DP. Put
2 3
6 λ1 0 0 ··· 07
6 7
6 7
60 ··· 7
6 λ2 0 0 7
D=6
6 .
7,
7 P = [P1 ; P2 ; · · · ; Pn ],
6 . .. .. .. 7
6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λn

where the Pk are the columns of P. Then

AP = DP =⇒ [AP1 ; AP2 ; · · · ; APn ] = [λ1 P1 ; λ2 P2 ; · · · ; λn Pn ],

from where it follows that APk = λk Pk . This motivates the following definition.

652 Definition Let V be a finite-dimensional vector space over a field F and let T : V → V be a linear transformation.
A scalar λ ∈ F is called an eigenvalue of T if there is a v 6= 0 (called an eigenvector) such that T (v) = λv.

653 Example Shew that if λ is an eigenvalue of T : V → V, then λk is an eigenvalue of T k : V → V, for k ∈ N \ {0}.

Solution: Assume that T (v) = λv. Then

T 2 (v) = TT (v) = T (λv) = λT (v) = λ(λv) = λ2 v.

Continuing the iterations we obtain T k (v) = λk v, which is what we want.

654 Theorem Let A ∈ Mn (F) be the matrix representation of T : V → V. Then λ ∈ F is an eigenvalue of T if an

only if det(λIn − A) = 0F .

Proof: λ is an eigenvalue of A ⇐⇒ there is v 6= 0 such that Av = λv ⇐⇒ λv − Av = 0 ⇐⇒

λIn v − Av = 0 ⇐⇒ det(λIn − A) = 0F by Corollary 648.❑

655 Definition The equation

det(λIn − A) = 0F
is called the characteristic equation of A or secular equation of A. The polynomial p(λ) = det(λIn − A) is the
characteristic polynomial of A.
Eigenvalues and Eigenvectors 159

2 3
61 1 0 07
6 7
6 7
61 07
6 1 0 7
656 Example Let A = 6
6
7. Find
7
6 7
60 0 1 17
6 7
4 5
0 0 1 1

➊ The characteristic polynomial of A.

➋ The eigenvalues of A.

➌ The corresponding eigenvectors.

Solution: We have

➊
2 3
6λ − 1 −1 0 0 7
6 7
6 7
6 −1 λ−1 0 0 7
6 7
det(λI4 − A) = det 6
6
7
7
6 7
6 0 0 λ−1 −1 7
6 7
4 5
0 0 −1 λ−1
2 3 2 3
6λ − 1 0 0 7 6−1 0 0 7
6 7 6 7
6 7 6 7
= (λ − 1) det 6
6 0 λ−1
7 + det 6
−1 7 6 0 λ−1
7
−1 7
6 7 6 7
4 5 4 5
0 −1 λ−1 0 −1 λ−1

= (λ − 1)((λ − 1)((λ − 1)2 − 1)) + (−((λ − 1)2 − 1))

= (λ − 1)((λ − 1)(λ − 2)(λ)) − (λ − 2)(λ)

= (λ − 2)(λ)((λ − 1)2 − 1)

= (λ − 2)2 (λ)2

➋ The eigenvalues are clearly λ = 0 and λ = 2.

➌ If λ = 0, then
2 3
6−1 −1 0 0 7
6 7
6 7
6−1 −1 0 0 7
6 7
0I4 − A = 6 7.
6 7
6 7
6 0 0 −1 −17
6 7
4 5
0 0 −1 −1
160 Chapter 8

This matrix has row-echelon form

2 3
6−1 −1 0 0 7
6 7
6 7
6 0 0 −1 −17
6 7
6 7,
6 7
6 7
6 0 0 0 0 7
6 7
4 5
0 0 0 0

and if

2 32 3 2 3
6−1 −1 0 0 7 6a7 607
6 76 7 6 7
6 76 7 6 7
6 0 0 −1 −17 6 7 6 7
6 7 6b7 607
6 76 7 = 6 7,
6 76 7 6 7
6 76 7 6 7
6 0 0 0 0 7 6 c 7 607
6 76 7 6 7
4 54 5 4 5
0 0 0 0 d 0

then c = −d and a = −b
Thus the general solution of the system (0I4 − A)X = 0n×1 is
2 3 2 3 2 3
6a7 6 1 7 6 0 7
6 7 6 7 6 7
6 7 6 7 6 7
6b7 6−17 6 0 7
6 7 6 7 6 7
6 7 = a6 7 +c6 7.
6 7 6 7 6 7
6 7 6 7 6 7
6 c7 6 0 7 6 1 7
6 7 6 7 6 7
4 5 4 5 4 5
d 0 −1

If λ = 2, then
2 3
6 1 −1 0 0 7
6 7
6 7
6−1 1 0 0 7
6 7
2I4 − A = 6 7.
6 7
6 7
6 0 0 1 −17
6 7
4 5
0 0 −1 1

This matrix has row-echelon form

2 3
6−1 1 0 0 7
6 7
6 7
6 0 0 1 −17
6 7
6 7,
6 7
6 7
6 0 0 0 0 7
6 7
4 5
0 0 0 0

and if
Eigenvalues and Eigenvectors 161

2 32 3 2 3
61 −1 0 07 6 a 7 607
6 76 7 6 7
6 76 7 6 7
60 −1 17 6 7 6 7
6 0 7 6b7 607
6 76 7 = 6 7,
6 76 7 6 7
6 76 7 6 7
60 0 0 07 6 c 7 607
6 76 7 6 7
4 54 5 4 5
0 0 0 0 d 0

then c = d and a = b
Thus the general solution of the system (2I4 − A)X = 0n×1 is

2 3 2 3 2 3
6a7 617 607
6 7 6 7 6 7
6 7 6 7 6 7
6b7 617 607
6 7 6 7 6 7
6 7 = a6 7 + c6 7.
6 7 6 7 6 7
6 7 6 7 6 7
6 c7 607 617
6 7 6 7 6 7
4 5 4 5 4 5
d 0 1

Thus for λ = 0 we have the eigenvectors

2 3 2 3
6 1 7 6 0 7
6 7 6 7
6 7 6 7
6−17 6 0 7
6 7 6 7
6 7,6 7
6 7 6 7
6 7 6 7
6 0 7 6 1 7
6 7 6 7
4 5 4 5
0 −1

and for λ = 2 we have the eigenvectors

2 3 2 3
617 607
6 7 6 7
6 7 6 7
617 607
6 7 6 7
6 7,6 7.
6 7 6 7
6 7 6 7
607 617
6 7 6 7
4 5 4 5
0 1

657 Theorem If λ = 0F is an eigenvalue of A, then A is non-invertible.

Proof: Put p(λ) = det(λIn − A). Then p(0F ) = det(−A) = (−1)n det A is the constant term
of the characteristic polynomial. If λ = 0F is an eigenvalue then

p(0F ) = 0F =⇒ det A = 0F ,

and hence A is non-invertible by Theorem 647. ❑

658 Theorem Similar matrices have the same characteristic polynomial.

162 Chapter 8

Proof: We have

det(λIn − SAS−1 ) = det(λSIn S−1 − SAS−1 )

= det S(λIn − A)S−1

= (det S)(det(λIn − A))(det S−1 )

 
1
= (det S)(det(λIn − A))
det S
= det(λIn − A),

from where the result follows.❑

2 3
659 Problem Find the eigenvalues and eigenvectors of
2 3 6 0 2 −17
−17 6 7
6 1 6
660 Problem Let A = 6 2
7
A=4 5 6 3 −27
7
. Find
−1 1 4 5
−1 −2 0

➊ The characteristic polynomial of A.

➋ The eigenvalues of A.
➌ The corresponding eigenvectors.

8.3 Diagonalisability
In this section we find conditions for diagonalisability.

661 Theorem Let {v1 , v2 , . . . , vk } ⊂ V be the eigenvectors corresponding to the different eigenvalues {λ1 , λ2 , . . . , λk }
(in that order). Then these eigenvectors are linearly independent.

Proof: Let T : V → V be the underlying linear transformation. We proceed by induction. For

k = 1 the result is clear. Assume that every set of k−1 eigenvectors that correspond to k−1 distinct
eigenvalues is linearly independent and let the eigenvalues λ1 , λ2 , . . . , λk−1 have corresponding
eigenvectors v1 , v2 , . . . , vk−1 . Let λ be a eigenvalue different from the λ1 , λ2 , . . . , λk−1 and let
its corresponding eigenvector be v. If v were linearly dependent of the v1 , v2 , . . . , vk−1 , we would
have
xv + x1 v1 + x2 v2 + · · · + xk−1 vk−1 = 0. (8.1)
Now
T (xv + x1 v1 + x2 v2 + · · · + xk−1 vk−1 ) = T (0) = 0,
by Theorem 538. This implies that

xλv + x1 λ1 v1 + x2 λ2 v2 + · · · + xk−1 λk−1 vk−1 = 0. (8.2)

From 8.2 take away λ times 8.1, obtaining

x1 (λ1 − λ)v1 + x2 (λ2 v2 + · · · + xk−1 (λk−1 − λ)vk−1 = 0 (8.3)

Since λ − λi 6= 0F 8.3 is saying that the eigenvectors v1 , v2 , . . . , vk−1 are linearly dependent,
a contradiction. Thus the claim follows for k distinct eigenvalues and the result is proven by
induction. ❑

662 Theorem A matrix A ∈ Mn (F) is diagonalisable if and only if it possesses n linearly independent eigenvectors.
Diagonalisability 163

Proof: Assume first that A is diagonalisable, so there exists P ∈ GLn (F) and
2 3
6λ 1 0 0 ··· 07
6 7
6 7
60 λ2 0 ··· 0 7
6 7
D=6
6 .
7
7
6 . .. .. .. 7
6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λn

such that 2 3
6 λ1 0 0 ··· 07
6 7
6 7
6 0 λ2 0 ··· 0 7
6 7
P−1 AP = 6
6
7.
7
6 .. .. .. .. 7
6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λn

Then
2 3
6 λ1 0 0 ··· 07
6 7
6 7
60 λ2 0 ··· 0 7
6 7
[AP1 ; AP2 ; · · · ; APn ] = AP = P 6
6 .
7 = [λ1 P1 ; λ2 P2 ; · · · ; λn Pn ],
7
6 . .. .. .. 7
6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λn

where the Pk are the columns of P. Since P is invertible, the Pk are linearly independent by virtue
of Theorems 473 and 647.

Conversely, suppose now that v1 , . . . , vn are n linearly independent eigenvectors, with correspond-
ing eigenvalues λ1 , λ2 , . . . , λn . Put
2 3
6 λ1 0 0 ··· 07
6 7
6 7
60 λ2 0 ··· 0 7
6 7
P = [v1 ; . . . ; vn ], D=6
6 .
7.
7
6 . .. .. .. 7
6 . . . ··· . 7
6 7
4 5
0 0 0 ··· λn

Since Avi = λi vi we see that AP = PD. Again P is invertible by Theorems 473 and 647 since
the vk are linearly independent. Left multiplying by P−1 we deduce P−1 AP = D, from where A is
diagonalisable. ❑

663 Example Shew that the following matrix is diagonalisable:

2 3
6 1 −1 −17
6 7
6 7
6
6 1 3 1 7
7
6 7
4 5
−3 1 −1
164 Chapter 8

and find a diagonal matrix D and an invertible matrix P such that

A = PDP−1 .

Solution: Verify that the characteristic polynomial of A is

λ3 − 3λ2 − 4λ + 12 = (λ − 2)(λ + 2)(λ − 3).

The eigenvector for λ = −2 is

2 3
6 1 7
6 7
6 7
6 7
6−17 .
6 7
4 5
4

The eigenvector for λ = 2 is

2 3
6−17
6 7
6 7
6 7
6 0 7.
6 7
4 5
1

The eigenvector for λ = 3 is

2 3
6−17
6 7
6 7
6 7
6 1 7.
6 7
4 5
1

We may take
2 3 2 3
6−2 0 07 6 1 −1 4 7
6 7 6 7
6 7 6 7
D=6
6 0 2
7, P = 6
07 6−1 0 1
7.
7
6 7 6 7
4 5 4 5
0 0 3 −1 1 1

We also find 2 3
1 1
65 −1 57
6 7
6 7
−1
P =6
60 −1
7
17 .
6 7
4 5
1 1
5
0 5
2 3 2 3
664 Problem Let A be a 2 × 2 matrix with eigenvalues 1 and
2 3 2 3 617 617
6 7 6 7
617 6 1 7 6 7 6 7
One of the eigenvalues has two eigenvectors 607 and 617.
−2 and corresponding eigenvectors 4 5 and 4 5, respec-
6 7 6 7
0 −1 4 5 4 5
0 0
tively. Determine A10 .

665 Problem Let A ∈ M3 (R) have characteristic polynomial

(λ + 1)2 (λ − 3).
The Minimal Polynomial 165

2 3
1. Demonstrate that U2 = nU.
61 7
6 7 2. Find det U.
6 7
The other eigenvalue has corresponding eigenvector 617. De-
6 7 3. Prove that det(λIn − U) = λn−1 (λ − n).
4 5
1 4. Shew that dim ker (U) = n − 1.
termine A. 5. Shew that
2 3
666 Problem Let 6n 0 ··· 07
2 3 6 7
6 7
60 0 ··· 07
60 0 0 17 6 7 −1
6 7 U = P6 7P ,
6 7 6 .. .. .. .. 7
60 0 1 07 6. . . .7
6 7 6 7
A=6 7. 4 5
6 7
60 1 0 07 0 0 ··· 0
6 7
4 5
1 0 0 0 where
2 3
1. Find det A.
2. Find A−1 . 61 1 0 ··· 0 0 7
6 7
3. Find rank (A − I4 ). 6 7
61 0 1 ··· 0 0 7
6 7
4. Find det(A − I4 ). 6 7
6 .. .. .. 7
5. Find the characteristic polynomial of A. 61 0 0 . . . 7
6 7
6. Find the eigenvalues of A. P=6 7.
6 .. .. .. .. .. .. 7
7. Find the eigenvectors of A. 6. . . . . . 7
6 7
6 7
8. Find A10 . 6 7
61 0 0 ··· 0 1 7
6 7
4 5
667 Problem Let U ∈ Mn (R) be a square matrix all whose 1 −1 −1 ··· −1 −1
entries are equal to 1.

8.4 The Minimal Polynomial

668 Theorem (Cayley-Hamilton) A matrix A ∈ Mn (F) satisfies its characteristic polynomial.

Proof: Put B = λIn − A. We can write

det B = det(λIn − A) = λn + b1 λn−1 + b2 λn−2 + · · · + bn ,

as det(λIn − A) is a polynomial of degree n.

Since adj (B) is a matrix obtained by using (n − 1) × (n − 1) determinants from B, we may write

adj (B) = λn−1 Bn−1 + λn−2 Bn−2 + · · · + B0 .

Hence

det(λIn − A)In = (B)(adj (B)) = (λIn − A)(adj (B)),

from where

λn In + b1 In λn−1 + b2 In λn−2 + · · · + bn In = (λIn − A)(λn−1 Bn−1 + λn−2 Bn−2 + · · · + B0 ).

By equating coefficients we deduce

166 Chapter 8

In = Bn−1

b1 In = −ABn−1 + Bn−2

b2 In = −ABn−2 + Bn−3
..
.

bn−1 In = −AB1 + B0

bn In = −AB0 .

Multiply now the k-th row by An−k (the first row appearing is really the 0-th row):

An = An Bn−1

b1 An−1 = −An Bn−1 + An−1 Bn−2

b2 An−2 = −An−1 Bn−2 + An−2 Bn−3

..
.

bn−1 A = −A2 B1 + AB0

bn In = −AB0 .

Add all the rows and through telescopic cancellation obtain

An + b1 An−1 + · · · + bn−1 A + bn In = 0n ,
from where the theorem follows. ❑

669 Example From example 663 the matrix

2 3
6 1 −1 −17
6 7
6 7
6 7
6 1 3 1 7
6 7
4 5
−3 1 −1
has characteristic polynomial
(λ − 3)(λ − 2)(λ + 2) = λ3 − 3λ2 − 4λ + 12,
hence the inverse of this matrix can be obtained by observing that
2 3
6 1/3 1/6 −1/6 7
6 7
6 7
1
6 7
A3 − 3A2 − 4A + 12I3 = 03 =⇒ A−1 =− A − 3A − 4I3 = 6
2
6 1/6 1/3 1/6
7.
7
12 6 7
6 7
4 5
−5/6 −1/6 −1/3
Having seen that the characteristic polynomial of a square matrix A annihilates A, the question now arises of whether
there exists a polynomial of minimal degree annihilating A.
Appendix A
Answers and Hints

17
x ∈ X \ (X \ A) ⇐⇒ x ∈ X ∧ x 6∈ (X \ A)

⇐⇒ x∈X∧x∈A

⇐⇒ x ∈ X ∩ A.

18
X \ (A ∪ B) ⇐⇒ x ∈ X ∧ (x 6∈ (A ∪ B))

⇐⇒ x ∈ X ∧ (x 6∈ A ∧ x 6∈ B)

⇐⇒ (x ∈ X ∧ x 6∈ A) ∧ (x ∈ X ∧ x 6∈ B)

⇐⇒ x ∈ (X \ A) ∧ x ∈ (X \ B)

⇐⇒ x ∈ (X \ A) ∩ (X \ B).

21 One possible solution is

A ∪ B ∪ C = A ∪ (B \ A) ∪ (C \ (A ∪ B)).

23 We have
|a| = |a − b + b| ≤ |a − b| + |b|,
giving
|a| − |b| ≤ |a − b|.
Similarly,
|b| = |b − a + a| ≤ |b − a| + |a| = |a − b| + |a|,
gives
|b| − |a| ≤ |a − b|.
The stated inequality follows from this.
a
38 a ∼ a since a = 1 ∈ Z, and so the relation is reflexive. The relation is not symmetric. For 2 ∼ 1 since 21 ∈ Z but 1 ≁ 2
since 2 6∈ Z. The relation is transitive. For assume a ∼ b and b ∼ c. Then there exist (m, n) ∈ Z2 such that a
1
b
= m, bc = n.
This gives
a a b
= · = mn ∈ Z,
c b c
and so a ∼ c.
a 2 +a a 2 +a
39 Here is one possible example: put a ∼ b ⇔ b
∈ Z. Then clearly if a ∈ Z \ {0} we have a ∼ a since a
= a + 1 ∈ Z.
5 2 +5 2 2 +2 6
On the other hand, the relation is not symmetric, since 5 ∼ 2 as 2
= 15 ∈ Z but 2 6∼ 5, as 5
= 5
∈
6 Z. It is not
2 2
5 +5 3 +3 5 2 +5
transitive either, since 3
∈ Z =⇒ 5 ∼ 3 and 12
∈ Z =⇒ 3 ∼ 12 but 12
6∈ Z and so 5 ≁ 12.

167
168 Appendix A

1
41 [B] [x] = x + Z. [C] No.
3
√
55 Let ω = − 21 + i 2
3
. Then ω2 + ω + 1 = 0 and ω3 = 1. Then

x = a3 + b3 + c3 − 3abc = (a + b + c)(a + ωb + ω2 c)(a + ω2 b + cω),

y = u3 + v3 + w3 − 3uvw = (u + v + w)(u + ωv + ω2 w)(u + ω2 v + ωw).

Then
(a + b + c)(u + v + w) = au + av + aw + bu + bv + bw + cu + cv + cw,

(a + ωb + ω2 c)(u + ωv + ω2 w) = au + bw + cv

+ω(av + bu + cw)

+ω2 (aw + bv + cu),

and
(a + ω2 b + ωc)(u + ω2 v + ωw) = au + bw + cv

+ω(aw + bv + cu)

+ω2 (av + bu + cw).

This proves that

xy = (au + bw + cv)3 + (aw + bv + cu)3 + (av + bu + cw)3

−3(au + bw + cv)(aw + bv + cu)(av + bu + cw),

which proves that S is closed under multiplication.

56 We have
x⊤(y⊤z) = x⊤(y ⊗ a ⊗ z) = (x) ⊗ (a) ⊗ (y ⊗ a ⊗ z) = x ⊗ a ⊗ y ⊗ a ⊗ z,
where we may drop the parentheses since ⊗ is associative. Similarly

(x⊤y)⊤z = (x ⊗ a ⊗ y)⊤z = (x ⊗ a ⊗ y) ⊗ (a) ⊗ (z) = x ⊗ a ⊗ y ⊗ a ⊗ z.

By virtue of having proved

x⊤(y⊤z) = (x⊤y)⊤z,
associativity is established.

57 We proceed in order.
➊ Clearly, if a, b are rational numbers,

|a| < 1, |b| < 1 =⇒ |ab| < 1 =⇒ −1 < ab < 1 =⇒ 1 + ab > 0,

whence the denominator never vanishes and since sums, multiplications and divisions of rational numbers are rational,
a+b a+b
is also rational. We must prove now that −1 < < 1 for (a, b) ∈] − 1; 1[2 . We have
1 + ab 1 + ab

a+b
−1 < <1 ⇔ −1 − ab < a + b < 1 + ab
1 + ab
⇔ −1 − ab − a − b < 0 < 1 + ab − a − b

⇔ −(a + 1)(b + 1) < 0 < (a − 1)(b − 1).

Since (a, b) ∈] − 1; 1[2 , (a + 1)(b + 1) > 0 and so −(a + 1)(b + 1) < 0 giving the sinistral inequality. Similarly
a − 1 < 0 and b − 1 < 0 give (a − 1)(b − 1) > 0, the dextral inequality. Since the steps are reversible, we have
a+b
established that indeed −1 < < 1.
1 + ab
Answers and Hints 169

a+b b+a
➋ Since a ⊗ b = = = b ⊗ a, commutativity follows trivially. Now
1 + ab 1 + ba

 
b+c
a ⊗ (b ⊗ c) = a⊗
1 + bc
 
b+c
a+
1 + bc
=  
b+c
1+a
1 + bc
a(1 + bc) + b + c a + b + c + abc
= = .
1 + bc + a(b + c) 1 + ab + bc + ca

One the other hand,

 
a+b
(a ⊗ b) ⊗ c = ⊗c
1 + ab
 
a+b
+c
1 + ab
=  
a+b
1+ c
1 + ab
(a + b) + c(1 + ab)
=
1 + ab + (a + b)c

a + b + c + abc
= ,
1 + ab + bc + ca
whence ⊗ is associative.
a+e
➌ If a ⊗ e = a then = a, which gives a + e = a + ea2 or e(a2 − 1) = 0. Since a 6= ±1, we must have e = 0.
1 + ae
a+b
➍ If a ⊗ b = 0, then = 0, which means that b = −a.
1 + ab

58 We proceed in order.

➊ Since a ⊗ b = a + b − ab = b + a − ba = b ⊗ a, commutativity follows trivially. Now

a ⊗ (b ⊗ c) = a ⊗ (b + c − bc)

= a + b + c − bc − a(b + c − bc)

= a + b + c − ab − bc − ca + abc.

One the other hand,

(a ⊗ b) ⊗ c = (a + b − ab) ⊗ c

= a + b − ab + c − (a + b − ab)c

= a + b + c − ab − bc − ca + abc,

whence ⊗ is associative.

➋ If a ⊗ e = a then a + e − ae = a, which gives e(1 − a) = 0. Since a 6= 1, we must have e = 0.

a
➌ If a ⊗ b = 0, then a + b − ab = 0, which means that b(1 − a) = −a. Since a 6= 1 we find b = − .
1−a
170 Appendix A

+ 0 1 2 3 4 5 6 7 8 9 10 · 0 1 2 3 4 5 6 7 8 9 10

0 0 1 2 3 4 5 6 7 8 9 10 0 0 0 0 0 0 0 0 0 0 0 0

1 1 2 3 4 5 6 7 8 9 10 0 1 0 1 2 3 4 5 6 7 8 9 10

2 2 3 4 5 6 7 8 9 10 0 1 2 0 2 4 6 8 10 1 3 5 7 9

3 3 4 5 6 7 8 9 10 0 1 2 3 0 3 6 9 1 4 7 10 2 5 8

4 4 5 6 7 8 9 10 0 1 2 3 4 0 4 8 1 5 9 2 6 10 3 7

5 5 6 7 8 9 10 0 1 2 3 4 5 0 5 10 4 9 3 8 2 7 1 6

6 6 7 8 9 10 0 1 2 3 4 5 6 0 6 1 7 2 8 3 9 4 10 5

7 7 0 9 10 0 1 2 3 4 5 6 7 0 7 3 10 6 2 9 5 1 8 4

8 8 9 10 0 1 2 3 4 5 6 7 8 0 8 5 2 10 7 4 1 9 6 3

9 9 10 0 1 2 3 4 5 6 7 8 9 0 9 7 5 3 1 10 8 6 4 2

10 10 0 1 2 3 4 5 6 7 8 9 10 0 10 9 8 7 6 5 4 3 2 1

Table A.1: Addition table for Z11 . Table A.2: Multiplication table Z11 .

59 We have
x◦y = (x ◦ y) ◦ (x ◦ y)

= [y ◦ (x ◦ y)] ◦ x

= [(x ◦ y) ◦ x] ◦ y

= [(y ◦ x) ◦ x] ◦ y

= [(x ◦ x) ◦ y] ◦ y

= (y ◦ y) ◦ (x ◦ x)

= y ◦ x,

proving commutativity.

71 The tables appear in tables A.1 and A.2.

72 From example 68
x2 = 5.
2 2 2 2 2 2 2
Now, the squares modulo 11 are 0 = 0, 1 = 1, 2 = 4, 3 = 9, 4 = 5, 5 = 3. Also, (11 − 4)2 = 7 = 5. Hence the
solutions are x = 4 or x = 7.

81 We have √ √ √
1 2+2 3−3 6
√ √ √ = √ √ √
2+2 3+3 6 ( 2 + 2 3)2 − (3 6)2
√ √ √
2+2 3−3 6
= √
√ 12 +
2 + √4 6 − √ 54
2+2 3−3 6
= √
√ −40 +
√4 6 √ √
( 2 + 2 3 − 3 6)(−40 − 4 6)
= √
402 − (4 6)2
√ √ √ √
( 2 + 2 3 − 3 6)(−40 − 4 6)
=
√ √1504 √
16 2 + 22 3 − 30 6 − 18
= −
376
82 Since
(−a)b−1 + ab−1 = (−a + a)b−1 = 0F b−1 = 0F ,
Answers and Hints 171

we obtain by adding −(ab−1 ) to both sides that

(−a)b−1 = −(ab−1 ).

Similarly, from
a(−b−1 ) + ab−1 = a(−b−1 + b−1 ) = a0F = 0F ,
we obtain by adding −(ab−1 ) to both sides that

a(−b−1 ) = −(ab−1 ).

93 Assume h(b) = h(a). Then

h(a) = h(b) =⇒ a 3 = b3

=⇒ a 3 − b3 = 0

=⇒ (a − b)(a2 + ab + b2 ) = 0

Now,
 
a 2 3a2
b2 + ab + a2 = b + + .
2 4
This shews that b2 + ab + a2 is positive unless both a and b are zero. Hence a − b = 0 in all cases. We have shewn that
h(b) = h(a) =⇒ a = b, and the function is thus injective.

94 We have
6a 6b
f(a) = f(b) ⇐⇒ =
2a − 3 2b − 3
⇐⇒ 6a(2b − 3) = 6b(2a − 3)

⇐⇒ 12ab − 18a = 12ab − 18b

⇐⇒ −18a = −18b

⇐⇒ a = b,

proving that f is injective. Now, if

f(x) = y, y 6= 3,
then
6x
= y,
2x − 3
that is 6x = y(2x − 3). Solving for x we find
3y
x= .
2y − 6
Since 2y − 6 6= 0, x is a real number, and so f is surjective. On combining the results we deduce that f is bijective.
2 3
61 1 17
6 7
6 7
104 A = 62 4 87 .
6 7
4 5
3 9 27

2 3
61 2 37
6 7
6 7
105 A = 62 4 67 .
6 7
4 5
3 6 9
172 Appendix A

2 3 2 3
6a + 1 0 2c 7 6 2a −4a 2c 7
6 7 6 7
6 7 6 7
106 M + N = 6 a b − 2a 0 7 , 2M = 6 0 −2a 2b 7 .
6 7 6 7
4 5 4 5
2a 0 −2 2a + 2b 0 −2

107 x = 1 and y = 4.
2 3 2 3
613 −17 65 07
108 A = 4 5, B = 4 5
15 3 6 1

111 The set of border elements is the union of two rows and two columns. Thus we may choose at most four elements from
the border, and at least one from the central 3 × 3 matrix. The largest element of this 3 × 3 matrix is 15, so any allowable
choice of does not exceed 15. The choice 25, 15, 18,l 23, 20 shews that the largest minimum is indeed 15.
2 3
62 2 7
120 4 5
0 −2

121 2 3 2 3
6 a b c7 6a + b + c b+c c7
6 7 6 7
6 7 6 7
AB = 6 c+a a+b b+c 7 , BA = 6a + b + c a+b b7
6 7 6 7
4 5 4 5
a+b+c a+b+c a+b+c a+b+c c+a a

2 3
60 2 0 37
6 7
6 7
60 2 0 37
6 7
122 AB = 04 and BA = 6 7.
6 7
60 2 0 37
6 7
4 5
0 2 0 3

123 Observe that

2 32 2 32 3 2 3
2
x7 6−4
6−4 x7 6−4 x7 616 − x 0 7
5 =4 4 54 5=4 5,
4 −x
−x 4 −x 4 0 16 − x2
√
and so we must have 16 − x2 = −1 or x = ± 17.
2 3 2 3
6 1 0 7 6 0 1 7
124 Disprove! Take A = 4 5 and B = 4 5. Then AB = B, but BA = 02 .
0 0 0 0

2 3 2 3
60 07 61 07
125 Disprove! Take for example A = 4 5 and B = 4 5. Then
1 1 1 0

2 3 2 3
6−1 07 6−1 07
A 2 − B2 = 4 5=
6 4 5 .(A + B)(A − B).
0 1 −2 1
Answers and Hints 173

2 3
6 32 −327
127 4 5.
−32 32

2 3
1001 1002
6 0 2 3 7
128 A2003 = 4 5.
1002 1001
2 3 0

130 The assertion is clearly true for n = 1. Assume that is it true for n, that is, assume
2 3
6cos(n)α − sin(n)α7
An = 4 5.
sin(n)α cos(n)α

Then

An+1 = AAn
2 32 3
6cos α − sin α7 6cos(n)α − sin(n)α7
= 4 54 5
sin α cos α sin(n)α cos(n)α
2 3
6cos α cos(n)α − sin α sin(n)α − cos α sin(n)α − sin α cos(n)α7
= 4 5
sin α cos(n)α + cos α sin(n)α − sin α sin(n)α + cos α cos(n)α
2 3
6cos(n + 1)α − sin(n + 1)α7
= 4 5,
sin(n + 1)α cos(n + 1)α

and the result follows by induction.

131 Let A = [aij ], B = [bij ] be checkered n × n matrices. Then A + B = (aij + bijP

). If j − i is odd, then aij + bij = 0 + 0 = 0,
n
which shows that A + B is checkered. Furthermore, let AB = [cij ] with cij = k=1 aik bkj . If i is even and j odd, then
aik = 0 for odd k and bkj = 0 for even k. Thus cij = 0 for i even and j odd. Similarly, cij = 0 for odd i and even j. This
proves that AB is checkered.

132 Put
2 3
60 1 17
6 7
6 7
J = 60 0 17 .
6 7
4 5
0 0 0

We first notice that

2 3
60 0 17
6 7
6 7
J 2 = 60 0 07 , J3 = 03 .
6 7
4 5
0 0 0

This means that the sum in the binomial expansion

n  
n n
X n n−k k
A = (I3 + J) = I J
k
k=0
174 Appendix A

is a sum of zero matrices for k ≥ 3. We thus have

n

n−2 2
An = In n−1
3 + nI3 J+ 2
I3 J
2 3 2 3 2 3
n


61 0 0 7 60 n n 7 60 0 2 7
6 7 6 7 6 7
6 7 6 7 6 7
= 60 1 07 +6 n7 +6 0 7
6 7 60 0 7 60 0 7
4 5 4 5 4 5
0 0 1 0 0 0 0 0 0
2 3
n(n+1)
61 n 2 7
6 7
6 7
= 60 1 n 7,
6 7
4 5
0 0 1

n

n(n−1) n

n(n+1)
giving the result, since 2
= 2
and n + 2
= 2
.
134 Argue inductively,

A2 B = A(AB) = AB = B

A3 B = A(A2 B) = A(AB) = AB = B
..
.

Am B = AB = B.
Hence B = Am B = 0n B = 0n .
2 3
6a b7
136 Put A = 4 5 . Using 135, deduce by iteration that
c d

Ak = (a + d)k−1 A.
2 3
6a b 7
2
137 4 5 , bc = −a
c −a
2 3
6a b 7
2
138 ±I2 , 4 5 , a = 1 − bc
c −a
2 3
6a b7
139 We complete squares by putting Y = 4 5 = X − I. Then
c d
2 3 2 3 2 3
2
6 a + bc b(a + d)7
2 2 2 6−1 07 60 07
4 5 = Y = X − 2X + I = (X − I) = 4 5+I=4 5.
c(a + d) bc + d2 6 3 6 4

This entails a = 0, b = 0, cd = 6, d2 = 4. Using X = Y + I, we find that there are two solutions,

2 3 2 3
61 07 6 1 0 7
4 5, 4 5.
3 3 −3 −1
Answers and Hints 175

150 There are infinitely many solutions. Here is one:

2 3 2 3 2 3
61 2 37 6−9 2 37 610 0 07
6 7 6 7 6 7
6 7 6 7 6 7
A = 62 3 17 =6 17 +6 07 .
6 7 6 2 3 7 60 0 7
4 5 4 5 4 5
3 1 2 3 1 2 0 0 0

151 If such matrices existed, then by the first equation

tr (AC) + tr (DB) = n.

By the second equation and by Theorem 141,

0 = tr (CA) + tr (BD) = tr (AC) + tr (DB) = n,

a contradiction, since n ≥ 1.
2 3 2 3
61 17 63 07
T T
152 Disprove! This is not generally true. Take A = 4 5 and B = 4 5. Clearly A = A and B = B. We have
1 2 0 1

2 3
63 17
AB = 4 5
3 2

but 2 3
63 37
(AB)T = 4 5.
1 2

154 We have 02 32 31 02 31
2

B6a b7 6 a b7C B6a + bc ab + bd7C
tr A2 = tr 4 54 5A = tr 4
2 2
5A = a + d + 2bc
2
c d c d ca + cd d + cb

and
0 02 3112
B B6a b7CC
2
tr 4 5AA = (a + d) .
c d

Thus

tr A2 = (tr (A))2 ⇐⇒ a2 + d2 + 2bc = (a + d)2 ⇐⇒ bc = ad,
is the condition sought.

155



tr (A − I4 )2 = tr A2 − 2A + I4


= tr A2 − 2tr (A) + tr (I4 )

= −4 − 2tr (A) + 4

= −2tr (A) ,

and tr (3I4 ) = 12. Hence −2tr (A) = 12 or tr (A) = −6.

156 Disprove! Take A = B = In and n > 1. Then tr (AB) = n < n2 = tr (A) tr (B).
176 Appendix A

2 3 2 3 2 3
61 07 60 17 60 07
157 Disprove! Take A = 4 5, B = 4 5, C = 4 5. Then tr (ABC) = 1 but tr (BAC) = 0.
0 0 0 0 1 0

158 We have

(AAT )T = (AT )T AT = AAT .

159 We have

(AB − BA)T = (AB)T − (BA)T = BT AT − AT BT = −BA − A(−B) = AB − BA.

161 Let X = [xij ] and put XXT = [cij ]. Then

n
X
0 = cii = x2ik =⇒ xik = 0.
k=1

187 Here is one possible approach. If we perform C1 ↔ C3 on A we obtain

2 3 2 3
61 0 1 07 60 0 1 07
6 7 6 7
6 7 6 7
60 1 0 17 60 1 0 07
6 7 6 7
A1 = 6 7 so take P=6 7.
6 7 6 7
61 1 −1 17 61 0 0 07
6 7 6 7
4 5 4 5
1 −1 1 1 0 0 0 1

Now perform 2R1 → R1 on A1 to obtain

2 3 2 3
62 0 2 07 62 0 0 07
6 7 6 7
6 7 6 7
60 1 0 17 60 1 0 07
6 7 6 7
A2 = 6 7 so take D=6 7.
6 7 6 7
61 1 −1 17 60 0 1 07
6 7 6 7
4 5 4 5
1 −1 1 1 0 0 0 1

Finally, perform R1 + 2R4 → R1 on A2 to obtain

2 3 2 3
64 −2 4 27 61 0 0 27
6 7 6 7
6 7 6 7
60 1 0 17 60 1 0 07
6 7 6 7
B=6 7 so take T =6 7.
6 7 6 7
61 1 −1 17 60 0 1 07
6 7 6 7
4 5 4 5
1 −1 1 1 0 0 0 1
Answers and Hints 177

188 Here is one possible approach.

2 3 2 3
6a b c7 6g h i7
6 7 6 7
6 7 P:ρ 3 ↔ ρ 1 6 7
6d e f7 6d e f7
6 7 6 7
4 5 4 5
g h i a b c
2 3
6h g i7
6 7
P ′ :C 1↔ C2 6 7
6e d f7
6 7
4 5
b a c
2 3
6h − g g i7
6 7
T :C 1 −C 2 −→ C 1 6 7
6e − d d f7
6 7
4 5
b−a a c
2 3
6 h−g g i7
6 7
D:2ρ 3 −→ ρ 3 6 7
6 e−d d f7
6 7
4 5
2b − 2a 2a 2c

Thus we take
2 3 2 3
60 0 17 60 1 07
6 7 6 7
6 7 6 7
P = 60 1 07 , P ′ = 61 0 07 ,
6 7 6 7
4 5 4 5
1 0 0 0 0 1

2 3 2 3
6 1 0 07 61 0 07
6 7 6 7
6 7 6 7
T = 6−1 1 07 , D = 60 1 07 .
6 7 6 7
4 5 4 5
0 0 1 0 0 2

189 Let Eij ∈ Mn (F). Then

2 3
60 0 ... a1i ... 07
6 7
6 .. .. 7
60 0 . a2i . 07
6 7
6 7
6. .. .. .. .. .. 7
AEij = 6 .. . . . . .7 ,
6 7
6 7
6 .. .. 7
60 0 . an−1i . 07
6 7
4 .. .. 5
0 0 . ani . 0

where the entries appear on the j-column. Then we see that tr (AEij ) = aji and similarly, by considering BEij , we see that
tr (BEij ) = bji . Therefore ∀i, j, aji = bji , which implies that A = B.
178 Appendix A

190 Let Est ∈ Mn (R). Then 2 3

6 0 0 ... 0 7
6 7
6 . .. .. 7
6 .. . ... . 7
6 7
6 7
6 7
Eij A = 6aj1 aj2 ... ajn 7 ,
6 7
6 7
6 .. .. .. 7
6 . . ... . 7
6 7
4 5
0 0 ... 0

where the entries appear on the i-th row. Thus

2 3
6 0 0 ... 0 7
6 7
6 .. .. .. 7
6 . . ... . 7
6 7
6 7
2 6 7
(Eij A) = 6aji aj1 aji aj2 ... aji ajn 7 ,
6 7
6 7
6 .. .. .. 7
6 . . ... . 7
6 7
4 5
0 0 ... 0

which means that ∀i, j, aji ajk = 0. In particular, a2ji = 0, which means that ∀i, j, aji = 0, i.e., A = 0n .

216 a = 1, b = −2.

217 Claim: A−1 = In − A + A2 − A3 . For observe that

(In + A)(In − A + A2 − A3 ) = In − A + A2 − A3 − A + A2 − A3 + A4 = In ,

proving the claim.

218 Disprove! It is enough to take A = B = 2In . Then (A + B)−1 = (4In )−1 = 1

I
4 n
but A−1 + B−1 = 1
I
2 n
+ 1
I
2 n
= In .

223 We argue by contradiction. If exactly one of the matrices is not invertible, the identities

A = AIn = (ABC)(BC)−1 = 0n ,

B = In BIn = (A)−1 (ABC)C−1 = 0n ,

C = In C = (AB)−1 (ABC) = 0n ,
shew a contradiction depending on which of the matrices are invertible. If all the matrices are invertible then

0n = 0n C−1 B−1 A−1 = (ABC)C−1 B−1 A−1 = In ,

also gives a contradiction.

224 Observe that A, B, AB are invertible. Hence

A2 B2 = In = (AB)2 =⇒ AABB = ABAB

=⇒ AB = BA,

by cancelling A on the left and B on the right. One can also argue that A = A−1 , B = B−1 , and so AB = (AB)−1 =
B−1 A−1 = BA.

225 Observe that A = (a − b)In + bU, where U is the n × n matrix with 1F ’s everywhere. Prove that

A2 = (2(a − b) + nb)A − ((a − b)2 + nb(a − b))In .

226 Compute (A − In )(B − In ).

Answers and Hints 179

227 By Theorem 141 we have tr SAS−1 = tr S−1 SA = tr (A).

243 The rank is 2.

244 If B is invertible, then rank (AB) = rank (A) = rank (BA). Similarly, if A is invertible rank (AB) = rank (B) =
2 3 2 3 2 3
6 1 0 7 6 0 1 7 6 0 0 7
rank (BA). Now, take A = 4 5 and B = 4 5. Then AB = B, and so rank (AB) = 1. But BA = 4 5,
0 0 0 0 0 0
and so rank (BA) = 0.

245 Effecting R3 − R1 → R3 ; aR4 − bR2 → R4 successively, we obtain

2 3 2 3
61 a 1 b7 61 a 1 b7
6 7 6 7
6 7 6 7
6a 1 b 17 6a 1 b 1 7
6 7 6 7
6 7 6 7.
6 7 6 7
61 b 1 a7 60 b−a 0 a − b7
6 7 6 7
4 5 4 5
b 1 a 1 0 a−b a 2 − b2 a−b

Performing R2 − aR1 → R2 ; R4 + R3 → R4 we have

2 3
61 a 1 b 7
6 7
6 7
60 1 − a2 b−a 1 − ab 7
6 7
6 7.
6 7
60 b−a 0 a−b 7
6 7
4 5
0 0 a 2 − b2 2(a − b)

Performing (1 − a2 )R3 − (b − a)R2 → R3 we have

2 3
61 a 1 b 7
6 7
6 7
60 1 − a2 b−a 1 − ab 7
6 7
6 7.
6 7
60 0 −a2 + 2ab − b2 2a − 2b − a3 + ab2 7
6 7
4 5
0 0 a 2 − b2 2(a − b)

Performing R3 − R4 → R3 we have
2 3
61 a 1 b 7
6 7
6 7
60 1 − a2 b−a 1 − ab 7
6 7
6 7.
6 7
60 0 −2a(a − b) −a(a2 − b2 )7
6 7
4 5
0 0 a 2 − b2 2(a − b)

Performing 2aR4 + (a + b)R3 → R4 we have

2 3
61 a 1 b 7
6 7
6 7
60 1 − a2 b−a 1 − ab 7
6 7
6 7.
6 7
60 0 −2a(a − b) −a(a2 − b2 ) 7
6 7
4 5
0 0 0 4a2 − 4ab − a4 + a2 b2 − ba3 + ab3
180 Appendix A

Factorising, this is 2 3
61 a 1 b 7
6 7
6 7
60 1 − a2 b−a 1 − ab 7
6 7
=6 7.
6 7
60 0 −2a(a − b) −a(a − b)(a + b) 7
6 7
4 5
0 0 0 −a(a + 2 + b)(a − b)(a − 2 + b)
Thus the rank is 4 if (a + 2 + b)(a − b)(a − 2 + b) 6= 0. The rank is 3 if a + b = 2 and (a, b) 6= (1, 1) or if a + b = −2
and (a, b) 6= (−1, −1). The rank is 2 if a = b 6= 1 and a 6= −1. The rank is 1 if a = b = ±1.
246 rank (A) = 4 if m3 + m2 + 2 6= 0, and rank (A) = 3 otherwise.

247 The rank is 4 if a 6= ±b. The rank is 1 is a = ±b 6= 0. The rank is 0 if a = b = 0.

248 The rank is 4 if (a − b)(c − d) 6= 0. The rank is 2 is a = b, c 6= d or if a 6= b, c = d. The rank is 1 if a = b and c = d.

249 rank (ABC) ≤ 2 =⇒ x = 13.
2 3
61 i 7
252 For the counterexample consider A = 4 5.
i −1

261 We form the augmented matrix

2 3
61 2 3 1 0 07
6 7
6 7
62 3 1 0 1 07
6 7
4 5
3 1 2 0 0 1

From R2 − 2R1 → R2 and R3 − 3R1 → R3 we obtain

2 3
61 2 3 1 0 07
6 7
6 7
60 6 2 5 1 07 .
6 7
4 5
0 2 0 4 0 1

From R2 ↔ R3 we obtain 2 3
61 2 3 1 0 07
6 7
6 7
60 2 0 4 0 17 .
6 7
4 5
0 6 2 5 1 0

Now, from R1 − R2 → R1 and R3 − 3R2 → R3 , we obtain

2 3
61 0 3 4 0 67
6 7
6 7
60 2 0 4 0 17 .
6 7
4 5
0 0 2 0 1 4

From 4R2 → R2 and 4R3 → R3 , we obtain

2 3
61 0 3 4 0 67
6 7
6 7
60 1 0 2 0 47 .
6 7
4 5
0 0 1 0 4 2
Answers and Hints 181

Finally, from R1 − 3R3 → R3 we obtain

2 3
61 0 0 4 2 07
6 7
6 7
60 1 0 2 0 47 .
6 7
4 5
0 0 1 0 4 2

We deduce that
2 3−1 2 3
61 2 37 64 2 07
6 7 6 7
6 7 6 7
62 3 17 = 62 0 47 .
6 7 6 7
4 5 4 5
3 1 2 0 4 2

262 To find the inverse of B we consider the augmented matrix

2 3
6 0 0 −1 1 0 07
6 7
6 7
6 0 −1 a 0 1 07 .
6 7
4 5
−1 a b 0 0 1

Performing R1 ↔ R3 , −R3 → R3 , in succession,

2 3
6−1 a b 0 0 17
6 7
6 7
6 0 −1 a 0 1 07 .
6 7
4 5
0 0 1 −1 0 0

Performing R1 + aR2 → R1 and R2 − aR3 → R2 in succession,

2 3
2
6−1 0 b+a 0 a 17
6 7
6 7
6 0 −1 0 a 1 07 .
6 7
4 5
0 0 1 −1 0 0

Performing R1 − (b + a2 )R3 → R3 , −R1 → R1 and −R2 → R2 in succession, we find

2 3
61 0 0 −b − a2 −a −17
6 7
6 7
60 1 0 −a −1 0 7 ,
6 7
4 5
0 0 1 −1 0 0

whence
2 3
2
6−b − a −a −17
6 7
6 7
B−1 = 6 −a −1 0 7 .
6 7
4 5
−1 0 0
182 Appendix A

Now,
2 32 32 3
6 0 0 −17 6a b c7 6−b − a2 −a −17
6 76 76 7
6 76 76 7
BAB−1 = 6 0 −1 a 7 6 07 6 0 7
6 7 61 0 7 6 −a −1 7
4 54 54 5
−1 a b 0 1 0 −1 0 0
2 32 3
6 0 −1 0 7 6−b − a2 −a −17
6 76 7
6 76 7
= 6−1 0 7 6 0 7
6 a 7 6 −a −1 7
4 54 5
0 0 −c −1 0 0
2 3
6a 1 07
6 7
6 7
= 6b 0 17
6 7
4 5
c 0 0

= AT ,

which is what we wanted to prove.

263 Operating formally, and using elementary row operations, we find

2 3
2
a −1 a 2 +2a−2 a−2
6 − a 2 −5+2a a 2 −5+2a a 2 −5+2a 7
6 7
−1 6 7
B =6 − 2 2 a+4 1
− a 2 −5+2a 7.
6 a −5+2a a 2 −5+2a 7
4 5
2a
a 2 −5+2a
− a 22a+5
−5+2a
a
a 2 −5+2a

√
Thus B is invertible whenever a 6= −1 ± 6.

264 Form the augmented matrix

2 3
6a 2a 3a 1 0 07
6 7
6 7
60 b 2b 0 1 07 .
6 7
4 5
0 0 c 0 0 1

1 1 1
Perform R1 → R1 , R2 → R2 , R3 → R3 , in succession, obtaining
a b a
2 3
61 2 3 1/a 0 0 7
6 7
6 7
60 1 2 0 1/b 0 7 .
6 7
4 5
0 0 1 0 0 1/c

Now perform R1 − 2R2 → R1 and R2 − 2R3 → R2 in succession, to obtain

2 3
61 0 −1 1/a −2/a 0
7
6 7
6 7
60 1 0 0 1/b −2/c7 .
6 7
4 5
0 0 1 0 0 1/c
Answers and Hints 183

Finally, perform R1 + R3 → R1 to obtain

2 3
61 0 0 1/a −2/b 1/c 7
6 7
6 7
60 1 0 0 1/b −2/c7 .
6 7
4 5
0 0 1 0 0 1/c

Whence
2 3−1 2 3
6a 2a 3a7 61/a −2/b 1/c 7
6 7 6 7
6 7 6 7
60 b 2b7 =6 0 1/b −2/c7 .
6 7 6 7
4 5 4 5
0 0 c 0 0 1/c

265 To compute the inverse matrix we proceed formally as follows. The augmented matrix is
2 3
6b a 0 1 0 07
6 7
6 7
6c 0 a 0 1 07 .
6 7
4 5
0 c b 0 0 1

Performing bR2 − cR1 → R2 we find

2 3
6b a 0 1 0 07
6 7
6 7
60 −ca ab −c b 07 .
6 7
4 5
0 c b 0 0 1
Performing aR3 + R2 → R3 we obtain
2 3
6b a 0 1 0 07
6 7
6 7
60 −ca ab −c b 07 .
6 7
4 5
0 0 2ab −c b a
Performing 2R2 − R3 → R2 we obtain
2 3
6b a 0 1 0 0 7
6 7
6 7
60 −2ca 0 −c b −a7 .
6 7
4 5
0 0 2ab −c b a
Performing 2cR1 + R2 → R1 we obtain
2 3
62bc 0 0 c b −a7
6 7
6 7
6 0 −2ca 0 −c b −a7 .
6 7
4 5
0 0 2ab −c b a

From here we easily conclude that

2 3−1 2 3
1 1 a
6b a 07 6 2b 2c
− 2bc 7
6 7 6 7
6 7 6 7
6c 0 a7 =6 1 b
− 2ac 1 7
6 7 6 2a 2c 7
4 5 4 5
c 1 1
0 c b − 2ba 2a 2b
184 Appendix A

as long as abc 6= 0.

266 Form the expanded matrix

2 3
61 + a 1 1 1 0 07
6 7
6 7
6 1 1+b 1 0 1 07 .
6 7
4 5
1 1 1+c 0 0 1

Perform bcR1 → R1 , abR3 → R3 , caR2 → R2 . The matrix turns into

2 3
6bc + abc bc bc bc 0 0 7
6 7
6 7
6 ca ca + abc ca 0 ca 0 7 .
6 7
4 5
ab ab ab + abc 0 0 ab

Perform R1 + R2 + R3 → R1 the matrix turns into

2 3
6ab + bc + ca + abc ab + bc + ca + abc ab + bc + ca + abc bc ca ab7
6 7
6 7
6 ca ca + abc ca 0 ca 0 7 .
6 7
4 5
ab ab ab + abc 0 0 ab

1
Perform R
ab+bc+ca+abc 1
→ R1 . The matrix turns into
2 3
bc ca ab
61 1 1 ab+bc+ca+abc ab+bc+ca+abc ab+bc+ca+abc 7
6 7
6 7
6 ca ca + abc ca 0 ca 0 7.
6 7
4 5
ab ab ab + abc 0 0 ab

Perform R2 − caR1 → R2 and R3 − abR3 → R3 . We get

2 3
bc ca ab
61 1 1 ab+bc+ca+abc ab+bc+ca+abc ab+bc+ca+abc 7
6 7
6 7
60 abc 0 abc
− ab+bc+ca+abc
2
ca − c a 2 2 2
a bc
− ab+bc+ca+abc 7.
6 ab+bc+ca+abc 7
4 5
2 2
ab c a bc a2 b2
0 0 abc − ab+bc+ca+abc − ab+bc+ca+abc ab − ab+bc+ca+abc

1 1
Perform R
ABC 2
→ R2 and R
ABC 3
→ R3 . We obtain
2 3
bc ca ab
61 1 1 ab+bc+ca+abc ab+bc+ca+abc ab+bc+ca+abc 7
6 7
6 7
60 1 0 c
− ab+bc+ca+abc 1
− ca a
− ab+bc+ca+abc 7.
6 b b(ab+bc+ca+abc) 7
4 5
b a 1 ab
0 0 1 − ab+bc+ca+abc − ab+bc+ca+abc c
− c(ab+bc+ca+abc)

Finally we perform R1 − R2 − R3 → R1 , getting

2 3
a+b+bc c b
61 0 0 ab+bc+ca+abc
− ab+bc+ca+abc − ab+bc+ca+abc 7
6 7
6 7
60 1 0 c
− ab+bc+ca+abc 1
− ca a
− ab+bc+ca+abc 7.
6 b b(ab+bc+ca+abc) 7
4 5
b a 1 ab
0 0 1 − ab+bc+ca+abc − ab+bc+ca+abc c
− c(ab+bc+ca+abc)
Answers and Hints 185

We conclude that the inverse is

2 3
b+c+bc c b
6 ab+bc+ca+abc
− ab+bc+ca+abc − ab+bc+ca+abc 7
6 7
6 7
6− c c+a+ca a
− ab+bc+ca+abc 7
6 ab+bc+ca+abc ab+bc+ca+abc 7
4 5
b a a+b+ab
− ab+bc+ca+abc − ab+bc+ca+abc ab+bc+ca+abc

271 Since rank A2 < 5, A2 is not invertible. But then A is not invertible and hence rank (A) < 5.

282 The free variables are z and w. We have

2y + w = 2 =⇒ 2y = 2 − w =⇒ y = 1 + w,

and

x + y + z + w = 0 =⇒ x = −y − z − w = 2y + 2z + 2w.

Hence
2 3 2 3 2 3 2 3
6 x 7 607 607 607
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 y 7 617 607 607
6 7 6 7 6 7 6 7
6 7 = 6 7 + z6 7 + w6 7.
6 7 6 7 6 7 6 7
6 z 7 607 617 607
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
w 0 0 1

This gives the 9 solutions.

283 We have
2 32 3 2 3
61 2 37 6 x 7 657
6 76 7 6 7
6 76 7 6 7
62 3 17 6 7 = 6 7,
6 7 6y7 667
4 54 5 4 5
3 1 2 z 0

Hence
2 3 2 3−1 2 3 2 32 3 2 3
6 x 7 61 2 37 657 64 2 07 657 647
6 7 6 7 6 7 6 76 7 6 7
6 7 6 7 6 7 6 76 7 6 7
6y7 = 62 3 17 667 = 62 0 47 6 7 = 6 7.
6 7 6 7 6 7 6 7 667 637
4 5 4 5 4 5 4 54 5 4 5
z 3 1 2 0 0 4 2 0 3
186 Appendix A

291 Observe that the third row is the sum of the first two rows and the fourth row is twice the third. So we have
2 3 2 3
61 1 1 1 1 1 7 61 1 1 1 1 1 7
6 7 6 7
6 7 6 7
61 0 1 0 1 −17 61 0 1 0 1 −17
6 7 6 7
6 7 6 7
6 7 R 3 −R 1 −R 2 → R 3 6 7
62 1 2 1 2 0 7 60 0 0 0 0 0 7
6 7 R 4 −2R 1 −2R 2 → R 4 6 7
6 7 6 7
6 7 6 7
64 2 4 2 4 0 7 60 0 0 0 0 0 7
6 7 6 7
4 5 4 5
1 0 0 0 1 0 1 0 0 0 1 0
2 3
60 1 1 1 0 1 7
6 7
6 7
60 0 1 0 0 −17
6 7
6 7
R 2 −R 5 → R 2 6 7
60 0 0 0 0 0 7
R 1 −R 5 → R 1 6 7
6 7
6 7
60 0 0 0 0 0 7
6 7
4 5
1 0 0 0 1 0

Rearranging the rows we obtain 2 3

61 0 0 0 1 0 7
6 7
6 7
60 1 1 1 0 1 7
6 7
6 7
6 7
60 0 1 0 0 −17 .
6 7
6 7
6 7
60 0 0 0 0 0 7
6 7
4 5
0 0 0 0 0 0
Hence d and f are free variables. We obtain
c = −1,
b = 1 − c − d = 2 − d,
a = −f.
The solution is 2 3 2 3 2 3 2 3
6a 7 6 0 7 6 0 7 6−17
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6b7 6 2 7 6−17 6 0 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 c 7 = 6−17 + d 6 0 7 + f 6 0 7 .
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6d7 6 0 7 6 1 7 6 0 7
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
f 0 0 1
2 3
6 1 7
6 7
6 7
6 1 7
6 7
6 7
6 7
292 The unique solution is 6−17.
6 7
6 7
6 7
6 1 7
6 7
4 5
1
Answers and Hints 187

293 The augmented matrix of the system is

2 3
62m 1 1 2 7
6 7
6 7
6 1 2m 1 4m 7 .
6 7
4 5
1 1 2m 2m2

Performing R1 ↔ R2 .
2 3
6 1 2m 1 4m 7
6 7
6 7
62m 1 1 2 7 .
6 7
4 5
1 1 2m 2m2
Performing R2 ↔ R3 .
2 3
6 1 2m 1 4m 7
6 7
6 7
6 1 1 2m 2m2 7 .
6 7
4 5
2m 1 1 2
Performing R2 − R1 → R1 and R3 − 2mR1 → R3 we obtain
2 3
61 2m 1 4m 7
6 7
6 7
60 1 − 2m 2m − 1 2m − 4m7
2 .
6 7
4 5
0 1 − 4m2 1 − 2m 2 − 8m2
1
If m = the matrix becomes
2
2 3
61 1 1 2 7
6 7
6 7
60 0 0 − 32 7
6 7
4 5
0 0 0 0

and hence it does not have a solution. If m 6= 21 , by performing 1

R
1−2m 2
→ R2 and 1
R
1−2m 3
→ R3 , the matrix becomes
2 3
61 2m 1 4m 7
6 7
6 7
60 1 −1 2m(m−2) 7 .
6 1−2m 7
4 5
0 1 + 2m 1 2(1 + 2m)

Performing R3 − (1 + 2m)R2 → R3 we obtain

2 3
61 2m 1 4m 7
6 7
6 7
60 1 −1 2m(m−2) 7.
6 1−2m 7
4 5
2(1+2m)(1−m 2 )
0 0 2 + 2m 1−2m

If m = −1 then the matrix reduces to 2 3

61 −2 1 −47
6 7
6 7
60 1 −1 2 7 .
6 7
4 5
0 0 0 0
188 Appendix A

The solution in this case is 2 3 2 3

6x7 6 z 7
6 7 6 7
6 7 6 7
6y7 = 62 + z7 .
6 7 6 7
4 5 4 5
z z

If m 6= −1, m 6= − 21 we have the solutions

2 3 2 3
m−1
6x7 6 1−2m 7
6 7 6 7
6 7 6 7
6y7 = 6 1−3m 7.
6 7 6 1−2m 7
4 5 4 5
(1+2m)(1−m)
z 1−2m

294 By performing the elementary row operations, we obtain the following triangular form:

ax + y − 2z = 1,
2
(a − 1) y + (1 − a)(a − 2)z = 1 − a,
(a − 2)z = 0.
If a = 2, there is an infinity of solutions:
2 3 2 3
6 x 7 61 + t7
6 7 6 7
6 7 6 7
6y7 = 6 −1 7 t ∈ R.
6 7 6 7
4 5 4 5
z t
Assume a 6= 2. Then z = 0 and the system becomes

ax + y = 1,
(a − 1)2 y = 1 − a,
2x + (3 − a)y = 1.
We see that if a = 1, the system becomes
x + y = 1,
2x + 2y = 1,
and so there is no solution. If (a − 1)(a − 2) 6= 0, we obtain the unique solution
2 3 2 3
1
6 x 7 6 a−1 7
6 7 6 7
6 7 6 7
6y7 = 6− 1 7 .
6 7 6 a−1 7
4 5 4 5
z 0

2 3 2 3
1
6 x 7 6 m−2 7
6 7 6 7
6 7 6 7
295 The system is solvable if m 6= 0, m 6= ±2. If m 6= 2 there is the solution 6y7 = 6 m+3 7 .
6 7 6 m−2 7
4 5 4 5
m+2
z m−2

2 3 2 3
6x7 6 a + d + b − c 7
6 7 6 7
6 7 6 7
6y7 6−c − d − b + a7
6 7 6 7
296 There is the unique solution 6 7 = 6 7.
6 7 6 7
6z7 6 d + c − b + a 7
6 7 6 7
4 5 4 5
t c−d+b+a
Answers and Hints 189

297 The system can be written as

2 32 3 2 3
6b a 07 6x7 6c7
6 76 7 6 7
6 76 7 6 7
6c 0 a7 6 7 = 6 7.
6 7 6 y 7 6b 7
4 54 5 4 5
0 c b z a

The system will have the unique solution

2 3 2 3−1 2 3
6x7 6b a 07 6c7
6 7 6 7 6 7
6 7 6 7 6 7
6y7 = 6c 0 a7 6 b7
6 7 6 7 6 7
4 5 4 5 4 5
z 0 c b a
2 32 3
1 1 a
6 2b 2c
− 2bc 7 6c7
6 76 7
6 76 7 ,
= 6 1 b
− 2ac 1 7 6 b7
6 2a 2c 76 7
4 54 5
c 1 1
− 2ba 2a 2b
a
2 3
b2 + c 2 − a 2
6 7
6 2 2bc 7
6 a + c 2 − b2 7
= 6 7
6 2ac 7
4 2 5
a + b2 − c 2
2ab

as long as the inverse matrix exists, which is as long as abc 6= 0

298

x = 2−2 36

y = 2−3 312

z = 22 3−7 .

299 Denote the addition operations applied to the rows by a1 , a2 , a3 , a4 and the subtraction operations to the columns
by b1 , b2 , b3 , b4 . Comparing A and AT we obtain 7 equations in 8 unknowns. By inspecting the diagonal entries, and the
entries of the first row of A and AT , we deduce the following equations

a 1 = b1 ,

a 2 = b2 ,

a 3 = b3 ,

a 4 = b4 ,

a1 − b2 = 3,

a1 − b3 = 6,

a1 − b4 = 9.
This is a system of 7 equations in 8 unknowns. We may let a4 = 0 and thus obtain a1 = b1 = 9, a2 = b2 = 6, a3 = b3 = 3,
a4 = b4 = 0.
190 Appendix A

300 The augmented matrix of this system is

2 3
6−y 1 0 0 1 07
6 7
6 7
6 1 −y 1 0 0 07
6 7
6 7
6 7
6 0 1 −y 1 0 07 .
6 7
6 7
6 7
6 0 0 1 −y 1 07
6 7
4 5
1 0 0 1 −y 0

Permute the rows to obtain

2 3
6 1 0 0 1 −y 07
6 7
6 7
6 0 1 −y 1 0 07
6 7
6 7
6 7
6 0 0 1 −y 1 07 .
6 7
6 7
6 7
6 1 −y 1 0 0 07
6 7
4 5
−y 1 0 0 1 0
Performing R5 + yR1 → R5 and R4 − R1 → R4 we get
2 3
61 0 0 1 −y 07
6 7
6 7
60 1 −y 1 0 07
6 7
6 7
6 7
60 0 1 −y 1 07 .
6 7
6 7
6 7
60 −y 1 −1 y 07
6 7
4 5
0 1 0 y 1 − y2 0

Performing R5 − R2 → R5 and R4 + yR2 → R4 we get

2 3
61 0 0 1 −y 07
6 7
6 7
60 1 −y 1 0 07
6 7
6 7
6 7
60 0 1 −y 1 07 .
6 7
6 7
6 7
60 0 1 − y2 y−1 y 07
6 7
4 5
0 0 y y−1 1 − y2 0

Performing R5 − yR3 → R5 and R4 + (y2 − 1)R3 → R4 we get

2 3
61 0 0 1 −y 07
6 7
6 7
60 1 −y 1 0 07
6 7
6 7
6 7
60 0 1 −y 1 07 .
6 7
6 7
6 7
60 0 0 −y3 + 2y − 1 y2 + y − 1 07
6 7
4 5
0 0 0 y2 + y − 1 1 − y − y2 0
Answers and Hints 191

Performing R5 + R4 → R5 we get
2 3
61 0 0 1 −y 07
6 7
6 7
60 1 −y 1 0 07
6 7
6 7
6 7
60 0 1 −y 1 07 .
6 7
6 7
6 7
60 0 0 −y3 + 2y − 1 y2 + y − 1 07
6 7
4 5
0 0 0 −y3 + y2 + 3y − 2 0 0

Upon factoring, the matrix is equivalent to

2 3
61 0 0 1 −y 07
6 7
6 7
60 1 −y 1 0 07
6 7
6 7
6 7
60 0 1 −y 1 07 .
6 7
6 7
6 7
60 0 0 −(y − 1)(y2 + y − 1) y2 + y − 1 07
6 7
4 5
0 0 0 −(y − 2)(y2 + y − 1) 0 0

Thus (y − 2)(y2 + y − 1)x4 = 0. If y = 2 then the system reduces to

2 3
61 0 0 1 −2 07
6 7
6 7
60 1 −2 1 0 07
6 7
6 7
6 7
60 0 1 −2 1 07 .
6 7
6 7
6 7
60 0 0 −5 5 07
6 7
4 5
0 0 0 0 0 0

In this case x5 is free and by backwards substitution we obtain

2 3 2 3
6 x 1 7 6 t7
6 7 6 7
6 7 6 7
6 x 2 7 6 t7
6 7 6 7
6 7 6 7
6 7 6 7
6 x 3 7 = 6 t7 , t ∈ R.
6 7 6 7
6 7 6 7
6 7 6 7
6 x 4 7 6 t7
6 7 6 7
4 5 4 5
x5 t

If y2 + y − 1 = 0 then the system reduces to

2 3
61 0 0 1 −y 07
6 7
6 7
60 1 −y 1 0 07
6 7
6 7
6 7
60 0 1 −y 1 07 .
6 7
6 7
6 7
60 0 0 0 0 07
6 7
4 5
0 0 0 0 0 0
192 Appendix A

In this case x4 , x5 are free, and 2 3 2 3

6x1 7 6 yt − s 7
6 7 6 7
6 7 6 7
6x2 7 6y2 s − yt − s7
6 7 6 7
6 7 6 7
6 7 6 7
6x3 7 = 6 ys − t 7, (s, t) ∈ R2 .
6 7 6 7
6 7 6 7
6 7 6 7
6x4 7 6 s 7
6 7 6 7
4 5 4 5
x5 t

Since y2 s − s = (y2 + y − 1)s − ys, this last solution can be also written as
2 3 2 3
6x1 7 6 yt − s 7
6 7 6 7
6 7 6 7
6x2 7 6−ys − yt7
6 7 6 7
6 7 6 7
6 7 6 7
6x3 7 = 6 ys − t 7 , (s, t) ∈ R2 .
6 7 6 7
6 7 6 7
6 7 6 7
6x4 7 6 s 7
6 7 6 7
4 5 4 5
x5 t

Finally, if (y − 2)(y2 + y − 1) 6= 0, then x4 = 0, and we obtain

2 3 2 3
6x1 7 607
6 7 6 7
6 7 6 7
6x2 7 607
6 7 6 7
6 7 6 7
6 7 6 7
6x3 7 = 607 .
6 7 6 7
6 7 6 7
6 7 6 7
6x4 7 607
6 7 6 7
4 5 4 5
x5 0

√
323 2a2 − 2a + 1
p 1
324 ||λv|| = 1
2
=⇒ (λ)2 + (−λ)2 = 1
2
=⇒ 2λ2 = 1
4
=⇒ λ = ± √ .
8

325 0

326 a = ±1 or a = −8.
1 1
327 [A] 2(x + y) − 2
z, [B] x + y − 2
z, [C] −(x + y + z)

328 [A]. 0, [B]. 0, [C]. 0, [D]. 0, [E]. 2c(= 2d)

329 [F]. 0, [G]. b, [H]. 20, [I]. 0.

330 Let the skew quadrilateral be ABCD and let P, Q, R, S be the midpoints of [A, B], [B, C], [C, D], [D, A], respectively.
Put x = OX, where X ∈ {A, B, C, D, P, Q, R, S}. Using the Section Formula 4.4 we have

a+b b+c c+d d+a

p= , q= , r= , s= .
2 2 2 2
This gives
a−c a−c
p−q= , s−r= .
2 2
−
−→ −→
This means that QP = RS and so PQRS is a parallelogram since one pair of sides are equal and parallel.
Answers and Hints 193

−→ −→ −
− −→ −−→ −
−→ − →
− −−→ −→ − →
−
331 We have 2BC = BE + EC. By Chasles’ Rule AC = AE + EC, and BD = BE + ED. We deduce that
−−→ −−→ − −→ −
−→ −→ − −→ −−→ − →
−
AC + BD = AE + EC + BE + ED = AD + BC.
−−→ −−→
But since ABCD is a parallelogram, AD = BC. Hence
−−→ −−→ −−→ − →
− −
−→
AC + BD = AD + BC = 2BC.
−→ −→ →
− −→ − →
− →
− −→
−
332 We have IA = −3IB ⇐⇒ IA = −3(IA + AB) = −3IA − 3AB. Thus we deduce

−→ −→ −
−→ −→ −
−→
IA + 3IA = −3AB ⇐⇒ 4IA = −3AB
→
− −
−→
⇐⇒ 4AI = 3AB
→
− 3−−→
⇐⇒ AI = 4
AB.

Similarly
−→ →
− −→ →
−
JA = − 13 JB ⇐⇒ 3JA = −JB
−→ −→ − −→
⇐⇒ 3JA = −JA − AB
−→ −
−→
⇐⇒ 4JA = −AB
→
− 1−−→
⇐⇒ AJ = 4
AB

.
→
− 3−−→ →
− 1−−→
Thus we take I such that AI = 4
AB and J such that AJ = 4
AB.

Now
−
−−→ −−→ −−→ − → →
−
MA + 3MB = MI + IA + 3IB
−−→ − → →
−
= 4MI + IA + 3IB
−−→
= 4MI,
and
−
−−→ −−→ −−→ −→ −−→ − →
3MA + MB = 3MJ + 3JA + MJ + JB
−−→ −→ − →
= 4MJ + 3JA + JB
−−→
= 4MJ.

333 Let G, O and P denote vectors from an arbitrary origin to the gallows, oak, and pine, respectively. The conditions of
the problem define X and Y, thought of similarly as vectors from the origin, by X = O + R(O − G), Y = P + R(P − G),
where R is the 90◦ rotation to the right, a linear transformation on vectors in the plane; the fact that −R is 90◦ leftward
rotation has been used in writing Y . Anyway, then
X+Y O+P R(O − P)
= +
2 2 2
is independent of the position of the gallows. This gives a simple algorithm for treasure-finding: take P as the (hitherto)
O + R(O)
arbitrary origin, then the treasure is at .
2
1
352 a = 2

354 2 3 2 3 2 3
647 6−17 627
p = 4 5 = 2 4 5 + 3 4 5 = 2r + 3s.
5 1 1
194 Appendix A

355 Since a1 = a•i, a2 = a•j, we may write

a = (a•i)i + (a•j)j
from where the assertion follows.

356
αa + βb = 0 =⇒ a•(αa + βb) = a•0

=⇒ α(a•a) = 0

=⇒ α||a||2 = 0.

Since a 6= 0, we must have ||a|| 6= 0 and thus α = 0. But if α = 0 then

αa + βb = 0 =⇒ βb = 0

=⇒ β = 0,

since b 6= 0.

357 We must shew that

(2x + 3y)•(2x − 3y) = 0.
But
9
(2x + 3y)•(2x − 3y) = 4||x||2 − 9||y||2 = 4( ||y||2 ) − 9||y||2 = 0.
4
358 We have ∀v ∈ R2 , v•(a − b) = 0. In particular, choosing v = a − b, we gather

(a − b)•(a − b) = ||a − b||2 = 0.

But the norm of a vector is 0 if and only if the vector is the 0 vector. Therefore a − b = 0, i.e., a = b.

359 We have
||a ± b||2 = (a ± b)•(a ± b)

= a•a ± 2a•b + b•b

= ||a||2 ± 2a•b + ||b||2 ,

whence the result follows.

360 We have
||u + v||2 − ||u − v||2 = (u + v)•(u + v) − (u − v)•(u − v)

= u•u + 2u•v + v•v − (u•u − 2u•v + v•v)

= 4u•v,
giving the result.

361 By definition

projax
projax •a
proj = a
||a||2
a
a•x
||x|| 2
x•a
= 2
a
||a||
(a•x)2
= a,
||x||2 ||a||2
(a•x)2
Since 0 ≤ ≤ 1 by the CBS Inequality, the result follows.
||x||2 ||a||2
Answers and Hints 195

362 Clearly, if a = 0 and λ 6= 0 then there are no solutions. If both a = 0 and λ = 0, then the solution set is the whole space
R2 . So assume that a 6= 0. By Theorem 347, we may write x = u + v with projxa = u||a and v ⊥ a. Thus there are infinitely
many solutions, each of the form
x •a λ
x=u+v = a+v = a + v,
||a||2 ||a||2
where v ∈ a⊥ .
2 3 2 3
6 2 7 6 1 7
372 Since a = 4 5 is normal to 2x − y = 1 and b = 4 5 is normal to x − 3y = 1, the desired angle can be obtained by
−1 −3
finding the angle between the normal vectors:

\ a• b 5 1 π
(a, b) = arccos = arccos √ √ = arccos √ = .
||a||||b|| 5 · 10 2 4

373 2(x − 1) + (y + 1) = 0 or 2x + y = 1.
−
−−→ −−→ − → −−→ −
−− −→ − →
−− −
−−→ −−→ − →
−−
374 By Chasles’ Rule AA ′ = AG + GA ′ , BB ′ = BG + GB ′ , and CC ′ = CG + GC ′ . Thus

−
−−→ −−→ − →
−−
0 = AA ′ + BB ′ + CC ′
−−→ − → −
−− −→ − → −−→ −
−− →
−−
= AG + GA ′ + BG + GB ′ + CG + GC ′
−−→ −−→ −−→ −
−−→ − → −
−− →
−−
= −(GA + GB + GC) + (GA ′ + GB ′ + GC ′ )
−
−−→ − → −
−− →
−−
= GA ′ + GB ′ + GC ′ ,

whence the result.

375 We have:
−→ −
−→ −→ −
−→
➊ The points F, A, D are collinear, and so FA is parallel to FD, meaning that there is k ∈ R \ {0} such that FA = kFD.
→
− →
− →
− −→
Since the lines (AB) and (DC) are parallel, we obtain through Thales’ Theorem that FI = kFJ and FB = kFC. This
gives
−→ → − −→ →
− − −→ −→
FA − FI = k(FD − FJ) =⇒ IA = kJD.
Similarly
−→ → − −→ → − −→ →
−
FB − FI = k(FC − FJ) =⇒ IB = kJC.
−→ − → −→ −→ −→ −→
Since I is the midpoint of [A, B], IA + IB = 0, and thus k(JC + JD) = 0. Since k 6= 0, we have JC + JD = 0, meaning
that J is the midpoint of [C, D]. Therefore the midpoints of [A, B] and [C, D] are aligned with F.
➋ Let J ′ be the intersection of the lines (EI) and (DC). Let us prove that J ′ = J.

−
−→ −
−→
Since the points E, A, C are collinear, there is l 6= 0 such that EA = lEC. Since the lines (ab) and (DC) are parallel,
→
− −→
− −→ −
−→
we obtain via Thales’ Theorem that EI = lEJ ′ and EB = lED. These equalities give
−→ − → → − −→ → −−→
EA − EI = l(EC − EJ ′ ) =⇒ IA = lJ ′ C,
− −
− −

−→ − → → − −→ → −−→
EB − EI = l(ED − EJ ′ ) =⇒ IB = lJ ′ D.
−
− −

−→ − → −−→ −−→ −−→ −−→

Since I is the midpoint of [A, B], IA + IB = 0, and thus l(J ′ C + J ′ D) = 0. Since l =
6 0, we deduce J ′ C + J ′ D = 0,
′ ′
that is, J is the midpoint of [C, D], and so J = J.

376 We have:
➊ By Chasles’ Rule
−
−→ 1−−→ −
−→ −→ 1−−→ ,
AE = 4
AC ⇐⇒ AB + BE = 4
AC

and
−→ 3−−→ −−→ −→
− 3−−→ .
AF = 4
AC ⇐⇒ AD + DF = 4
AC
196 Appendix A

−
−→ −−→
Adding, and observing that since ABCD is a parallelogram, AB = CD,

−
−→ −→ −−→ − → −−→
− −→ − → −−→ −
− −→ −−→
AB + BE + AD + DF = AC ⇐⇒ BE + DF = AC − AB − AD
−→ − → −−→ −−→ − → −−→
BE + DF = AD + DC − AB − AD .
− −
⇐⇒
−→ −
−→
⇐⇒ BE = −DF.

The last equality shews that the lines (BE) and (DF) are parallel.
−→ −
−→ −−→ − → −→
➋ Observe that BJ = 21 BC = 12 AD = AI = −IA . Hence
→
− − → − −→ − → − −→
IJ = IA + AB + BJ = AB,
proving that the lines (AB) and (IJ) are parallel.

Observe that
→ −
− → − → 1 −−→
− 1 −−→ 1 −
−→ −→ − → −→ −→ − → → −
IE = IA + AE = DA + AC = CB + FC = CJ + FC = FC + CJ = FJ,
2 4 2
whence IEJF is a parallelogram.
→
− 1− →
377 Since IE = 3
ID and [I, D] is a median of △ABD, E is the centre of gravity of △ABD. Let M be the midpoint of
−
−−→ −−→ −−→
[B, D], and observe that M is the centre of the parallelogram, and so 2AM = AB + AD. Thus
−→ 2−
− −−→ 1 − →
−− 1 −−→ −−→
AE = AM = (2AM) = (AB + AD).
3 3 3
−
−→ −−→
To shew that A, C, E are collinear it is enough to notice that AE = 31 AC.

||[A, B]|| λ
378 Suppose A, B, C are collinear and that = . Then by the Section Formula 4.4,
||[B, C]|| µ
λc + µa
b= ,
λ+µ
whence µa − (λ + µ)b + λc = 0 and clearly µ − (λ + µ) + λ = 0. Thus we may take α = µ, β = λ + µ, and γ = λ. Conversely,

suppose that
αa + βb + γc = 0, α+β+γ =0
for some real numbers α, β, γ, not all zero. Assume without loss of generality that γ 6= 0. Otherwise we simply change the
roles of γ, and α and β Then γ = −(α + β) 6= 0. Hence
αa + βb
αa + βb = (α + β)c =⇒ c = ,
α+β
β
and thus [O, C] divides [A, B] into the ratio , and therefore, A, B, C are collinear.
α
−−→
379 Put OX = x for X ∈ {A, A ′ , B, B ′ , C, C ′ , L, M, N, V}. Using problem 378 we deduce
v + αa + α ′ a ′ = 0, 1 + α + α ′ = 0, (A.1)
′ ′ ′
v + βa + β a = 0, 1 + β + β = 0, (A.2)
v + γa + γ ′ a ′ = 0, 1 + γ + γ ′ = 0. (A.3)
From A.2, A.3, and the Section Formula 4.4 we find
βb − γc β ′b′ − γ ′c′
= = l,
β−γ β′ − γ′
whence (β − γ)l = βb − γc. In a similar fashion, we deduce
(γ − α)m = γc − αa,
(α − β)n = αa − βb.
This gives
(β − γ)l + (γ − α)m + (α − β)n = 0,
(β − γ) + (γ − α) + (α − β) = 0,
and appealing to problem 378 once again, we deduce that L, M, N are collinear.
Answers and Hints 197

−
−→ −
−→
391 [A] AS, [B] AB.

392 Put 2 3 2 3 2 3
617 617 6−17
6 7 6 7 6 7
6 7 6 7 6 7
a = 617 × 617 = (i + j + k)×(i + j) = j − i = 6 1 7 .
6 7 6 7 6 7
4 5 4 5 4 5
1 0 0
Then either 2 3
3
6− √ 2 7
6 7
3a 3a 6 7
= √ = 6 √3 7 ,
||a|| 2 6 2 7
4 5
0
or 2 3
3
√
6 7 2
6 7
3a 6 3 7
− = 6− √ 7
||a|| 6 27
4 5
0
will satisfy the requirements.

393 The desired area is 2 3 2 3 2 3

   
 0   
6 7 6 1 7 6−17
  6 7 6 7 6 7 √
−→ −→  6 7 6 7 6 7
−
PQ×PR = 6 1 7 × 6 0 7 = 6−17 = 3.
6 7 6 7 6 7
4 5 4 5 4 5
   
 −1 −1   −1 

394 It is not associative, since i×(i×j) = i×k = −j but (i×i)×j = 0×j = 0.

395 We have x×x = −x×x by letting y = x in 4.15. Thus 2x×x = 0 and hence x×x = 0.

396 2a×b

397
a×(x×b) = b×(x×a) ⇐⇒ (a•b)x − (a•x)b = (b•a)x − (b•x)a ⇐⇒ a•x = b•x = 0.
The answer is thus {x : x ∈ Ra×b}.

398
(a•b)a + 6b + 2a×c
x=
12 + 2||a||2
(a•c)a + 6c + 3a×b
y=
18 + 3||a||2
2π
399 Assume contrariwise that a, b, c are three unit vectors in R3 such that the angle between any two of them is > . Then
3
1 1 1
a•b < − , b•c < − , and c•a < − . Thus
2 2 2

||a + b + c||2 = ||a||2 + ||b||2 + ||c||2

+2a•b + 2b•c + 2c•a

< 1+1+1−1−1−1

= 0,

which is impossible, since a norm of vectors is always ≥ 0.

198 Appendix A

410 The vectors 2 3 2 3

6a − (−a)7 6 2a 7
6 7 6 7
6 7 6 7
6 0 − 1 7 = 6−17
6 7 6 7
4 5 4 5
a−0 a

and 2 3 2 3
60 − (−a)7 6 a 7
6 7 6 7
6 7 6 7
6 1−1 7=6 0 7
6 7 6 7
4 5 4 5
2a − 0 2a

are coplanar. A vector normal to the plane is

2 3 2 3 2 3
6 2a 7 6 a 7 6 −2a 7
6 7 6 7 6 7
6 7 6 7 6 7
6−17 × 6 0 7 = 6−3a2 7 .
6 7 6 7 6 7
4 5 4 5 4 5
a 2a a

The equation of the plane is thus given by

2 3 2 3
6 −2a 7 6x − a7
6 7 6 7
6 7 6 7
6−3a2 7 • 6y − 07 = 0,
6 7 6 7
4 5 4 5
a z−a

that is,
2ax + 3a2 y − az = a2 .

411 The vectorial form of the equation of the line is

2 3 2 3
617 6 1 7
6 7 6 7
6 7 6 7
r = 607 + t 6−27 .
6 7 6 7
4 5 4 5
1 −1

2 3 2 3
6 1 7 6 1 7
6 7 6 7
6 7 6 7
Since the line follows the direction of 6−27, this means that 6−27 is normal to the plane, and thus the equation of the
6 7 6 7
4 5 4 5
−1 −1
desired plane is
(x − 1) − 2(y − 1) − (z − 1) = 0.

412 Observe that (0, 0, 0) (as 0 = 2(0) = 3(0)) is on the line, and hence on the plane. Thus the vector
2 3 2 3
6 1−0 7 6 1 7
6 7 6 7
6 7 6 7
6−1 − 07 = 6−17
6 7 6 7
4 5 4 5
−1 − 0 −1
Answers and Hints 199

lies on the plane. Now, if x = 2y = 3z = t, then x = t, y = t/2, z = t/3. Hence, the vectorial form of the equation of the line
is 2 3 2 3 2 3
607 6 1 7 6 1 7
6 7 6 7 6 7
6 7 6 7 6 7
r = 607 + t 61/27 = t 61/27 .
6 7 6 7 6 7
4 5 4 5 4 5
0 1/3 1/3
2 3
6 1 7
6 7
6 7
This means that 61/27 also lies on the plane, and thus
6 7
4 5
1/3

2 3 2 3 2 3
6 1 7 6 1 7 6 1/6 7
6 7 6 7 6 7
6 7 6 7 6 7
6−17 × 61/27 = 6−4/37
6 7 6 7 6 7
4 5 4 5 4 5
−1 1/3 3/2

is normal to the plane. The desired equation is thus

1 4 3
x − y + z = 0.
6 3 2

413 Put ax = by = cz = t, so x = t/a; y = t/b; z = t/c. The parametric equation of the line is
2 3 2 3
6x7 6 1/a 7
6 7 6 7
6 7 6 7
6y7 = t 61/b7 , t ∈ R.
6 7 6 7
4 5 4 5
z 1/c

2 3
6 1/a 7
6 7
6 7
Thus the vector 61/b7 is perpendicular to the plane. Therefore, the equation of the plane is
6 7
4 5
1/c

2 3 2 3 2 3
6 1/a 7 6 x − 1 7 607
6 7 6 7 6 7
6 7 6 7 6 7
61/b7 • 6y − 17 = 607 ,
6 7 6 7 6 7
4 5 4 5 4 5
1/c z−1 0

or
x y z 1 1 1
+ + = + + .
a b c a b c
We may also write this as
bcx + cay + abz = ab + bc + ca.
2 3
6a7
6 7
6 7
414 A vector normal to the plane is 6a2 7. The line sought has the same direction as this vector, thus the equation of the
6 7
4 5
a2
200 Appendix A

line is 2 3 2 3 2 3
6 x 7 607 6a7
6 7 6 7 6 7
6 7 6 7 6 7
6y7 = 607 + t 6a2 7 , t ∈ R.
6 7 6 7 6 7
4 5 4 5 4 5
z 1 a2

415 We have
x − z − y = 1 =⇒ −1 − y = 1 =⇒ y = −2.
Hence if z = t,
2 3 2 3 2 3 2 3
6 x 7 6t − 17 6−17 61 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6y7 = 6 −2 7 = 6−27 + t 607 .
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
z t 0 1

416 The vector 2 3 2 3

6 2 − 1 7 61 7
6 7 6 7
6 7 6 7
6 1 − 0 7 = 61 7
6 7 6 7
4 5 4 5
1 − (−1) 2

lies on the plane. The vector

2 3 2 3 2 3
617 617 6 1 7
6 7 6 7 6 7
6 7 6 7 6 7
607 × 617 = 6 1 7
6 7 6 7 6 7
4 5 4 5 4 5
1 2 −1

is normal to the plane. Hence the equation of the plane is

2 3 2 3
6 1 7 6x − 17
6 7 6 7
6 7•6 7
6 1 7 6 y 7 = 0 =⇒ x + y − z = 2.
6 7 6 7
4 5 4 5
−1 z+1

417 We have c×a = −i + 2j and a×b = 2k − 3i. By Theorem 408, we have

b×c = −a×b − c×a = −2k + 3i + i − 2j = 4i − 2j − 2k.

418 4x + 6y = 1

419 There are 7 vertices (V0 = (0, 0, 0), V1 = (11, 0, 0), V2 = (0, 9, 0), V3 = (0, 0, 8), V4 = (0, 3, 8), V5 = (9, 0, 2),
V6 = (4, 7, 0)) and 11 edges (V0 V1 , V0 V2 , V0 V3 , V1 V5 , V1 V6 , V2 V4 , V3 V4 , V3 V5 , V4 V5 , and V4 V6 ).
P 2
427 Expand  ni=1 ai
 = 0.

Pn
428 Observe that k=1 1 = n. Then we have
!2 ! ! !
n
X n
X   2 n
X Xn
2 1 1
n = 1 = (ak ) ≤ a2k 2
,
ak a k
k=1 k=1 k=1 k=1

giving the result.

Answers and Hints 201

429 This follows at once from the CBS Inequality by putting

2 3 2 3
a1
6 1 7 617
6 7 6 7
6a 7 6 7
6 27 627
6 2 7 6 7
v = 6 7, u = 6 7
6 7 6 7
6. . .7 6. . .7
6 7 6 7
4 5 4 5
an
n
n

and noticing that

n
X n(n + 1)(2n + 1)
k2 = .
6
k=1

442 No, since 1F v = v is not fulfilled. For example

2 3 2 3 2 3
1
6 7 6 1 · 1 7 617
1·4 5=4 5 6= 4 5 .
1 0 1

443 We expand (1F + 1F )(a + b) in two ways, first using 5.7 first and then 5.8, obtaining

(1F + 1F )(a + b) = (1F + 1F )a + (1F + 1F )b = a + a + b + b,

and then using 5.8 first and then 5.7, obtaining

(1F + 1F )(a + b) = 1F (a + b) + 1F (a + b) = a + b + a + b.

We thus have the equality

a + a + b + b = a + b + a + b.
Cancelling a from the left and b from the right, we obtain

a + b = b + a,

which is what we wanted to shew.

444 We must prove that each of the axioms of a vector space are satisfied. Clearly if (x, y, α) ∈ R+ × R+ × R then
x ⊕ y = xy > 0 and α ⊗ x = xα > 0, so V is closed under vector addition and scalar multiplication. Commutativity and
associativity of vector addition are obvious.

Let A be additive identity. Then we need

x ⊕ A = x =⇒ xA = x =⇒ A = 1.

Thus the additive identity is 1. Suppose I is the additive inverse of x. Then

1
x ⊕ I = 1 =⇒ xI = 1 =⇒ I = .
x
1
Hence the additive inverse of x is .
x

Now
α ⊗ (x ⊕ y) = (xy)α = xα yα = xα ⊕ yα = (α ⊗ x) ⊕ (α ⊗ y),
and
(α + β) ⊗ x = xα+β = xα xβ = (xα ) ⊕ (xβ ) = (α ⊗ x) ⊕ (β ⊗ x),
whence the distributive laws hold.

Finally,
1 ⊗ x = x1 = x,
and
α ⊗ (β ⊗ x) = (β ⊗ x)α = (xβ )α = xαβ = (αβ) ⊗ x,
and the last two axioms also hold.
202 Appendix A

445 C is a vector space over R, the proof is trivial. But R is not a vector space over C, since, for example taking i as a scalar
(from C) and 1 as a vector (from R) the scalar multiple i · 1 = i 6∈ R and so there is no closure under scalar multiplication.

446 One example is

 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3

 
607 607 607 607 617 617 617 617
 

 
 6 7 6 7 6 7 6 7 6 7 6 7 6 7 6 7
 
3 6 7 6 7 6 7 6 7 6 7 6 7 6 7 6 7
(Z2 ) = 607 , 607 , 617 , 617 , 607 , 607 , 617 , 617 .

 6 7 6 7 6 7 6 7 6 7 6 7 6 7 6 7 
 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5

 


 0 
1 0 1 0 1 0 1 

Addition is the natural element-wise addition and scalar multiplication is ordinary element-wise scalar multiplication.

447 One example is

2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3

 

 0 
2 6 7 607 607 617 617 617 627 627 627
(Z3 ) = 4 5 , 4 5 , 4 5 , 4 5 , 4 5 , 4 5 , 4 5 , 4 5 , 4 5 .

 

 0 1 2 0 1 2 0 1 2 

Addition is the natural element-wise addition and scalar multiplication is ordinary element-wise scalar multiplication.

454 Take α ∈ R and

2 3 2 3
′
6a 7 6a 7
6 7 6 7
6 7 6 ′7
6 b7 6b 7
6 7 6 7
x = 6 7 ∈ X, a − b − 3d = 0, y = 6 7 ∈ X, a ′ − b ′ − 3d ′ = 0.
6 7 6 ′7
6c7 6c 7
6 7 6 7
4 5 4 5
d d′

Then
2 3 2 3 2 3
′ ′
6a 7 6a 7 6a + αa 7
6 7 6 7 6 7
6 7 6 ′7 6 7
6b 7 6b 7 6b + αb ′ 7
6 7 6 7 6 7
x + αy = 6 7 + α 6 7 = 6 7.
6 7 6 ′7 6 7
6c7 6 c 7 6 c + αc ′ 7
6 7 6 7 6 7
4 5 4 5 4 5
d d′ d + αd ′

Observe that

(a + αa ′ ) − (b + αb ′ ) − 3(d + αd ′ ) = (a − b − 3d) + α(a ′ − b ′ − 3d ′ ) = 0 + α0 = 0,

meaning that x + αy ∈ X, and so X is a vector subspace of R4 .

455 Take
2 3 2 3
6 a1 7 6 a2 7
6 7 6 7
6 7 6 7
62a1 − 3b1 7 62a2 − 3b2 7
6 7 6 7
6 7 6 7
6 7 6 7
u=6 5b 7 , v = 6 5b 7 , α ∈ R.
6 1
7 6 2
7
6 7 6 7
6 7 6 7
6 a1 + 2b1 7 6 a2 + 2b2 7
6 7 6 7
4 5 4 5
a1 a2
Answers and Hints 203

Put s = a1 + αa2 , t = b1 + αb2 . Then

2 3 2 3
6 a1 + αa2 7 6 s 7
6 7 6 7
6 7 6 7
62(a1 + αa2 ) − 3(b1 + αb2 )7 62s − 3t7
6 7 6 7
6 7 6 7
6 7 6 7
u + αv = 6 5(b + αb ) 7 = 6 5t 7 ∈ X,
6 1 2
7 6 7
6 7 6 7
6 7 6 7
6 (a1 + αa2 ) + 2(b1 + αb2 ) 7 6 s + 2t 7
6 7 6 7
4 5 4 5
a1 + αa2 s

since this last matrix has the basic shape of matrices in X. This shews that X is a vector subspace of R5 .

456 Take (u, v) ∈ X2 and α ∈ R. Then

a•(u + αv) = a•u + αa•v = 0 + 0 = 0,

proving that X is a vector subspace of Rn .

457 Take (u, v) ∈ X2 and α ∈ R. Then

a×(u + αv) = a×u + αa×v = 0 + α0 = 0,

proving that X is a vector subspace of Rn .

462 We shew that some of the properties in the definition of vector subspace fail to hold in these sets.
2 3 2 3
607 607
6 7 6 7
6 7 6 7
➊ Take x = 617 , α = 2. Then x ∈ V but 2x = 627 6∈ V as 02 + 22 = 4 6= 1. So V is not closed under scalar
6 7 6 7
4 5 4 5
0 0
multiplication.
2 3 2 3 2 3
607 617 61 7
6 7 6 7 6 7
6 7 6 7 6 7
➋ Take x = 617 , y = 607. Then x ∈ W, y ∈ W but x + y = 617 6∈ W as 1 · 1 = 1 6= 0. Hence W is not closed under
6 7 6 7 6 7
4 5 4 5 4 5
0 0 0
vector addition.
2 3 2 3 2 3
6−1 17 6−1 17 61 −17
2
➌ Take x = 4 5 . Then x ∈ Z but −x = −4 5=4 5 6∈ Z as 1 + (−1) = 2 6= 0. So Z is not closed
0 0 0 0 0 0
under scalar multiplication.

463 Assume U1 * U2 and U2 * U1 . Take v ∈ U2 \ U1 (which is possible because U2 * U1 ) and u ∈ U1 \ U2 (which is

possible because U1 * U2 ). If u + v ∈ U1 , then—as −u is also in U1 —the sum of two vectors in U1 must also be in U1
giving
u + v − u = v ∈ U1 ,
a contradiction. Similarly if u + v ∈ U2 , then—as −v also in U2 —the sum of two vectors in U2 must also be in U1 giving

u + v − u = u ∈ U2 ,

another contradiction. Hence either U1 ⊆ U2 or U2 ⊆ U1 (or possibly both).

S S S
464 Assume contrariwiseSthat S
V = U1 U2 · · · Uk is the shortest such list. Since the UT j are proper subspaces, k
T> 1.
Choose x ∈ U1 , x 6∈ U2 · · · Uk and choose y 6∈ U1 . Put L = {y + αx|α ∈ F}. Claim: L U1 = ∅. For if u ∈ L U1
then ∃a0 ∈ F with u = y + a0 x and so y = u − a0 x ∈ U1 , a contradiction. So L and U1 are disjoint.
204 Appendix A

We now shew that L has at most one vector in common with Uj , 2 ≤ j ≤ k. For, if there were two elements of F, a 6= b
with y + ax, y + bx ∈ Uj , j ≥ 2 then

(a − b)x = (y + ax) − (y + bx) ∈ Uj ,

contrary to the choice of x.

Conclusion: since F is infinite, L is infinite. But we have shewn that L can have at most one element in common with the
Uj . This means that there are not enough Uj to go around to cover the whole of L. So V cannot be a finite union of proper
subspaces.

465 Take F = Z2 , V = F × F. Then V has the four elements

2 3 2 3 2 3 2 3
607 607 617 617
4 5,4 5,4 5,4 5,
0 1 0 1

with the following subspaces

 2 3 2 3  2 3 2 3 2 3 2 3

 
 
 
 
 

 0   0   0 
6 7 607 6 7 617 6 7 61 7
V1 = 4 5 , 4 5 , V2 = 4 5 , 4 5 , V3 = 4 5 , 4 5 .

 
 
 
 
 

 0 1   0 0   0 1 

It is easy to verify that these subspaces satisfy the conditions of the problem.

475 If 2 3 2 3 2 3
617 617 617
6 7 6 7 6 7
6 7 6 7 6 7
a 607 + b 617 + c 617 = 0,
6 7 6 7 6 7
4 5 4 5 4 5
0 0 1

then 2 3 2 3
6a + b + c7 607
6 7 6 7
6 7 6 7
6 b + c 7 = 607 .
6 7 6 7
4 5 4 5
c 0

This clearly entails that c = b = a = 0, and so the family is free.

476 Assume 2 3 2 3 2 3 2 3 2 3
617 6 1 7 6 1 7 61 7 60 7
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
617 6 1 7 6−17 61 7 60 7
6 7 6 7 6 7 6 7 6 7
a6 7 +b6 7 + c6 7 + d6 7 = 6 7.
6 7 6 7 6 7 6 7 6 7
617 6−17 6 1 7 60 7 60 7
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
1 −1 −1 1 0

Then
a + b + c + d = 0,
a + b − c + d = 0,
a − b + c = 0,
a − b − c + d = 0.
Subtracting the second equation from the first, we deduce 2c = 0, that is, c = 0. Subtracting the third equation from the
fourth, we deduce −2c + d = 0 or d = 0. From the first and third equations, we then deduce a + b = 0 and a − b = 0, which
entails a = b = 0. In conclusion, a = b = c = d = 0.
Answers and Hints 205

Now, put
2 3 2 3 2 3 2 3 2 3
617 6 1 7 6 1 7 617 617
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
617 6 1 7 6−17 617 627
6 7 6 7 6 7 6 7 6 7
x6 7 +y6 7 +z6 7 + w6 7 = 6 7.
6 7 6 7 6 7 6 7 6 7
617 6−17 6 1 7 607 617
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
1 −1 −1 1 1

Then

x + y + z + w = 1,
x + y − z + w = 2,
x − y + z = 1,
x − y − z + w = 1.
Solving as before, we find
2 3 2 3 2 3 2 3 2 3
617 6 1 7 6 1 7 617 617
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
617 6 1 7 6 7 6 7 6 7
6 7 16 7 1 6−17 617 627
26 7 + 6 7 − 6 7 − 6 7 = 6 7.
6 7 26 7 26 7 6 7 6 7
617 6−17 6 1 7 607 617
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
1 −1 −1 1 1

477 Since a, b are linearly independent, none of them is 0. Assume that there are (α, β, γ) ∈ R3 such that

αa + βb + γa×b = 0. (A.4)

Since a•(a×b) = 0, taking the dot product of A.4 with a yields α||a||2 = 0, which means that α = 0, since ||a|| 6= 0. Similarly,
we take the dot product with b and a×b obtaining respectively, β = 0 and γ = 0. This establishes linear independence.

478 Assume that

λ1 a1 + · · · + λk ak = 0.
Taking the dot product with aj and using the fact that ai •aj = 0 for i 6= j we obtain

0 = 0•aj = λj aj •aj = λj ||aj ||2 .

Since aj 6= 0 =⇒ ||aj ||2 6= 0, we must have λj = 0. Thus the only linear combination giving the zero vector is the trivial
linear combination, which proves that the vectors are linearly independent.

481 We have
(v1 + v2 ) − (v2 + v3 ) + (v3 + v4 ) − (v4 + v1 ) = 0,
a non-trivial linear combination of these vectors equalling the zero-vector.
√ √
483 Yes. Suppose that a + b 2 = 0 is a non-trivial linear combination of 1 and 2 with rational numbers a and b. If one
of a, b is different from 0 then so is the other. Hence
√ √ b
a + b 2 = 0 =⇒ 2=− .
a
√b
The sinistral side of the equality is irrational whereas the dextral side is rational, a contradiction.
2=−
a
√ √ √
484 No. The representation 2 · 1 + (− 2) 2 = 0 is a non-trivial linear combination of 1 and 2.

485 1. Assume that √ √

a + b 2 + c 3 = 0, a, b, c, ∈ Q, a2 + b2 + c2 6= 0.
If ac 6= 0, then
√ √ √ 2b2 − a2 − 3c2 √
b 2 = −a − c 3 ⇔ 2b2 = a2 + 2ac 3 + 3c2 ⇔ = 3.
2ac
206 Appendix A

The dextral side of the last implication is irrational, whereas the sinistral side is rational. Thus it must be the case that
ac = 0. If a = 0, c 6= 0 then r
√ √ b 3
b 2+c 3 =0⇔ − = ,
c 2
and again the dextral side is irrational and the sinistral side is rational. Thus if a = 0 then also c = 0. We can similarly
prove that c = 0 entails a = 0. Thus we have √
b 2 = 0,
which means that b = 0. Therefore
√ √
a + b 2 + c 3 = 0, a, b, c, ∈ Q, ⇔ a = b = c = 0.
√ √
This proves that {1, 2, 3} are linearly independent over Q.
2. Rationalising denominators,
√ √
1 2 1+ 2 2 12 + 4
√ +√ = +
1− 2 12 − 2 1−2 12 − 4
√ 1√ 1
= −1 − 2 + 3+
2 2
1 √ 1√
= − − 2+ 3.
2 2
486 Assume that
aex + be2x + ce3x = 0.
Then
c = −ae−2x − be−x .
Letting x → +∞, we obtain c = 0. Thus
aex + be2x = 0,
and so
b = −ae−x .
Again, letting x → +∞, we obtain b = 0. This yields
aex = 0.
Since the exponential function never vanishes, we deduce that a = 0. Thus a = b = c = 0 and the family is linearly
independent over R.

487 This follows at once from the identity

cos 2x = cos2 x − sin2 x,
which implies
cos 2x − cos2 x + sin2 x = 0.

500 Given an arbitrary polynomial

p(x) = a + bx + cx2 + dx3 ,
we must shew that there are real numbers s, t, u, v such that
p(x) = s + t(1 + x) + u(1 + x)2 + v(1 + x)3 .
In order to do this we find the Taylor expansion of p around x = −1. Letting x = −1 in this last equality,
s = p(−1) = a − b + c − d ∈ R.
Now,
p ′ (x) = b + 2cx + 3dx2 = t + 2u(1 + x) + 3v(1 + x)2 .
Letting x = −1 we find
t = p ′ (−1) = b − 2c + 3d ∈ R.
Again,
p ′′ (x) = 2c + 6dx = 2u + 6v(1 + x).
Letting x = −1 we find
u = p ′′ (−1) = c − 3d ∈ R.
Finally,
p ′′′ (x) = 6d = 6v,
so we let v = d ∈ R. In other words, we have
p(x) = a + bx + cx2 + dx3 = (a − b + c − d) + (b − 2c + 3d)(1 + x) + (c − 3d)(1 + x)2 + d(1 + x)3 .
Answers and Hints 207

501 Assume contrariwise that 2 3 2 3 2 3

6 1 7 6 1 7 6 0 7
6 7 6 7 6 7
6 7 6 7 6 7
6 1 7 = a6 0 7 + b6 1 7.
6 7 6 7 6 7
4 5 4 5 4 5
−1 −1 −1

Then we must have

a = 1,
b = 1,
−a − b = −1,
2 3 2 3 2 3 02 3 2 31
6 1 7 6 1 7 6 0 7 B6 1 7 6 0 7C
6 7 6 7 6 7 B6 7 6 7C
6 7 6 7 6 7 B6 7 6 7C
which is impossible. Thus 6 1 7 is not a linear combination of 6 0 7 , 6 1 7 and hence is not in span B6 0 7 , 6 1 7C.
6 7 6 7 6 7 B6 7 6 7C
4 5 4 5 4 5 4 5 4 5A
−1 −1 −1 −1 −1

502 It is 2 3 2 3 2 3 2 3
61 07 60 07 6 0 17 6 a c7
a4 5+b4 5 + c4 5=4 5,
0 0 0 1 −1 0 −c b

i.e., this family spans the set of all skew-symmetric 2 × 2 matrices over R.

515 We have 2 3 2 3 2 3
6 a 7 617 6 0 7
6 7 6 7 6 7
6 7 6 7 6 7
62a − 3b7 627 6−37
6 7 6 7 6 7
6 7 6 7 6 7
6 7 6 7 6 7
6 5b 7 = a 607 + b 6 5 7 ,
6 7 6 7 6 7
6 7 6 7 6 7
6 7 6 7 6 7
6 a + 2b 7 617 6 2 7
6 7 6 7 6 7
4 5 4 5 4 5
a 1 0

so clearly the family 2 3 2 3


 


 61 7 6 7
0 
 

 6 7 6 7 


 6 7 6 7 


 62 7 6−37

 6 7 6 7 
6 7
 6 7 

6 7 6 7
60 7 , 6 5 7

 6 7 6 7

 6 7 6 7 


 6 7 6 7 
 61 7
 6 2 7



 6 7 6 7 


 4 5 4 5 



 1 
0 

spans the subspace. To shew that this is a linearly independent family, assume that
2 3 2 3 2 3
617 6 0 7 607
6 7 6 7 6 7
6 7 6 7 6 7
627 6−37 607
6 7 6 7 6 7
6 7 6 7 6 7
6 7 6 7 6 7
a 607 + b 6 5 7 = 607 .
6 7 6 7 6 7
6 7 6 7 6 7
6 7 6 7 6 7
617 6 2 7 607
6 7 6 7 6 7
4 5 4 5 4 5
1 0 0
208 Appendix A

Then it follows clearly that a = b = 0, and so this is a linearly independent family. Conclusion:
2 3 2 3

 


 617 6 0 7 
 
 6 7
6 7 6 7
 6 7

 


 627 6−37 


 6 7 6 7

6 7 6 7
 
6 7 6 7
60 7 , 6 5 7
6


7 6 7
6 7 6 7 
 

 6 7 6 7 


 617 6 2 7 


 6 7 6 7
4 5 4 5
 


 

 1 0 

is a basis for the subspace.

516 Suppose

0 = a(v1 + v2 ) + b(v2 + v3 ) + c(v3 + v4 ) + d(v4 + v5 ) + f(v5 + v1 )

= (a + f)v1 + (a + b)v2 + (b + c)v3 + (c + d)v4 + (d + f)v5 .

Since {v1 , v2 , . . . , v5 } are linearly independent, we have

a + f = 0,

a+b =0
b+c =0
c+d =0
d + f = 0.
Solving we find a = b = c = d = f = 0, which means that the

{v1 + v2 , v2 + v3 , v3 + v4 , v4 + v5 , v5 + v1 }

are linearly independent. Since the dimension of V is 5, and we have 5 linearly independent vectors, they must also be a basis
for V.

517 The matrix of coefficients is already in echelon form. The dimension of the solution space is n − 1 and the following
vectors in R2n form a basis for the solution space
2 3 2 3
2 3
6−17 6−17
6 7 6 7
6 7 6 7 6−17
6 1 7 6 0 7 6 7
6 7 6 7 6 7
6 7 6 7 6 0 7
6 7 6 7 6 7
6 0 7 6 1 7 6 7
6 7 6 7 6 7
6 . 7 6 . 7 6. . .7
6 . 7 6 . 7 6 7
6 . 7 6 . 7 6 7
6 7 6 7 6 7
6 7 6 7 6 1 7
6 7 6 7 6 7
6 0 7 6 0 7 6 7
a1 = 6 7 , a2 = 6 7 , . . . , an−1 = 6
6 7 6 7
6−17 .
7
6−17 6−17 6 7
6 7 6 7 6 7
6 7 6 7 6 0 7
6 7 6 7 6 7
6 1 7 6 0 7 6 7
6 7 6 7 6 . 7
6 7 6 7 6 .. 7
6 7 6 7 6 7
6 0 7 6 1 7 6 7
6 7 6 7 6 7
6 7 6 7 6 0 7
6 .. 7 6 .. 7 6 7
6 . 7 6 . 7 4 5
6 7 6 7
4 5 4 5 1
0 0

(The “second” −1 occurs on the n-th position. The 1’s migrate from the 2nd and n + 1-th position on a1 to the n − 1-th
and 2n-th position on an−1 .)
Answers and Hints 209

518 Take (u, v) ∈ X2 and α ∈ R. Then

2 3 2 3
′
6a 7 6a 7
6 7 6 7
6 7 6 ′7
6b7 6b 7
6 7 6 7
u = 6 7 , b + 2c = 0, v = 6 7 , b ′ + 2c ′ = 0.
6 7 6 ′7
6c7 6c 7
6 7 6 7
4 5 4 5
d d′

We have
2 3
′
6a + αa 7
6 7
6 7
6b + αb ′ 7
6 7
u + αv = 6 7,
6 7
6 c + αc ′ 7
6 7
4 5
′
d + αd

and to demonstrate that u + αv ∈ X we need to shew that (b + αb ′ ) + 2(c + αc ′ ) = 0. But this is easy, as

(b + αb ′ ) + 2(c + αc ′ ) = (b + 2c) + α(b ′ + 2c ′ ) = 0 + α0 = 0.

Now
2 3 2 3 2 3 2 3 2 3
6a7 6 a 7 617 6 0 7 607
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
6b7 6−2c7 607 6−27 607
6 7 6 7 6 7 6 7 6 7
6 7=6 7 = a6 7 +c6 7 + d6 7
6 7 6 7 6 7 6 7 6 7
6c7 6 c 7 607 6 1 7 607
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
d d 0 0 1

It is clear that
2 3 2 3 2 3
617 6 0 7 607
6 7 6 7 6 7
6 7 6 7 6 7
607 6−27 607
6 7 6 7 6 7
6 7,6 7,6 7
6 7 6 7 6 7
607 6 1 7 607
6 7 6 7 6 7
4 5 4 5 4 5
0 0 1

are linearly independent and span X. They thus constitute a basis for X.

n(n + 1)
519 As a basis we may take the matrices Eij ∈ Mn (F) for 1 ≤ i ≤ j ≤ n.
2

520 dim X = 2, as basis one may take {v1 , v2 }.

521 dim X = 3, as basis one may take {v1 , v2 , v3 }.

522 dim X = 3, as basis one may take {v1 , v2 , v3 }.

523 Since a ⊥ a×x = b, there are no solutions if a•b 6= 0. Neither are there solutions if a = 0 and b 6= 0. If both a = b = 0,
then the solution set is the whole of R3 . Assume thus that a•b = 0 and that a and b are linearly independent. Then a, b, a×b
are linearly independent, and so they constitute a basis for R3 . Any x ∈ R3 can be written in the form

x = αa + βb + γa×b.
210 Appendix A

We then have
b = a×x

= βa×b + γa×(a×b)

= βa×b + γ((a•b)a − (a•a)b).

= βa×b − γ(a•ab)

= βa×b − γ||a||2 b,
from where
βa×b + (−γ||a||2 − 1)b = 0,
1
which means that β = 0 and γ = − ||a|| 2 , since a, b, a×b are linearly independent. Thus

1
x = αa − a×b
||a||2
in this last case.

531 1. It is enough to prove that the matrix

2 3
61 1 1 1 7
6 7
6 7
61 1 −1 −17
6 7
A=6 7
6 7
61 −1 1 −17
6 7
4 5
1 −1 −1 1

is invertible. But an easy computation shews that

2 32
61 1 1 1 7
6 7
6 7
61 1 −1 −17
2 6 7
A =6 7 = 4I4 ,
6 7
61 −1 1 −17
6 7
4 5
1 −1 −1 1

whence the inverse sought is

2 3 2 3
61 1 1 1 7 61/4 1/4 1/4 1/4 7
6 7 6 7
6 7 6 7
6 −1 −17 6 −1/4 −1/47
1 1 61 1 7 61/4 1/4 7
A−1 = A= 6 7=6 7.
4 46 7 6 7
61 −1 1 −17 61/4 −1/4 1/4 −1/47
6 7 6 7
4 5 4 5
1 −1 −1 1 1/4 −1/4 −1/4 1/4

2. Since the ak are four linearly independent vectors in R4 and dim R4 = 4, they form a basis for R4 . Now, we want to
solve 2 3 2 32 3 2 3
6 x 7 61 1 1 1 7 6 x 7 61 7
6 7 6 76 7 6 7
6 7 6 76 7 6 7
6 y 7 61 1 −1 −17 6 7 6 7
6 7 6 7 6 y 7 62 7
A6 7 = 6 76 7 = 6 7
6 7 6 76 7 6 7
6 z 7 61 −1 1 −17 6 z 7 617
6 7 6 76 7 6 7
4 5 4 54 5 4 5
w 1 −1 −1 1 w 1
Answers and Hints 211

and so
2 3 2 3 2 32 3 2 3
6x7 617 61/4 1/4 1/4 1/4 7 617 6 5/4 7
6 7 6 7 6 76 7 6 7
6 7 6 7 6 76 7 6 7
6y7 627 61/4 1/4 −1/4 −1/47 627 6 1/4 7
7 6 7 6
6 7 −1 6 7 6 7
6 7=A 6 7=6 76 7 = 6 7.
6 7 6 7 6 76 7 6 7
6z7 617 61/4 −1/4 1/4 −1/4 7 6 1 7 6 −1/47
6 7 6 7 6 76 7 6 7
4 5 4 5 4 54 5 4 5
w 1 1/4 −1/4 −1/4 1/4 1 −1/4

It follows that
2 3 2 3 2 3 2 3 2 3
617 617 6 1 7 6 1 7 6 1 7
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
627 617 6 1 7 6−17 6 7
6 7 56 7 16 7 16 7 1 6−17
6 7 = 6 7 + 6 7 − 6 7 − 6 7.
6 7 46 7 46 7 46 7 46 7
617 617 6−17 6 1 7 6−17
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
1 1 −1 −1 1

The coordinates sought are

 
5 1 1 1
, ,− ,− .
4 4 4 4

3. Since we have
2 3 2 3 2 3 2 3 2 3
617 617 6 1 7 6 1 7 6 1 7
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
627 617 6−17 6 1 7 6 7
6 7 56 7 16 7 16 7 1 6−17
6 7 = 6 7 − 6 7 + 6 7 − 6 7,
6 7 46 7 46 7 46 7 46 7
617 617 6 1 7 6−17 6−17
6 7 6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5 4 5
1 1 −1 −1 1

the coordinates sought are

 
5 1 1 1
,− , ,− .
4 4 4 4

2 3
1 1
6 a−1 0 0 −
6 a − 177
6 7
6 −1 1−a −1 a+1 7
6 7
532 [1] a = 1, [2] (A(a))−1 =6 7 [3]
6 1 a 7
6− −1 0 7
6 a−1 a−1 7
4 5
1 a 1 −a − 1

2 3
1 1 1
60 a−1 a−1 a−1 7
6 7
6 7
60 −a − 1 −a −1 7
6 7
6 7
6 a a 1 7
60 −− − − 7
6 a−1 a−1 a − 17
4 5
1 2+a a+1 1
212 Appendix A

539 Let α ∈ R. Then

2 3 2 3
6 x + αa 7 6(x + αa) − (y + αb) − (z + αc)7
6 7 6 7
6 7 6 7
L 6y + αb7 = 6(x + αa) + (y + αb) + (z + αc)7
6 7 6 7
4 5 4 5
z + αc z + αc
2 3 2 3
6x − y − z7 6a − b − c 7
6 7 6 7
6 7 6 7
= 6x + y + z7 + α 6a + b + c7
6 7 6 7
4 5 4 5
z c
2 3 2 3
6x 7 6a 7
6 7 6 7
6 7 6 7
= L 6y7 + αL 6b7 ,
6 7 6 7
4 5 4 5
z c

proving that L is a linear transformation.

540 Let x, y, x ′ , y ′ be vectors in R3 and let α ∈ R be a scalar. Then

L((x, y) + α(x ′ , y ′ )) = L(x + αx ′ , y + αy ′ )

= (x + αx ′ )×k + h×(y + αy ′ )

= x×k + αx ′ ×k + h×y + h×αy ′

= L(x, y) + αL(x ′ , y ′ )

541

L(H + αH ′ ) = −A−1 (H + αH ′ )A−1

= −A−1 HA−1 + α(−A−1 H ′ A−1 )

= L(H) + αL(H ′ ),

proving that L is linear.

542 Let S be convex and let a, b ∈ T (S). We must prove that ∀α ∈ [0; 1], (1 − α)a + αb ∈ T (S). But since a, b belong
to T (S), ∃x ∈ S, y ∈ S with T (x) = a, T (y) = b. Since S is convex, (1 − α)x + αy ∈ S. Thus

T ((1 − α)x + αy) ∈ T (S),

which means that

(1 − α)T (x) + αT (y) ∈ T (S),

that is,
(1 − α)a + αb ∈ T (S),

as we wished to show.
Answers and Hints 213

2 3
6x7
6 7
6 7
550 Assume 6y7 ∈ ker (L). Then
6 7
4 5
z

2 3 2 3
6 x 7 60 7
6 7 6 7
6 7 6 7
L 6y 7 = 60 7 ,
6 7 6 7
4 5 4 5
z 0

that is

x − y − z = 0,

x + y + z = 0,

z = 0.

This implies that x − y = 0 and x + y = 0, and so x = y = z = 0. This means that

 2 3

 
607
 

 
 6 7
 
6 7
ker (L) = 607 ,

 6 7
 4 5

 


 0  

and L is injective.

By the Dimension Theorem 546, dim Im (L) = dim V − dim ker (L) = 3 − 0 = 3, which means that

Im (L) = R3

and L is surjective.

2 3
6a 7
6 7
6 7
551 Assume that 6b7 ∈ ker (T ),
6 7
4 5
c

2 3 2 3 2 3 2 3
6a7 617 617 607
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6b7 = (a − b) 607 + b 617 + c 607 .
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
c 0 0 1
214 Appendix A

Then
2 3
2 3
607
6 7
6 7 6a 7
607 6 7
6 7 6 7
6 7 = T 6b 7
6 7 6 7
607 4 5
6 7
4 5 c
0
2 3 2 3 2 3
617 617 607
6 7 6 7 6 7
6 7 6 7 6 7
= (a − b)T 607 + bT 617 + cT 607
6 7 6 7 6 7
4 5 4 5 4 5
0 0 1
2 3 2 3 2 3
6 1 7 6 2 7 6 1 7
6 7 6 7 6 7
6 7 6 7 6 7
6 0 7 6−17 6−17
6 7 6 7 6 7
= (a − b) 6 7 + b 6 7 + c 6 7
6 7 6 7 6 7
6−17 6 0 7 6 1 7
6 7 6 7 6 7
4 5 4 5 4 5
0 0 0
2 3
6 a+b+c 7
6 7
6 7
6 −b − c 7
6 7
= 6 7.
6 7
6−a + b + c7
6 7
4 5
0

It follows that a = 0 and b = −c. Thus

 2 3 

 


 6 0 7 

 
 6 7
 

6 7
ker (T ) = c 6−17 : c ∈ R ,

 6 7 



 4 5 



 

1

and so dim ker (T ) = 1.

By the Dimension Theorem 546,

dim Im (T ) = dim V − dim ker (T ) = 3 − 1 = 2.

We readily see that

2 3 2 3 2 3
6 2 7 6 1 7 6 1 7
6 7 6 7 6 7
6 7 6 7 6 7
6−17 6 0 7 6−17
6 7 6 7 6 7
6 7 = 6 7 +6 7,
6 7 6 7 6 7
6 0 7 6−17 6 1 7
6 7 6 7 6 7
4 5 4 5 4 5
0 0 0
Answers and Hints 215

and so 02 3 2 31
B6 1 7 6 1 7C
B6 7 6 7C
B6 7 6 7C
B6 0 7 6−17C
B6 7 6 7C
Im (T ) = span B6 7 , 6 7C .
B6 7 6 7C
B6−17 6 1 7C
B6 7 6 7C
4 5 4 5A
0 0

552 Assume that 2 3 2 3

2 3
6x + 2y7 607
x 6 7 6 7
6 7 6 7 6 7
L 4 5 = 6x + 2y7 = 607 .
6 7 6 7
y 4 5 4 5
0 0

Then x = −2y and so

2 3 2 3
6x7 6−27
4 5 = y4 5.
y 1

This means that dim ker (L) = 1 and ker (L) is the line through the origin and (−2, 1). Observe that L is not injective.

2 3
6a 7
6 7
6 7
By the Dimension Theorem 546, dim Im (L) = dim V − dim ker (L) = 2 − 1 = 1. Assume that 6b7 ∈ Im (L). Then
6 7
4 5
c
∃(x, y) ∈ R2 such that
2 3 2 3
2 3
6x + 2y7 6a7
x 6 7 6 7
6 7 6 7 6 7
L 4 5 = 6x + 2y7 = 6b7 .
6 7 6 7
y 4 5 4 5
0 c

This means that 2 3 2 3 2 3

6a7 6x + 2y7 617
6 7 6 7 6 7
6 7 6 7 6 7
6b7 = 6x + 2y7 = (x + 2y) 617 .
6 7 6 7 6 7
4 5 4 5 4 5
c 0 0

Observe that L is not surjective.

553 Assume that 2 3 2 3

2 3
6x − y7 607
x 6 7 6 7
6 7 6 7 6 7
L 4 5 = 6x + y7 = 607 .
6 7 6 7
y 4 5 4 5
0 0

Then x + y = 0 = x − y, that is, x = y = 0, meaning that

 2 3

 
 0  
6 7
ker (L) = 4 5 ,

 
 0  
216 Appendix A

and so L is injective.
2 3
6a 7
6 7
6 7
By the Dimension Theorem 546, dim Im (L) = dim V − dim ker (L) = 2 − 0 = 2. Assume that 6b7 ∈ Im (L). Then
6 7
4 5
c
∃(x, y) ∈ R2 such that
2 3 2 3
2 3 x − y
6 7 6a 7
x 6 7 6 7
6 7 6 7 6 7
L 4 5 = 6x + y7 = 6b7 .
6 7 6 7
y 4 5 4 5
0 c

This means that 2 3 2 3 2 3 2 3

6 a 7 6x − y 7 617 6−17
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
6b7 = 6x + y7 = x 617 + y 6 1 7 .
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
c 0 0 0

Since 2 3 2 3
617 6−17
6 7 6 7
6 7 6 7
61 7 , 6 1 7
6 7 6 7
4 5 4 5
0 0

are linearly independent, they span a subspace of dimension 2 in R3 , that is, a plane containing the origin. Observe that L
is not surjective.

554 Assume that 2 3

2 3 2 3
6x7
6 7
6 7 6x − y − z 7 60 7
L 6y7 = 4 5 = 4 5.
6 7
4 5 y − 2z 0
z
 2 3 

 


 3
6 7 

 
 6 7
 

6 7
Then y = 2z; x = y + z = 3z. This means that ker (L) = z 627 : z ∈ R . Hence dim ker (L) = 1, and so L is not

 6 7 



 4 5 



 

1
injective.

Now, if
2 3
2 3 2 3
6x7
6 7
6 7 6 x − y − z7 6 a 7
L 6y7 = 4 5 = 4 5.
6 7
4 5 y − 2z b
z

Then 2 3 2 3 2 3 2 3 2 3
a
6 7 6 x − y − z 7 1
6 7 −1
6 7 6−17
4 5=4 5 = x4 5 +y4 5 + z4 5.
b y − 2z 0 1 −2
Answers and Hints 217

Now,
2 3 2 3 2 3
617 6−17 6−17
−3 4 5 − 2 4 5 = 4 5
0 1 −2

and
2 3 2 3
617 6−17
4 5,4 5
0 1

are linearly independent. Since dim Im (L) = 2, we have Im (L) = R2 , and so L is surjective.

555 Assume that

02 31
B6a b 7C
0 = tr 4 5A = a + d.
c d

Then a = −d and so,

2 3 2 3 2 3 2 3 2 3
6a b7 6−d b7 6−1 07 60 17 60 07
4 5=4 5 = d4 5+b4 5 + c4 5,
c d c d 0 1 0 0 1 0

and so dim ker (L) = 3. Thus L is not injective. L is surjective, however. For if α ∈ R, then
02 31
B6α 0 7C
α = tr 4 5A .
0 0

556 1. Let (A, B)2 ∈ M2 (R), α ∈ R. Then

L(A + αB) = (A + αB)T + (A + αB)

= AT + BT + A + αB

= AT + A + αBT + αB

= L(A) + αL(B),

proving that L is linear.

2 3
6a b7
2. Assume that A = 4 5 ∈ ker (L). Then
c d

2 3 2 3 2 3 2 3
60 07 6a b 7 6a c 7 6 2a b + c7
4 5 = L(A) = 4 5+4 5=4 5,
0 0 c d b d b+c 2d

whence a = d = 0 and b = −c. Hence

02 31
B60 −17C
ker (L) = span 4 5A ,
1 0

and so dim ker (L) = 1.

218 Appendix A

3. By the Dimension Theorem, dim Im (L) = 4 − 1 = 3. As above,

2 3
6 2a b + c7
L(A) = 4 5
b+c 2d
2 3 2 3 2 3
62 07 60 17 60 07
= a4 5 + (b + c) 4 5+d4 5,
0 0 1 0 0 2

from where 02 3 2 3 2 31
B62 07 60 17 60 0 7C
Im (L) = span 4 5,4 5,4 5A .
0 0 1 0 0 2

557 ➊ Observe that

(I − T )2 = I − 2T + T 2 = I − 2T + T = I − T,
proving the result.
1
➋ The inverse is I − 2
T, for

1 1 1 1 1
(I + T )(I − T ) = I + T − T − T 2 = I + T − T − T = I,
2 2 2 2 2
proving the claim.
➌ We have
x ∈ ker (T ) ⇐⇒ x − T (x) ∈ ker (T )

⇐⇒ I(x) − T (x) ∈ ker (T )

⇐⇒ (I − T )(x) ∈ ker (T )

⇐⇒ x ∈ Im (I − T ) .

562 1. Since the image of T is the plane x + y + z = 0, we must have

a + 0 + 1 = 0 =⇒ a = −1,

3 + b − 5 = 0 =⇒ b = 2,
−1 + 2 + c = 0 =⇒ c = −1.
2 3
617
6 7
6 7
2. Observe that 617 ∈ ker (T ) and so
6 7
4 5
1
2 3 2 3
617 607
6 7 6 7
6 7 6 7
T 617 = 607 .
6 7 6 7
4 5 4 5
1 0

Thus 2 3 2 3 2 3 2 3
617 627 61 7 6 3 7
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
T 607 = T 617 − T 61 7 = 6 2 7 ,
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
0 1 1 −5
Answers and Hints 219

2 3 2 3 2 3 2 3
607 617 617 6−17
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
T 617 = T 627 − T 61 7 = 6 2 7 ,
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
0 1 1 −1
2 3 2 3 2 3 2 3
607 617 617 6−17
6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7
T 607 = T 617 − T 61 7 = 6 0 7 .
6 7 6 7 6 7 6 7
4 5 4 5 4 5 4 5
1 2 1 1

The required matrix is therefore

2 3
6 3 −1 −17
6 7
6 7
6 2 2 0 7 .
6 7
4 5
−5 −1 1

563 1. Let α ∈ R. We have

02 3 2 31 02 31
x
B6 7 6 u 7C B6x + αu7C
T 4 5 + α 4 5A = T 4 5A
y v y + αv
2 3
6 x + αu + y + αv 7
6 7
6 7
= 6 x + αu − y − αv 7
6 7
4 5
2(x + αu) + 3(y + αv)
2 3 2 3
6 x+y 7 6 u+v 7
6 7 6 7
6 7 6 7
= 6 x−y 7+ α6 u −v 7
6 7 6 7
4 5 4 5
2x + 3y 2u + 3v
02 31 02 31
x
B6 7C B6u7C
= T 4 5A + αT 4 5A ,
y v

proving that T is linear.

2. We have 2 3 2 3 2 3
02 31
607 6 x + y 7 607
x 6 7 6 7 6 7
B6 7C 6 7 6 7 6 7
T 4 5A = 607 ⇐⇒ 6 x − y 7 = 607 ⇐⇒ x = y = 0,
6 7 6 7 6 7
y 4 5 4 5 4 5
0 2x + 3y 0

dim ker (T ) = 0, and whence T is injective.

3. By the Dimension Theorem, dim Im (T ) = 2 − 0 = 2. Now, since
2 3 2 3 2 3
02 31 x + y 1
6 7 6 7 6 1 7
x 6 7 6 7 6 7
B6 7C 6 7 6 7 6 7
T 4 5A = 6 x − y 7 = x 617 + y 6−17 ,
6 7 6 7 6 7
y 4 5 4 5 4 5
2x + 3y 2 3
220 Appendix A

whence
02 3 2 31
B617 6 1 7C
B6 7 6 7C
B6 7 6 7C
Im (T ) = span B617 , 6−17C .
B6 7 6 7C
4 5 4 5A
2 3

4. We have
2 3 2 3 2 3 2 3 2 3
02 31
6 3 7 617 6 1 7 607 6 11/2 7
1 6 7 6 7 6 7 6 7 6 7
B6 7C 6 7 11 6 7 5 6 7 13 6 7 6 7
T 4 5A = 6−17 = 617 − 6 0 7 − 617 = 6 −5/2 7 ,
6 7 2 6 7 26 7 2 6 7 6 7
2 4 5 4 5 4 5 4 5 4 5
8 1 −1 0 −13/2
B

and
2 3 2 3 2 3 2 3 2 3
02 31 4 1 1 0 15/2
6 7 6 7 6 7 6 7 6 7
6 7 6 7 6 7 6 7 6 7
B617C 6 7 15 6 7 76 7 19 6 7 6 7
T 4 5A = 6−27 = 617 − 6 0 7 − 617 = 6 −7/2 7 .
6 7 2 6 7 26 7 2 6 7 6 7
3 4 5 4 5 4 5 4 5 4 5
11 1 −1 0 −19/2
B

The required matrix is

2 3
6 11/2 15/2 7
6 7
6 7
6 −5/2 −7/2 7 .
6 7
4 5
−13/2 −19/2
B

564 The matrix will be a 2 × 3 matrix. In each case, we find the action of L on the basis elements of R3 and express the
result in the given basis for R3 .
1. We have
02 31 02 31 02 31
1 2 3 0 2 3 2 3
B6 7C B6 7C B607C
B6 7C 1 B6 7C 2 B6 7C
B6 7C 6 7 B6 7C 6 7 B6 7C 6 0 7
L B607C = 4 5 , L B617C = 4 5 , L B607C = 4 5 .
B6 7C B6 7C B6 7C
4 5A 3 4 5A 0 4 5A −1
0 0 1

The required matrix is

2 3
61 2 0 7
4 5.
3 0 −1

2. We have
02 31 02 31 02 31
1 2 3 1 2 3 2 3
B6 7C B6 7C B617C
B6 7C 1 B6 7C 3 B6 7C
B6 7C 6 7 B6 7C 6 7 B6 7C 637
L B607C = 4 5 , L B617C = 4 5 , L B617C = 4 5 .
B6 7C B6 7C B6 7C
4 5A 3 4 5A 3 4 5A 2
0 0 1

The required matrix is

2 3
61 3 37
4 5.
3 3 2
Answers and Hints 221

3. We have 02 31
2 3 2 3 2 3 2 3
B617C
B6 7C
B6 7C 6 71 1
6 7 617 6−27
L B607C = 4 5 = −2 4 5 + 3 4 5 = 4 5 ,
B6 7C
4 5A 3 0 1 3
A
0
02 31
2 3 2 3 2 3 2 3
B617C
B6 7C
B6 7C 637 617 6 1 7 60 7
L B617C = 4 5 = 0 4 5 + 3 4 5 = 4 5 ,
B6 7C
4 5A 3 0 1 3
A
0
02 31
2 3 2 3 2 3 2 3
B617C
B6 7C 3 1
B6 7C 6 7 6 7 6 1 7 61 7
L B617C = 4 5 = 1 4 5 + 2 4 5 = 4 5 .
B6 7C
4 5A 2 0 1 2
A
1
The required matrix is
2 3
6−2 0 17
4 5.
3 3 2
2 3
62 7
565 Observe that 4 5 ∈ Im (T ) = ker (T ) and so
3
2 3 2 3
627 607
T 4 5 = 4 5.
3 0

Now 2 3 0 2 3 2 31 2 3 2 3 2 3
1
6 7 1
B 6 7 6 7C2 1
6 7 627 667
T 4 5 = T 3 4 5 − 4 5A = 3T 4 5 − T 4 5 = 4 5 ,
0 1 3 1 3 9
and 2 3 02 3 2 31 2 3 2 3 2 3
0
6 7 2
B6 7 1
6 7C 2
6 7 617 6−47
T 4 5 = T 4 5 − 2 4 5A = T 4 5 − 2T 4 5 = 4 5 .
1 3 1 3 1 −6
The required matrix is thus
2 3
66 −47
4 5.
9 −6

566 The matrix will be a 1 × 4 matrix. We have

02 31
B61 07C
tr 4 5A = 1,
0 0
02 31
B60 17C
tr 4 5A = 0,
0 0
222 Appendix A

02 31
B60 07C
tr 4 5A = 0,
1 0
02 31
B60 07C
tr 4 5A = 1.
0 1
Thus
ML = (1 0 0 1).
567 First observe that ker (B) ⊆ ker (AB) since ∀X ∈ Mq×1 (R),
BX = 0 =⇒ (AB)X = A(BX) = 0.
Now
dim ker (B) = q − dim Im (B)

= q − rank (B)

= q − rank (AB)

= q − dim Im (AB)

= dim ker (AB) .

Thus ker (B) = ker (AB) . Similarly, we can demonstrate that ker (ABC) = ker (BC) . Thus

rank (ABC) = dim Im (ABC)

= r − dim ker (ABC)

= r − dim ker (BC)

= dim Im (BC)

= rank (BC) .

594 This is clearly (1 2 3 4)(6 8 7) of order lcm(4, 3) = 12.

621 Multiplying the first column of the given matrix by a, its second column by b, and its third column by c, we obtain
2 3
6abc abc abc7
6 7
6 7
abcΩ = 6 a2 b2 c2 7 .
6 7
4 5
a3 b3 c3

We may factor out abc from the first row of this last matrix thereby obtaining
2 3
61 1 17
6 7
6 7
abcΩ = abc det 6a2 b2 c2 7 .
6 7
4 5
a3 b3 c3

Upon dividing by abc,

2 3
61 1 17
6 7
6 7
Ω = det 6a2 b2 c2 7 .
6 7
4 5
a3 b3 c3
Answers and Hints 223

622 Performing R1 + R2 + R3 → R1 we have

2 3
6a − b − c 2a 2a 7
6 7
6 7
Ω = det 6 2b b−c−a 2b 7
6 7
4 5
2c 2c c−a−b
2 3
6a + b + c a+b+c a + b + c7
6 7
6 7
= det 6 2b b−c−a 2b 7.
6 7
4 5
2c 2c c−a−b

Factorising (a + b + c) from the first row of this last determinant, we have

2 3
61 1 1 7
6 7
6 7
Ω = (a + b + c) det 62b b−c−a 2b 7.
6 7
4 5
2c 2c c−a−b

Performing C2 − C1 → C2 and C3 − C1 → C3 ,
2 3
61 0 0 7
6 7
6 7
Ω = (a + b + c) det 62b −b − c − a 0 7.
6 7
4 5
2c 0 −c − a − b

This last matrix is triangular, hence

Ω = (a + b + c)(−b − c − a)(−c − a − b) = (a + b + c)3 ,

as wanted.

623 det A1 = det A = −540 by multilinearity. det A2 = − det A1 = 540 by alternancy. det A3 = 3 det A2 = 1620 by both
multilinearity and homogeneity from one column. det A4 = det A3 = 1620 by multilinearity, and det A5 = 2 det A4 = 3240
by homogeneity from one column.

624 From the given data, det B = −2. Hence

det ABC = (det A)(det B)(det C) = −12,

det 5AC = 53 det AC = (125)(det A)(det C) = 750,

(det A)3 27
(det A3 B−3 C−1 ) = =− .
(det B)3 (det C) 16

625 Pick λ ∈ R \ {0, a11 , a22 , . . . , ann }. Put

2 3
6a11 − λ 0 0 ··· 0 7
6 7
6 7
6 a21 a22 − λ 0 ··· 0 7
6 7
6 7
6 7
X = 6 a31 a32 a33 − λ ··· 0 7
6 7
6 7
6 .. .. .. .. .. 7
6 . . . . . 7
6 7
4 5
an1 an2 an3 ··· ann − λ
224 Appendix A

and

2 3
..
6λ a12 a13 . a1n 7
6 7
6 .. 7
60 λ a23 . a2n 7
6 7
6 7
6 .. 7
Y = 60 0 λ . a3n 7
6 7
6 7
6 .. .. .. .. .. 7
6. . . . . 7
6 7
4 .. 5
0 0 0 . λ

Clearly A = X + Y , det X = (a11 − λ)(a22 − λ) · · · (ann − λ) 6= 0, and det Y = λn 6= 0. This completes the proof.

626 No.

636 We have

2 3 2 3 2 3
64 67 61 37 61 37
det A = 2(−1)1+2 det 4 5 + 5(−1)
2+2
det 4 5 + 8(−1)
2+3
det 4 5
7 9 7 9 4 6
= −2(36 − 42) + 5(9 − 21) − 8(6 − 12) = 0.

637 Since the second column has three 0’s, it is advantageous to expand along it, and thus we are reduced to calculate

2 3
61 −1 17
6 7
6 7
−3(−1)3+2 det 62 0 17
6 7
4 5
1 0 1

Expanding this last determinant along the second column, the original determinant is thus

2 3
62 17
−3(−1)3+2 (−1)(−1)1+2 det 4 5 = −3(−1)(−1)(−1)(1) = 3.
1 1
Answers and Hints 225

638 Expanding along the first column,

2 3
61 1 1 17
6 7
6 7
6x a 0 07
6 7
0 = det 6 7
6 7
6x 0 b 07
6 7
4 5
x 0 0 c
2 3 2 3
6a 0 07 61 1 17
6 7 6 7
6 7 6 7
= det 6 0 b 07 − x det 60 b 07
6 7 6 7
4 5 4 5
0 0 c 0 0 c
2 3 2 3
61 1 17 61 1 17
6 7 6 7
6 7 6 7
+x det 6a 0 07 − x det 6a 0 07
6 7 6 7
4 5 4 5
0 0 c 0 b 0
2 3 2 3
61 1 17 61 1 17
6 7 6 7
6 7 6 7
= xabc − xbc + x det 6a 0 077 − x det 6a 0 07 .
6 6 7
4 5 4 5
0 0 c 0 b 0

Expanding these last two determinants along the third row,

2 3 2 3
61 1 17 61 1 17
6 7 6 7
6 7 6 7
0 = abc − xbc + x det 6a 0 077 − x det 6a a 07
6 6 7
4 5 4 5
0 0 c 0 b 0
2 3 2 3
61 17 61 17
= abc − xbc + xc det 4 5 + xb det 4 5
a 0 a 0

= abc − xbc − xca − xab.

It follows that
abc = x(bc + ab + ca),
whence
1 bc + ab + ca 1 1 1
= = + + ,
x abc a b c
as wanted.

639 Expanding along the first row the determinant equals

2 3 2 3
2 3 2 3
6a b 07 6a 0 07
6 7 6 7
6 7 6 7 6a b7 6a b7
−a det 6 0 0 b77 + b det 60 a b7 = ab det 4 5 + ab det 4 5
6 6 7
4 5 4 5 1 1 1 1
1 1 1 1 1 1

= 2ab(a − b),

as wanted.
226 Appendix A

640 Expanding along the first row, the determinant equals

2 3 2 3
6a 0 b7 60 a b7
6 7 6 7
6 7 6 7
a det 6 0 d 07 + b det 6c 0 07 .
6 7 6 7
4 5 4 5
c 0 d 0 c d

Expanding the resulting two determinants along the second row, we obtain

2 3 2 3
6a b7 6a b7
2
ad det 4 5 + b(−c) det 4 5 = ad(ad − bc) − bc(ad − bc) = (ad − bc) ,
c d c d

as wanted.

641 For n = 1 we have det(1) = 1 = (−1)1+1 . For n = 2 we have

2 3
61 17
2+1
det 4 5 = −1 = (−1) .
1 0

Assume that the result is true for n − 1. Expanding the determinant along the first column

2 3
··· 2 3
61 1 1 1 17 ..
6 7 .
6 .. 7 60 0 0 07
61 0 0 . 0 07 6 7
6 7 6 7
6 7 61 0 ··· 0 07
6 7 6 7
60 1 0 ··· 0 07 6 7
6 7 6 7
det 6 7 = 1 det 60 1 ··· 0 07
6 7 6 7
60 0 1 ··· 0 07 6 7
6 7 6 .. .. .. 7
6. 7 6. ··· . . 7
6. .. .. .. 7 6 7
6. . ··· . . 7 4 5
6 7
4 5 0 0 ··· 1 0
0 0 0 ··· 1 0
2 3
61 1 ··· 1 17
6 7
6 .. 7
61 0 ··· . 07
6 7
6 7
6 7
60 1 ··· ··· 07
6 7
−1 det 6 7
6 7
60 0 ··· ··· 07
6 7
6. 7
6. .. .. 7
6. . ··· . 7
6 7
4 5
0 0 ··· 1 0

= 1(0) − (1)(−1)n

= (−1)n+1 ,

giving the result.

Answers and Hints 227

642 Perform Ck − C1 → Ck for k ∈ [2; n]. Observe that these operations do not affect the value of the determinant. Then
2 3
61 n−1 n−1 n−1 ··· n − 17
6 7
6 .. 7
6n 2−n 0 0 . 0 7
6 7
6 7
6 7
6n 0 3−n 0 ··· 0 7
6 7
det A = det 6 7.
6 7
6n 0 0 4−n ··· 0 7
6 7
6. 7
6. .. .. .. 7
6. . . ··· . 7
6 7
4 5
n 0 0 0 0 0

Expand this last determinant along the n-th row, obtaining,

2 3
6n − 1 n−1 n−1 ··· n−1 n − 17
6 7
6 .. 7
62 − n 0 0 . 0 0 7
6 7
6 7
6 7
6 0 3−n 0 ··· 0 0 7
6 7
det A = (−1)1+n n det 6 7
6 7
6 0 0 4−n ··· 0 0 7
6 7
6 . 7
6 . .. .. .. 7
6 . . ··· . . 7
6 7
4 5
0 0 0 ··· −1 0

= (−1)1+n n(n − 1)(2 − n)(3 − n)

2 3
61 1 1 ··· 1 17
6 7
6 .. 7
61 0 0 . 0 07
6 7
6 7
6 7
60 1 0 ··· 0 07
6 7
· · · (−2)(−1) det 6 7
6 7
60 0 1 ··· 0 07
6 7
6. 7
6. .. .. .. 7
6. . ··· . . 7
6 7
4 5
0 0 0 ··· 1 0
2 3
61 1 1 ··· 1 17
6 7
6 .. 7
61 0 0 . 0 07
6 7
6 7
6 7
60 1 0 ··· 0 07
6 7
= −(n!) det 6 7
6 7
60 0 1 ··· 0 07
6 7
6. 7
6. .. .. .. 7
6. . ··· . . 7
6 7
4 5
0 0 0 ··· 1 0

= −(n!)(−1)n

= (−1)n+1 n!,

upon using the result of problem 641.

228 Appendix A

n

n


643 Recall that k
= n−k
,
n  
X n
= 2n
k
k=0

and  
n
X nk
(−1) = 0, if n > 0.
k
k=0


Assume that n is odd. Observe that then there are n + 1 (an even number) of columns and that on the same row, n is on a
n

k
column of opposite parity to that of n−k . By performing C1 − C2 + C3 − C4 + · · · + Cn − Cn+1 → C1 , the first column
becomes all 0’s, whence the determinant if 0 if n is odd.

659 We have
2 3
6λ − 1 1 7 2
det(λI2 − A) = det 4 5 = (λ − 1) − 1 = λ(λ − 2),
1 λ−1

whence the eigenvalues are 0 and 2. For λ = 0 we have

2 3
6−1 1 7
0I2 − A = 4 5.
1 −1

This has row-echelon form

2 3
61 −17
4 5.
0 0

If
2 32 3 2 3
61 −17 6a7 607
4 54 5 = 4 5
0 0 b 0

then a = b. Thus
2 3 2 3
a
6 7 61 7
4 5 = a4 5
b 1
2 3
617
and we can take 4 5 as the eigenvector corresponding to λ = 0. Similarly, for λ = 2,
1

2 3
61 37
2I2 − A = 4 5,
1 3

which has row-echelon form

2 3
61 37
4 5.
0 0

If
2 32 3 2 3
61 37 6a7 607
4 54 5 = 4 5
0 0 b 0
Answers and Hints 229

then a = −3b. Thus

2 3 2 3
6a 7 6 1 7
4 5 = a4 5
b −3

2 3
6 1 7
and we can take 4 5 as the eigenvector corresponding to λ = 2.
−3

660 ➊ We have

2 3
6 λ −2 17
6 7
6 7
det(λI3 − A) = det 6−2 λ−3 27
6 7
4 5
1 2 λ
2 3 2 3 2 3
6λ − 3 27 6−2 27 6−2 λ − 37
= λ det 4 5 + 2 det 4 5 + det 4 5
2 λ 1 λ 1 2

= λ(λ2 − 3λ − 4) + 2(−2λ − 2) + (−λ − 1)

= λ(λ − 4)(λ + 1) − 5(λ + 1)

= (λ2 − 4λ − 5)(λ + 1)

= (λ + 1)2 (λ − 5)

➋ The eigenvalues are −1, −1, 5.

➌ If λ = −1,
2 32 3 2 3
6−1 −2 1 7 6 a 7 60 7
6 76 7 6 7
6 76 7 6 7
(−I3 − A) = 6−2 −4 2 7 6 7=6 7 ⇐⇒ a = −2b + c
6 7 6 b 7 60 7
4 54 5 4 5
1 2 −1 c 0
2 3 2 3 2 3
6a 7 6−27 617
6 7 6 7 6 7
6 7 6 7 6 7
⇐⇒ 6b7 = b 6 1 7 + c 607 .
6 7 6 7 6 7
4 5 4 5 4 5
c 0 1

2 3 2 3
6−27 617
6 7 6 7
6 7 6 7
We may take as eigenvectors 6 1 7 , 607, which are clearly linearly independent.
6 7 6 7
4 5 4 5
0 1
230 Appendix A

If λ = 5,
2 32 3 2 3
6 5 −2 1 7 6 a 7 60 7
6 76 7 6 7
6 76 7 6 7
(5I3 − A) = 6−2 2 27 6 7=6 7 ⇐⇒ a = −c, b = −2c,
6 7 6 b 7 60 7
4 54 5 4 5
1 2 5 c 0
2 3 2 3
6a7 6−17
6 7 6 7
6 7 6 7
⇐⇒ 6b7 = c 6−27 .
6 7 6 7
4 5 4 5
c 1
2 3
6 1 7
6 7
6 7
We may take as eigenvector 6 2 7.
6 7
4 5
−1

664 Solution: Put

2 3 2 3
61 0 7 61 0 7
D=4 5, ,P = 4 5.
0 −2 1 −1

We find
2 3
61 1 7
P−1 =4 5.
0 −1

Since A = PDP−1 2 32 32 3 2 3
61 0 7 61 0 7 61 1 7 61 −10237
A10 = PD10 P−1 =4 54 54 5=4 5.
1 −1 0 1024 0 −1 0 1024

665 Put 2 3 2 3
6−1 0 07 61 1 17
6 7 6 7
6 7 6 7
D=6 0 −1 07 , X = 60 1 17 .
6 7 6 7
4 5 4 5
0 0 3 0 0 1

Then we know that A = XDX−1 and so we need to find X−1 . But this is readily obtained by performing R1 − R2 → R1 and
R2 − R3 → R3 in the augmented matrix
2 3
61 1 1 1 0 07
6 7
6 7
60 1 1 0 1 07 ,
6 7
4 5
0 0 1 0 0 1

getting
2 3
61 −1 0 7
6 7
6 7
X−1 = 60 1 −17 .
6 7
4 5
0 0 1
Answers and Hints 231

Thus
A = XDX−1
2 32 32 3
61 1 17 6−1 0 07 61 −1 0 7
6 76 76 7
6 76 76 7
= 60 1 17 6 −1 07 6 −17
6 76 0 7 60 1 7
4 54 54 5
0 0 1 0 0 3 0 0 1
2 3
6−1 0 47
6 7
6 7
= 6 0 −1 47 .
6 7
4 5
0 0 3

666 The determinant is 1, A = A−1 , and the characteristic polynomial is (λ2 − 1)2 .
Bibliography

[Cul] CULLEN, C., Matrices and Linear Transformations, 2nd ed., New York: Dover Publications, 1990.
[Del] DELODE, C., Géométrie Affine et Euclidienne, Paris: Dunod, 2000.
[Fad] FADDÉEV, D., SOMINSKY, I., Recueil d’Exercises d’Algèbre Supérieure, 7th ed., Paris: Ellipses, 1994.
[Hau] HAUSNER, M., A Vector Space Approach to Geometry, New York: Dover, 1998.
[Lan] LANG, S., Introduction to Linear Algebra, 2nd ed., New York: Springer-Verlag, 1986.
[Pro] PROSKURYAKOV, I. V., Problems in Linear Algebra, Moscow: Mir Publishers, 1978.
[Riv] RIVAUD, J., Ejercicios de álgebra, Madrid: Aguilar, 1968.
[Tru] TRUFFAULT, B., Géométrie Élémentaire: Cours et exercises, Paris: Ellipses, 2001.

232