Part 2.

Linear Algebra

Chapter 1: Systems of Linear Equations

Main subjects

1. Solutions and Elementary Operations

2. Gaussian Elimination

3. Homogeneous Equations

1
 1.1
Solutions and Elementary
Operations

2
REVIEW: SOLUTIONS OF A SYSTEM OF LINEAR EQUATIONS

Solve the following systems of linear equations:

5𝑥 − 2𝑦 = 4 𝑥 − 2𝑦 = 5 3𝑥 + 4𝑦 = 4
(1) ቊ (2) ቊ (3) ቊ
3𝑥 − 𝑦 = 3 −3𝑥 + 6𝑦 = −11 9𝑥 + 12𝑦 = 12

Solution.
1. 𝑥 = 2 and 𝑦 = 3.
2. The system has no solutions.
4−3𝑡
3. The system has infinitely many solutions: 𝑥 = 𝑡, 𝑦 = 𝑡 ∈ ℝ}.
4

3
DISCUSSION

Problem. Solve the following system of linear equations

𝑥 + 2𝑦 = −2 (1)
Ι ቊ
−3𝑥 − 𝑦 = 11 (2)
Solution.

𝑥 + 2𝑦 = −2
(ΙΙ) ቊ (adding three times the first to the second)
0𝑥 + 5𝑦 = 5

𝑥 + 2𝑦 = −2 1
ΙΙΙ ቊ (multiplying the second by )
𝑦=1 5

𝑥 = −4
ΙV ቊ (subtract twice the second from the first)
𝑦=1
4
DISCUSSION (cont)

Question. What have we done to get the solutions 𝑥 and 𝑦?

Recall. Two systems of equations are called equivalent if they have the same set of
solutions.
For instance, all the systems Ι , ΙΙ , ΙΙΙ and ΙV in the previous page are equivalent.

Answer. What we have performed are called the elementary operations.

The purpose is to obtain equivalent systems that are easier to solve.

5
ELEMENTARY OPERATIONS

To produce equivalent systems of linear equations, the following elementary

operations can be used:

(1) Interchange any two equations.

(2) Multiply one equation by a nonzero number.

(3) Add a multiple of one equation to another equation.

6
DISCUSSION (cont)

Observe how coefficients and constant terms change if we apply elementary

operations.

𝑥 + 2𝑦 = −2 1 2 −2
ቊ
−3𝑥 − 𝑦 = 11 −3 −1 11

𝑥 + 2𝑦 = −2 1 2 −2
ቊ
0𝑥 + 5𝑦 = 5 0 5 5

𝑥 + 2𝑦 = −2 1 2 −2
ቊ
𝑦=1 0 1 1

𝑥 = −4 1 0 −4
ቊ
𝑦=1 0 1 1
7
AUGMENTED MATRICES
𝑥 + 2𝑦 = −2
∎ The augmented matrix of the system ቊ
−3𝑥 − 𝑦 = 11
is the array of numbers
1 2 −2
−3 −1 11

∎ For a system of the form

𝑎𝑥1 +𝑏𝑥2 + 𝑐𝑥3 = 𝑑

൞ 𝑎′𝑥1 + 𝑏′𝑥2 + 𝑐′𝑥3 = 𝑑′
𝑎′′𝑥1 + 𝑏 ′′ 𝑥2 + 𝑐 ′′ 𝑥3 = 𝑑 ′′ ,
the corresponding augmented matrix is
𝑎 𝑏 𝑐 𝑑
𝑎′ 𝑏′ 𝑐′ 𝑑′
𝑎′′ 𝑏 ′′ 𝑐 ′′ 𝑑 ′′
8
EXERCISE
1. Write down the augmented matrix of the following system
3𝑥 + 𝑦 − 4𝑧 = −1
ቐ 𝑥 + 10𝑧 = 5
4𝑥 + 𝑦 =1
Solution.
3 1 −4 −1
1 0 10 5 .
4 1 0 1

2. Write down the system of linear equations corresponding to the following

2 0 0 −1 15
augmented matrix 1 0 5 13 5 .
4 1 0 0 −31

2𝑥 − 𝑡 = 15
ቐ𝑥 + 5𝑧 + 13𝑡 = 5
4𝑥 + 𝑦 = −31
9
EXAMPLE
3𝑥 + 2𝑦 = 1
Solve the system of linear equations ቊ using augmented matrices.
6𝑥 − 𝑦 = 7

3 2 1
Solution. The augmented matrix of this system is
6 −1 7
We transform this as follows

3 2 1
(subtract twice the first row from the second)
0 −5 5
3 2 1 1
(multiply the second row by − )
0 1 −1 5

3 0 3
(subtract twice the second row from the first)
0 1 −1
1 0 1 1
(multiply the first row by )
0 1 −1 3

From this, we deduce that 𝑥 = 1 and 𝑦 = −1 are the solution.

10
ELEMENTARY OPERATIONS (cont)

∎ To solve a system of linear equations, the following elementary operations

can be applied to augmented matrices:
(1) Interchange any two rows.
(2) Multiply one row by a nonzero number.
(3) Add a multiple of one row to another row.

∎ The ultimate goal is to try to obtain the augmented matrix of the form
∗ 1 0 0 ∗
1 0 ∗ .
or 0 1 0
0 1 ∗
0 0 1 ∗

11
EXAMPLE
Solve the system of linear equations

using augmented matrices.

Solution. We apply the elementary operations on the augmented matrix of the system.

Row1 − Row2

Row2 −2 ×Row1 Row3 − 4 ×Row1

→ →
12
 Row1 − Row2 1 0 3 3
0 1 −2 −2
→ 0 −1 −5 −6

1
Row3 + Row2 − × Row3
7

→ →

Row1 −3 × Row3

→
Row2 +2 × Row3

3 2 8
We deduce that 𝑥 = − , 𝑦 = and 𝑧 = are the solution.
7 7 7
13
EXERCISE
Exercise 1. Write down the systems of linear equations whose augmented matrices
are given below.

Exercise 2. 1.1.10 ([2], p.8).

14
 1.2
Gaussian Elimination

15
ROW-ECHELON MATRICES

∗ refers to any real number.

Definition. A row-echelon matrix is a matrix that satisfies three conditions:

1. All zero rows are at the bottom.
2. The first nonzero entry from the left in each nonzero row is a 1, called the leading 𝟏 for
that row.
3. Each leading 1 is to the right of all leading 1’s in the rows above it.

16
REDUCED ROW-ECHELON MATRICES

Definition. A reduced row-echelon matrix is a row-echelon matrix that satisfies the

following extra condition:
4. Each leading 1 is the only nonzero entry in its column.

17
EXAMPLE
Exercise 1.2.1 (page 17, [2]).

Solution. (a), (b) and (f) are not row-echelon matrices.

(c), (d) and (e) are row-echelon matrices, but not reduced row-echelon matrices.

18
EXERCISE

Write down a row-echelon matrix and a reduced row-echelon matrix with 4 rows and
5 columns.

19
GAUSSIAN ALGORITHM
Theorem. Every matrix can be brought to (reduced) row-echelon form by a sequence of
elementary row operations.

To bring a matrix to row-echelon form proceed as follows:

Step 1. If the matrix consists entirely of zeros, stop.
Step 2. Otherwise, find the first column from the left containing a nonzero entry (call it a),
and move the row containing that entry to the top position.
Step 3. Multiply the new top row by 1/𝑎 to create a leading 1.
Step 4. By subtracting multiples of that row from rows below it, make each entry below the
leading 1 zero.
These four steps completes the first row.
Step 5. Repeat steps 1–4 on the matrix consisting of the remaining rows.
20
EXAMPLE

1
0 2 3 Row1 ↔ Row2 3 1 2 3
×Row1 1 1/3 2/3
3 1 2 0 2 3 0 2 3
3 4 1 → 3 4 1 → 3 4 1

Row3 − 3 ×Row1 1 1/3 2/3

1
×Row2 1 1/3 2/3 Row3 − 3 ×Row2
1 1/3 2/3
2
0 2 3 0 1 3/2 0 1 3/2
→ 0 3 −1 → 0 3 −1 → 0 0 −11/2

(complete the first row) (complete the second row)

1
Row1 − ×Row3
−
2
×Row3 1 1/3 2/3 1
Row1 − ×Row2 1 0 1/6 6 1 0 0
11 3

→
0
0
1
0
3/2
1 → 0 1 3/2
0 0 1
→ 3
0
0
1
0
0
1
Row2 − ×Row3
2
(row-echelon form) (reduced row-echelon form)

21
EXAMPLE
Solve the following system of equations

3𝑥 + 𝑦 − 4𝑧 = −1
ቐ 𝑥 + 10𝑧 = 5
4𝑥 + 𝑦 + 6𝑧 = 1
Solution. We use the Gaussian algorithm to carry the augmented matrix of the system
into reduced row-echelon form.
Row2 −3 ×Row1
Row2 ↔ Row1

→ →
Row3 −4 ×Row1

Row3 − Row2
1
− ×Row3
3
1 0 10 5
0 1 −34 −16
→ → 0 0 0 1
(row-echelon form)
22
 1 0 10 5 Row1 −5 × Row3 1 0 10 0
0 1 −34 −16 → 0 1 −34 0
0 0 0 1 Row2 +16 ×Row3 0 0 0 1

(reduced row-echelon form)

Therefore, the original system is equivalent to the system

𝑥 + 10𝑧 = 0
ቐ𝑦 − 34𝑧 = 0
0=1

From this, we deduce that the system has no solution.

23
EXAMPLE
Find a condition on 𝑎, 𝑏 and 𝑐 such that the following system admits solution

𝑥 + 3𝑦 + 𝑧 = 𝑎
ቐ−𝑥 − 2𝑦 + 𝑧 = 𝑏
3𝑥 + 7𝑦 − 𝑧 = 𝑐
Solution. We use the Gaussian algorithm to carry the augmented matrix of the system
(closely) to reduced row-echelon form.

Row2 + Row1

→
Row3 − 3 ×Row1

Row1 − 3 × Row2

→
Row3 + 2 ×Row2

24
Therefore, the original system is equivalent to the system

𝑥 − 5𝑧 = −2𝑎 − 3𝑏
ቐ 𝑦 + 2𝑧 = 𝑎 + 𝑏
0 = 𝑐 − 𝑎 + 2𝑏

We see that the system admits solution if and only if 𝑐 = 𝑎 − 2𝑏.

Taking 𝑧 = 𝑡, we find that 𝑦 = 𝑎 + 𝑏 − 2𝑡 and 𝑥 = 5𝑡 − 2𝑎 − 3𝑏.

25
RANK OF A MATRIX

Definition. Rank of a matrix is the number of nonzero rows of its row-echelon

form.

Example. Determine the rank of each of the following matrices

1 0 0
0 1 0
0 0 1
Rank = 3 Rank = 3 Rank = 2

In general, to compute the rank of a matrix, we first need to carry it to the row-echelon form.

26
EXAMPLE
Find the rank of the matrix

𝐴=

Let us carry 𝐴 into the row-echelon form.

Row2 − 2 ×Row1 −1 ×Row2 1 1 −1 4

→ 0 1 −5 8
→ 0 1 −5 8

Row3 − Row2

→ We deduce that Rank(𝐴) = 2.

(row-echelon form)
27
EXERCISE
Consider the system of equations

𝑥1 − 2𝑥2 − 𝑥3 + 3𝑥4 = 1
ቐ 2𝑥1 − 4𝑥2 + 𝑥3 = 5
𝑥1 − 2𝑥2 + 2𝑥3 − 3𝑥4 = 4
a. Find the rank of the augmented matrix of the system.
b. Solve the system.

Row2 −2 ×Row1

→
Row3 − Row1

1
3
×Row2 1 −2 −1 3 1 Row3 − 3 ×Row2 1 −2 −1 3 1
0 0 1 −2 1 → 0 0 1 −2 1
→ 0 0 3 −6 3 0 0 0 0 0
(row-echelon form)
28
 1 −2 −1 3 1 Row1 + Row2 1 −2 0 1 2
0 0 1 −2 1 0 0 1 −2 1
0 0 0 0 0
→ 0 0 0 0 0
(row-echelon form) (reduced row-echelon form)

(a) The row-echelon form of 𝐴 has two nonzero rows. So we have Rank(𝐴) = 2.
(b) The original system is equivalent to the system

𝑥1 − 2𝑥2 + 𝑥4 = 2
ቊ
𝑥3 − 2𝑥4 = 1

𝑥1 and 𝑥3 are leading variables. So we put 𝑥2 = 𝑠 and 𝑥4 = 𝑡.

The solution is {𝑥1 = 2 + 2𝑠 − 𝑡, 𝑥2 = 𝑠, 𝑥3 = 1 + 2𝑡, 𝑥4 = 𝑡 𝑠, 𝑡 ∈ ℝ .
RANK AND SOLUTIONS

Theorem. Suppose a system of linear equations in 𝑛 variables admits solutions.

Denote by 𝑟 the rank of the augmented matrix of the system.

Then

1. 𝑟 ≤ 𝑛.

2. The set of solutions involves exactly 𝑛 − 𝑟 parameters.

3. If 𝑟 < 𝑛: the system has infinitely many solutions.

4. If 𝑟 = 𝑛: the system admits a unique solution.

30
 1.3
Homogeneous Equations

31
DEFINITION

Definition. A system of equations is said to be homogeneous if all the constant

terms are zero.

Example. The following system of equations is homogeneous

2𝑥 + 𝑦 − 3𝑧 + 𝑡 = 0
ቐ 𝑦 − 5𝑧 + 3𝑡 = 0
−𝑥 + 𝑦 − 𝑧 + 𝑡 = 0

Exercise. Solve this system of equations.

32
EXAMPLE
Solve the following homogeneous system of equations
𝑥1 − 𝑥2 + 2𝑥3 − 𝑥4 = 0
ቐ 2𝑥1 + 2𝑥2 + 𝑥4 = 0
3𝑥1 + 𝑥2 + 2𝑥3 − 𝑥4 = 0
Solution. Let us transform the augmented matrix of the system into the reduced row-echelon form.

Row2 −2 × Row1

→
Row3 −3 ×Row1

1
×Row2 1 −1 2 −1 0 Row1 + Row2 1 0 1 −1/4 0
→
4
0 1 −1 3/4 0 0 1 −1 3/4 0
→ 0 4 −4 2 0 Row3 −4 ×Row2 0 0 0 −1 0

1
Row1 + ×Row3
−1 ×Row3 1 0 1 −1/4 0 4

→ 0 1 −1 3/4 0 →
0 0 0 1 0 3
Row2 − ×Row3
4 33
So the original system is equivalent to the system
𝑥1 + 𝑥3 = 0
ቐ𝑥2 − 𝑥3 = 0
𝑥4 = 0

𝑥1 , 𝑥2 and 𝑥4 are leading variables. So we put 𝑥3 = 𝑡. The solution is

{ 𝑥1 = −𝑡, 𝑥2 = 𝑥3 = 𝑡, 𝑥4 = 0 𝑡∈ℝ .

Remark. A homogeneous system of linear equations always has a trivial

solution: 𝑥1 = 𝑥2 = ⋯ = 𝑥𝑛 = 0.

Theorem. If a homogeneous system of linear equations has more variables

than equations, then it has infinitely many solutions.
34
QUIZ QUESTION

How many solutions must a homogeneous linear system of 6 equations

in 7 unknowns have?
A. Infinitely many solutions.
B. Exactly one solution.
C. None of the other choices is correct.
D. No solutions.

35
QUIZ QUESTION

The row-echelon form of the augmented matrix of a system is given as follows

𝐴=

where ∗ is any real number. Choose the correct answer.

A. The system has no solution.
B. The system has a unique solution.
C. The system has infinitely many solutions.

Answer. C (since the system has 5 variables and moreover, Rank 𝐴 = 4 < 5).
36
EXERCISE

1.2.5, 1.2.7, 1.2.11 (a, b, c, d), 1.2.12, 1.3.2.

37
 Part 2. Linear Algebra
Chapter 2: Matrix Algebra
Main subjects

1. Matrix Addition, Scalar Multiplication and Transposition

2. Matrix Multiplication

3. Matrices and systems of linear equations

4. Matrix Inverses

5. Matrix Transformations

38
 2.1
Matrix addition, scalar multiplication
and transposition

39
MATRICES
Definition.

∎ A matrix is a rectangular array of numbers.

∎ The numbers in a matrix are called the entries.

For example, the entries of 𝐵 are 1, −1, 0 and 2.

∎ A matrix with 𝑚 rows and 𝑛 columns is called an 𝒎 × 𝒏 matrix or a matrix of

size 𝒎 × 𝒏.

For example, 𝐴 is an 2 × 3 matrix, 𝐵 is an 2 × 2 matrix and 𝐶 is an 3 × 1 matrix.

40
MATRICES (cont.)

∎ The number lying in row 𝒊 and column 𝒋 is called the (𝒊, 𝒋)-entry.

1 3 −1 4
𝐴= 0 −4 6 8
2 1 −5 9

For example, the (3,2)-entry of the matrix 𝐴 is 1, while the (2,4)-entry is 8.

41
MATRICES (cont.)
If 𝐴 is a matrix of size 𝑚 × 𝑛 then it is usually displayed as follows

For simply, one writes 𝐴 = 𝑎𝑖,𝑗 .

𝑚×𝑛

For example, if 𝐴 = 𝑎𝑖,𝑗 then 𝐴 is displayed as follows

3×4

42
EXERCISE

Write down the matrix 𝐴 = 𝑎𝑖,𝑗 with 𝑎1,2 = 1, 𝑎1,1 = 9, 𝑎2,2 = −4,
3×2

𝑎3,2 = 0, 𝑎3,1 = −7 and 𝑎2,1 = 5.

Solution.

9 1
𝐴 = 5 −4
−7 0

43
EXERCISE

Write down the matrix 𝐴 = 𝑎𝑖,𝑗 with 𝑎𝑖,𝑗 = 𝑖 + 2𝑗.

2×3

Solution. We have 𝑎1,1 = 1 + 2 × 1 = 3, 𝑎1,2 = 1 + 2 × 2 = 5, 𝑎1,3 = 1 + 2 × 3 = 7,

𝑎2,1 = 4, 𝑎2,2 = 6 and 𝑎2,3 = 8.

3 5 7
𝐴=
4 6 8

44
SQUARE MATRICES AND SYMMETRIC MATRICES

∎ An (𝑚, 𝑛)-matrix is called a square matrix if 𝑚 = 𝑛.

For instance, the following matrices are square matrices.

1 −5 1 2 2
4 7 −7 −4 5
8 0.5 6

∎ A square matrix 𝐴 = 𝑎𝑖,𝑗 is said to be symmetric if 𝑎𝑖,𝑗 = 𝑎𝑗,𝑖 for every 𝑖, 𝑗.

𝑛×𝑛

7 12 5 9
1 2 − 2
1 −5 2 4 5 12 3 4 0
−5 7 5 4 13 45
− 2 5 6
9 0 45 57
45
EXERCISE

Fill in the blanks with suitable numbers to make a symmetric matrix:

1 −3

8 −7 6

Answer. 1 −3 8
−3 any −7
8 −7 6

46
MAIN DIAGONAL

Definition. The main diagonal of a matrix 𝐴 = 𝑎𝑖,𝑗 𝑚×𝑛

is the tuple (𝑎1,1 , 𝑎2,2 , 𝑎3,3 , ⋯ ).

1 4 7 6
4 2 5 8
7 5 3 9
6 8 9 5

47
UPPER TRIANGULAR MATRICES

∎ An upper triangular matrix: if all the entries below and to the left the main
diagonal are zeros.

∎ Equivalently, a matrix 𝐴 = 𝑎𝑖,𝑗 is an upper triangular matrix if

𝑚×𝑛

𝒂𝒊,𝒋 = 𝟎 whenever 𝒊 > 𝒋.

48
TRIANGULAR MATRICES
∎ A lower triangular matrix: if all the entries above and to the right the main
diagonal are zeros.

Equivalently, a matrix 𝐴 = 𝑎𝑖,𝑗 is an lower triangular matrix if

𝑚×𝑛

𝒂𝒊,𝒋 = 𝟎 whenever 𝐣 > 𝒊.

Definition. A matrix is said to be triangular if it is either upper triangular or lower

triangular.
Remark. Every row-echelon matrix is upper triangular.

49
TRANSPOSE OF A MATRIX
1 −3 4 1 0 8
0 2 5 −3 2 −7
8 −7 6 4 5 6

Definition. The transpose of a matrix 𝑨 is the matrix whose column 𝒊 is given by the
row 𝒊 of 𝑨.
∎ We denote by 𝐴𝑇 the transposed matrix of 𝐴.

Properties. (i) (𝑖, 𝑗)-entry of 𝐴𝑇 = (𝑗, 𝑖)-entry of 𝐴.

(ii) If 𝐴 is an 𝑚 × 𝑛 matrix then 𝐴𝑇 is an 𝑛 × 𝑚 matrix.
(iii) (𝐴𝑇 )𝑇 = 𝐴, i.e. the transpose of transpose of 𝐴 is itself.

50
EXERCISES

1 7 0
1 5 3 4
1. Let 𝐴 = 7 −1 −2 3 . Find 𝐴𝑇 . 𝐴𝑇 = 5 −1 5
3 −2 1
0 5 1 6 4 3 6

1 4 7 6
2. Let 𝐵 = 4 2 5 8 . Find 𝐵𝑇 . Answer: 𝐵𝑇 = 𝐵.
7 5 3 9
6 8 9 5

Theorem (important). 𝐴 is symmetric ⇔ 𝑨𝑻 = 𝑨.

51
MATRIX ADDITION

Definition. Let 𝐴 = 𝑎𝑖,𝑗 and 𝐵 = 𝑏𝑖,𝑗 be two matrices of same size.

𝑚×𝑛 𝑚×𝑛

The sum of 𝑨 and 𝑩 is the matrix

𝑨 + 𝑩 ∶= 𝒂𝒊,𝒋 + 𝒃𝒊,𝒋 .
𝒎×𝒏

1 3 2 −3
0 −4 + 5 2
7 5 −1 7

1+2 3−3 3 0
= 0+5 −4 + 2 = 5 −2 .
7−1 5+7 6 12

52
SCALAR MULTIPLICATION

Definition. Let 𝐴 = 𝑎𝑖,𝑗 𝑚×𝑛

be a matrix and let 𝜆 be a real number.

The scalar multiplication of 𝑨 by 𝝀 is the matrix

𝜆𝐴 ∶= 𝜆 × 𝑎𝑖,𝑗 .
𝑚×𝑛

1 5 3
Example. Suppose 𝐴 = 7 −1 −2 . Then
0 5 1

3×1 3×5 3×3 3 15 9

3𝐴 = 3 × 7 3 × (−1) 3 × (−2) = 21 −3 −6
3×0 3×5 3×1 0 15 3
53
EXERCISES

Convention. The zero matrix is the matrix whose entries are all zeros.

0 0 0
For example, 0 2×3 = . Sometimes we denote it simply by 0.
0 0 0

1 0 8 2 3 2
Exercise. Consider the matrices 𝐴 = −3 2 −7 and 𝐵 = 3 2 −2 .
4 5 6 −1 0 4
(i) Find 2𝐴, 3𝐵 and 3𝐵 − 2𝐴.
1
(ii) Find matrix 𝐶 such that 2𝐴 + 𝐶 − 𝐵 = 0.
2

54
BASIC PROPERTIES
Theorem. Let 𝐴, 𝐵 and 𝐶 be the matrices of the same size and let 𝑝, 𝑞 ∈ ℝ. One has

(1) 𝐴 + 𝐵 = 𝐵 + 𝐴.

(2) 𝐴+𝐵 +𝐶 =𝐴+ 𝐵+𝐶 .

(3) 𝑝 + 𝑞 𝐴 = 𝑝𝐴 + 𝑞𝐴.

(4) 𝑝 𝐴 + 𝐵 = 𝑝𝐴 + 𝑝𝐵.

(5) 𝑝 𝑞𝐴 = 𝑝𝑞 𝐴.

(6) 𝑝𝐴 𝑇 = 𝑝𝐴𝑇 .

(7) 𝐴+𝐵 𝑇 = 𝐴𝑇 + 𝐵𝑇 .
55
EXERCISES

1. Determine whether the following statements are true or false.

(i) For any two symmetric matrices 𝐴 and 𝐵, the sum 𝐴 + 𝐵 is symmetric as well.
Answer. This is false (trick). When 𝐴 and 𝐵 are not the same size, the sum A + B cannot
be defined. Otherwise, the statement is true.

(ii) For any two matrices of the same size 𝐴 and 𝐵,

2𝐴 − 3𝐵 𝑇 = 2𝐴𝑇 − 3𝐵𝑇 .
Answer. This is true.
2𝐴 − 3𝐵 𝑇 = (2𝐴)𝑇 + −3𝐵 𝑇
= 2𝐴𝑇 − 3𝐵𝑇 .

56
EXERCISES

1. Is it true that if 𝐴 and 𝐵 are symmetric matrices of the same size then 3𝐴 + 4𝐵
is symmetric?
Answer. Yes, it is.
Explanation: 3𝐴 + 4𝐵 𝑇 = 3𝐴 𝑇 + 4𝐵 𝑇

= 3𝐴𝑇 + 4𝐵𝑇
= 3𝐴 + 4𝐵 (since 𝐴 and 𝐵 are symmetric).
This means 3𝐴 + 4𝐵 is symmetric.

57
2. Suppose that 𝐴 is a matrix with 4 rows and such that 𝐴 = 2𝐴𝑇 . Find 𝐴.

Answer. Firstly, 𝐴 must be a square matrix of size 4 × 4.

Secondly, 𝐴 = 2𝐴𝑇 ⇒ 𝐴𝑇 = 2𝐴𝑇 𝑇 = 2 𝐴𝑇 𝑇 = 2𝐴.

So we have 𝐴 = 4𝐴. This means 𝐴 = 0.

3. Is it true that 𝐴 + 𝐴𝑇 is symmetric for every square matrix 𝐴?

Answer. Yes, it is.
Explanation: 𝐴 + 𝐴𝑇 𝑇 = 𝐴𝑇 + 𝐴𝑇 𝑇

= 𝐴𝑇 + 𝐴 = 𝐴 + 𝐴𝑇 .
This means 𝐴 + 𝐴𝑇 is symmetric.
58
EXERCISES

2.1.14, 2.1.15, 2.1.16, 2.1.17.

59
 2.2
Matrix multiplication

60
DOT PRODUCT

Definition. For two vectors (𝑎1 , 𝑎2 , ⋯ , 𝑎𝑛 ) and (𝑏1 , 𝑏2 , ⋯ , 𝑏𝑛 ), their dot product is
𝑎1 , 𝑎2 , ⋯ , 𝑎𝑛 ∙ 𝑏1 , 𝑏2 , ⋯ , 𝑏𝑛 = 𝑎1 × 𝑏1 +𝑎2 × 𝑏2 + ⋯ + 𝑎𝑛 × 𝑏𝑛 .

For example,
2, 1, 5 ∙ −1, 3, 7 = 2 × −1 + 1 × 3 + 5 × 7 = 36.

61
MULTIPLICATION MATRIX
Definition. Let 𝐴 be a matrix of size 𝑚 × 𝑛 and 𝐵 a matrix of size 𝑛 × 𝑝.
Then the product 𝐴 × 𝐵 is a matrix of size 𝑚 × 𝑝 whose 𝒊, 𝒋 -entry is the dot
product of the row 𝒊 of 𝑨 and the column 𝒋 of 𝑩:
𝑖, 𝑗 -entry = (row 𝑖 of 𝐴) ∙ (𝑐𝑜𝑙𝑢𝑚𝑛 𝑗 of 𝐵).

2 −3 𝑎 −1 10
1 4 5
Example. 5 2 × = 1 26 𝑏 .
−1 7 −2 3 0
−15 17 −5

𝑎 = 2, −3 ∙ 1, −2 = 2 × 1 + −3 × −2 = 8
𝑏 = 5,2 ∙ 5,0 = 25.
62
EXERCISES
1. Let 𝐴 be an (𝑚 × 𝑛)-matrix. When can product 𝐴 × 𝐴 be formed?
Answer. If and only if 𝑚 = 𝑛. This means that 𝐴 is a square matrix.

1 0 8
2. Consider the matrix 𝐴 = −3 2 −7 . Find 𝐴2 (i.e. 𝐴 × 𝐴).
4 6 5

1 0 8 1 0 8
−3 2 −7 × −3 2 −7 = −37 −73 .
4 6 5 4 6 5

3. Let 𝐴 be an (𝑚 × 𝑛)-matrix and 𝐵 a (𝑝 × 𝑞)-matrix. When can both 𝐴 × 𝐵 and 𝐵 × 𝐴 be

formed?
Answer. If and only if 𝑛 = 𝑝 and 𝑚 = 𝑞.
63
4. Suppose 𝐴𝐵𝐶 can be formed, 𝐴 is 3 × 3 and 𝐶 is 5 × 5. What sizes are 𝐵 and 𝐴𝐵𝐶?
Answer. 𝐵 and 𝐴𝐵𝐶 are both (3 × 5)-matrices.

5. (IMPORTANT) Let 𝐴 and 𝐵 be two matrices of size 𝑚 × 𝑚.

Is it true that 𝐴 × 𝐵 = 𝐵 × 𝐴?
Answer. No, it isn’t.

6. Find two matrices 𝐴 and 𝐵 such that both 𝐴 × 𝐵 and 𝐵 × 𝐴 can be formed and are
equal to each other.

64
EXERCISE
4 0
4 5 7 4 11
Let 𝐴 = , 𝐵 = −15 3 and 𝐶 = . Determine the size
−2 0 −3 2 −6
8 1
of (𝐴 × 𝐵) × 𝐶. Then find the (2,1)-entry of (𝐴 × 𝐵) × 𝐶.

Solution. (𝐴 × 𝐵) × 𝐶 is a (2 × 2)-matrix.
(2,1)-entry of 𝐴 × 𝐵 × 𝐶 = (Row2 of 𝐴 × 𝐵) ∙ (Column1 of 𝐶).

4 0
4 5 7
× −15 3 =
−2 0 −3 −32 −3
8 1
Result: −32 × 4 + (−3) × 2 = −134.

65
IDENTITY MATRICES

Definition. The identity matrix 𝐼𝑛 is the square matrix of size 𝒏 such that

1 if 𝑖 = 𝑗
(𝑖, 𝑗)-entry = ቊ
0 if 𝑖 ≠ 𝑗.

66
BASIC PROPERTIES MULTIPLICATION MATRIX
Suppose that 𝐴, 𝐵 and 𝐶 are matrices with suitable sizes so that the operations given

below can be formed.

1. 𝐼 × 𝐴 = 𝐴, 𝐵 × 𝐼 = 𝐵 where 𝐼 is the identity matrix of a suitable size.

2. 𝐴 × 𝐵 × 𝐶 = (𝐴 × 𝐵) × 𝐶.

3. 𝐴 × 𝐵 + 𝐶 = 𝐴 × 𝐵 + 𝐴 × 𝐶.

4. 𝐴 + 𝐵 × 𝐶 = 𝐴 × 𝐶 + 𝐵 × 𝐶.

5. 𝑘 𝐴 × 𝐵 = 𝑘𝐴 × 𝐵 for every real number 𝑘.

6. 𝐴 × 𝐵 𝑇 = 𝐵𝑇 × 𝐴𝑇 . More generally, 𝐴1 × 𝐴2 × ⋯ × 𝐴𝑛 𝑇 = 𝐴𝑇𝑛 × ⋯ × 𝐴1𝑇 .

67
EXERCISES

Convention. We usually write 𝐴𝐵 standing for the product 𝐴 × 𝐵.

1. Find two matrices 𝐴 and 𝐵 such that

(a) 𝐴𝐵 = 𝐼2 .

(b) 𝐴𝐵 = 𝐼3 .

2. Find a matrix 𝐴 of size 2 × 3 and a matrix 𝐵 of size 3 × 2 such that

(a) 𝐴𝐵 = 0.

(b) 𝐵𝐴 = 0.

68
EXERCISE

Let 𝐴 and 𝐵 be two symmetric matrices. Which of the following statements are true?
(i) 2𝐴 + 𝐵 is symmetric.
(ii) 𝐴𝐵 is symmetric.
(iii) 𝐴𝑘 is symmetric for every 𝑘 ∈ ℕ.
Solution. (i) This is true.
(ii) 𝐴𝐵 is symmetric ⇔ 𝐴𝐵 𝑇 = 𝐴𝐵.
We have 𝐴𝐵 𝑇 = 𝐵𝑇 𝐴𝑇 = 𝐵𝐴. But in general, 𝐵𝐴 ≠ 𝐴𝐵. So (ii) is false.
𝑘 𝑇
(iii) This is true because 𝐴 = 𝐴𝑇 × ⋯ × 𝐴𝑇 = 𝐴 × ⋯ × 𝐴 (𝑘 times)
= 𝐴𝑘 .

69
EXERCISES

2.3.16, 2.3.19, 2.3.20, 2.3.27.

70
BLOCK MATRIX

1 4 0 0
𝐴= 4 2 0 0
1 5 0 1
0 −1 1 5

𝐴1 𝐴2
=
𝐴3 𝐴4
where

1 4 0 0 1 5 0 1
𝐴1 = , 𝐴2 = , 𝐴3 = and 𝐴4 = are called the blocks of 𝑨.
4 2 0 0 0 −1 1 5

71
BLOCK MATRIX MULTIPLICATION
Consider the matrices

𝐼 0 𝑋 𝐼𝑋 + 0𝑌
Then we have 𝐴𝐵 = × =
𝑃 𝑄 𝑌 𝑃𝑋 + 𝑄𝑌

4 −2
𝑋 5 6
= =
𝑃𝑋 + 𝑄𝑌 3 + 27

72
BLOCK MATRIX MULTIPLICATION (cont)

𝐴1 𝑋 𝐵1 𝑌
Theorem. Suppose 𝐴 = and 𝐵 = are block matrices such that
0 𝐴2 0 𝐵2
(i) 𝐴1 and 𝐵1 are square matrices of the same size, and
(ii) 𝐴2 and 𝐵2 are also square matrices of the same size.

Then we have

𝐴1 𝑋 𝐵1 𝑌
𝐴𝐵 = ×
0 𝐴2 0 𝐵2

𝐴1 𝐵1 𝐴1 𝑌 + 𝑋𝐵2
= .
0 𝐴2 𝐵2
73
EXAMPLE
Compute 𝐴𝐵 where

1 3 2 2 1 2 1 2
𝐴= 0 1 1 1 and 𝐵 = 3 1 1 1
0 0 1 0 0 0 1 0
0 0 4 1 0 0 0 1

𝐴1 𝑋 𝐵1 𝑌
= =
0 𝐴2 0 𝐵2

10
𝐴1 𝐵1 𝐴1 𝑌 + 𝑋𝐵2 𝐴1 𝐵1 𝐴1 𝑌 + 𝑋 1
𝐴𝐵 = = =
0 𝐴2 𝐵2 0 𝐴2 0 0 1 0
0 0 4 1

74
 2.2
Matrices and systems of linear
equations

75
A REPRESENTATION
2 3 2 𝑥 1
3
Question. Consider three matrices 𝐴 = −1 2 −2 , 𝑋 = 𝑦 and 𝐵 = 4 .
0 4 0
𝑧
5 7 8 5
What does "𝐴𝑋 = 𝐵" mean?

2 3 2 2𝑥 + 3𝑦 + 2𝑧
𝑥
3
Answer. 𝐴𝑋 = −1 2 −2 × 𝑦 = 3𝑥 + 2𝑦 − 2𝑧 .
0 4 −𝑥 + 4𝑧
𝑧
5 7 8 5𝑥 + 7𝑦 + 8𝑧

So the equation 𝐴𝑋 = 𝐵 is equivalent to the system of linear of equations

2𝑥 + 3𝑦 + 2𝑧 = 1
3𝑥 + 2𝑦 − 2𝑧 = 4
−𝑥 + 4𝑧 = 0
5𝑥 + 7𝑦 + 8𝑧 = 5
76
EXERCISE

Write down the system corresponding to the equation 𝐴𝑋 = 𝐵 where

1 0 4 𝑥 10
𝐴 = 0 3 5 , X = 𝑦 and 𝐵 = −2 .
−2 4 12 𝑧 7
Then solve the obtained system using Gaussian elimination.

𝑥 + 4𝑧 = 10
Solution. ቐ 3𝑦 + 5𝑧 = −2
−2𝑥 + 4𝑦 + 12𝑧 = 7

77
BASIC SOLUTIONS
Problem. Write down and then solve the system corresponding to the equation
𝐴𝑋 = 0 where

𝑥1
𝑥2
Solution. Put 𝑋 = 𝑥 . The system that corresponds to the equation 𝐴𝑋 = 0 is
3
𝑥4

𝑥1 − 2𝑥2 + 3𝑥3 − 2𝑥4 = 0

Ι ቐ −3𝑥1 + 6𝑥2 + 𝑥3 = 0
−2𝑥1 + 4𝑥2 + 4𝑥3 − 2𝑥4 = 0

78
We carry the augmented matrix of the system into the reduced row-echelon form:

So the system Ι is equivalent to the system

1
𝑥1 − 2𝑥2 − 𝑥4 = 0
5
ቐ 3
𝑥3 − 𝑥4 = 0
5

1 3
Put 𝑥2 = 𝑠 and 𝑥4 = 𝑡, we obtain 𝑥1 = 2𝑠 + 𝑡, 𝑥3 = 𝑡.
5 5

79
 2𝑠 + 𝑡/5 2 1/5
Let us write 𝑋 = 𝑠 =𝑠 1 +𝑡 0 .
3𝑡/5 0 3/5
𝑡 0 1

2 1/5
𝑋1 = 1 and 𝑋2 = 0 are called the basic solutions of the system Ι .
0 3/5
0 1

* The solution 𝑋 = 𝑠𝑋1 + 𝑡𝑋2 is called a linear combination of 𝑋1 and 𝑋2 .

* The number of basic solutions is equal to the number of parameters.

80
EXERCISE

5 2 1
1. Consider a homogeneous system which admits 1 , 3 and 1 as basic
−4 0 0
0 −5 0
solutions. Find its (general) solution.

2. Exercise 1.3.5 (p.26, [2]).

81
A THEOREM

Let 𝐴 be a matrix of size 𝑚 × 𝑛, and consider the homogeneous system 𝐴𝑋 = 0.

Theorem. If rank 𝐴 = 𝑟 then
1. The system has exactly 𝑛 − 𝑟 basic solutions.
2. Every solution is a linear combination of these basic solutions.

82
 2.3
Matrix inverse

83
DEFINITION
Definition. Let 𝐴 be a square matrix. A matrix 𝐵 is called an inverse of 𝑨 if
𝑨𝑩 = 𝑰 and 𝑩𝑨 = 𝑰

1 1
Example. Find an inverse of the matrix 𝐴 = .
1 0
𝑥 𝑦
Solution. We need to find a matrix 𝐵 of the form 𝐵 = such that
𝑧 𝑡
1 1 𝑥 𝑦 1 0
× = and
1 0 𝑧 𝑡 0 1
𝑥 𝑦 1 1 1 0
× = .
𝑧 𝑡 1 0 0 1
0 1
And we find 𝐵 = .
1 −1
84
INVERSE MATRIX IS UNIQUE

1 1
Example. Find an inverse of the matrix 𝐴 =
0 0
Solution. 𝐴 does not have any inverse.

Theorem. If 𝐵 and 𝐶 are both inverses of 𝐴 then 𝐵 = 𝐶.

Proof. 𝐵 = 𝐼𝐵
= 𝐶𝐴 𝐵 (since 𝐶 is an inverse 𝐴)
= 𝐶(𝐴𝐵) (by the associativity)
= 𝐶𝐼 (since 𝐵 is an inverse 𝐴)
= 𝐶.
85
Convention. If the matrix 𝐴 has an inverse then we say that 𝐴 is invertible and
we will denote by 𝑨−𝟏 the inverse of 𝑨.

Exercise. Find the inverse matrix (if any) of

0 0 1
𝐴= 0 1 0
1 0 0

86
EXAMPLE

3𝑥 + 5𝑦 = 2
Example. Solve the system ቊ
2𝑥 + 𝑦 = 3

3 5 𝑥 2
Solution. The system can be rewritten as 𝐴𝑋 = 𝐵 where 𝐴 = , 𝑋 = 𝑦 and 𝐵 = .
2 1 3

−1/7 5/7
𝐴−1 = .
2/7 −3/7

Observe that
𝐴𝑋 = 𝐵 ⇒ 𝐴−1 𝐴𝑋 = 𝐴−1 𝐵 ⇒ 𝑋 = 𝐴−1 𝐵.

So we find that
𝑥 −1/7 5/7 2 13/7
𝑦 = × = .
2/7 −3/7 3 −5/7

87
A THEOREM

Theorem. Suppose 𝐴𝑋 = 𝐵 is a system of 𝑛 equations and 𝑛 variables and the

matrix 𝐴 is invertible. Then the system has a unique solution given by
𝑋 = 𝐴−1 𝐵.

88
INVERSE OF A (𝟐 × 𝟐)-MATRIX

𝑎 𝑏
Problem. Find the inverse (if any) of 𝐴 = .
𝑐 𝑑
Solution.

∎ Determinant: det 𝐴 = 𝑎𝑑 − 𝑏𝑐.

𝑑 −𝑏
∎ Adjugate of 𝐴: adj 𝐴 = .
−𝑐 𝑎

𝐴 is invertible ⟺ det 𝐴 ≠ 0. When det 𝐴 ≠ 0, we have

1 1 𝑑 −𝑏
𝐴−1 = adj 𝐴 = .
det 𝐴 𝑎𝑑−𝑏𝑐 −𝑐 𝑎

89
EXERCISES

Find the inverse (if any) for each of the following matrices:

1 2 1 3 −2 5
𝐴= 𝐵= 𝐶=
2 4 5 7 4 11

90
BASIC PROPERTIES

1. 𝐼 −1 = 𝐼.

2. 𝐴−1 −1 = 𝐴.

−1 −1
3. 𝐴𝐵 = 𝐵 −1 𝐴−1 . More generally, 𝐴1 𝐴2 ⋯ 𝐴𝑛 −1 = 𝐴−1
𝑛 ⋯ 𝐴1 .

4. 𝐴𝑇 −1
= 𝐴−1 𝑇 .

−1 1 −1
5. 𝑎𝐴 = 𝐴 for every real number 𝑎 ≠ 0.
𝑎

91
 EXERCISES

1. Consider the matrices

1 2 0 1
𝐴= and 𝐵 = .
2 3 2 0
Calculate
(a) 𝐴𝐵 −1 (b) 𝐴2 −1 (c) 𝐼 − 𝐴𝑇 𝐵 −1 .

2. Find 𝐴 when

−1 1 −1 1 −1 2 0
(a) 3𝐴 = (b) 2𝐴−1 𝑇 = (c) 𝐼 + 3𝐴 −1 = .
0 1 2 3 1 −1

92
INVERTIBLE MATRIX THEOREM

Let 𝐴 be a square matrix of size 𝑛 × 𝑛. The following conditions are equivalent.

1. 𝐴 is invertible.

2. The system of equations 𝐴𝑋 = 𝐵 has a unique solution for every column 𝐵.

3. The homogeneous system 𝐴𝑋 = 0 has a unique solution (𝑋 = 0).

4. The reduced row-echelon form of 𝐴 is 𝐼𝑛 .

5. There exists a matrix 𝐶 of size 𝑛 × 𝑛 such that 𝐴𝐶 = 𝐼𝑛 .

6. There exists a matrix 𝐶 of size 𝑛 × 𝑛 such that 𝐶𝐴 = 𝐼𝑛 .

7. Rank 𝐴 = 𝑛.
93
EXERCISES
Determine whether the following statements are true or false.
(1) Let 𝐴, 𝐵 and 𝐶 be square matrices. If 𝐴 ≠ 0 and 𝐴𝐵 = 𝐴𝐶 then 𝐵 = 𝐶.

1 1 0 1 1 0
Answer. This is false. For example, we take 𝐴 = , 𝐵= and 𝐶 = 𝐼2 = .
0 0 1 0 0 1

(2) Let 𝐴, 𝐵 and 𝐶 be square matrices. If 𝐴 is invertible and 𝐴𝐵 = 𝐴𝐶 then 𝐵 = 𝐶.

Answer. This is true. Firstly, since 𝐴 is invertible, there exists 𝐴−1 .
𝐴𝐵 = 𝐴𝐶
⇒ 𝐴−1 𝐴𝐵 = 𝐴−1 𝐴𝐶
𝐼×𝐵 =𝐼×𝐶 (since 𝐴−1 𝐴 = 𝐼)
𝐵 = 𝐶.
94
(3) If 𝐴 is a square matrix and 𝐴 ≠ 0 then 𝐴 is invertible.

1 1
Answer. This is false. For example, 𝐴 = is non-zero but not invertible.
0 0

(4) If 𝐴 is invertible then 𝑝𝐴 is invertible for every 𝑝 ∈ ℝ.

Answer. This is false. If 𝑝 ≠ 0 then it is true.

(5) If 𝐴 and 𝐵 are invertible matrices of the same size then 𝐴𝐵 is invertible.
Answer. This is true. Recall that 𝐴𝐵 −1 = 𝐵−1 𝐴−1 .

95
(6) Let 𝐴 and 𝐵 be two square matrices of the same size. If 𝐴𝐵 is invertible then both 𝐴
and 𝐵 are invertible.
Answer. This is true.
Explanation: since 𝐴𝐵 is invertible, there exists 𝐶 such that
𝐴𝐵 𝐶 = 𝐼
⇒ 𝐴 𝐵𝐶 = 𝐼.
This implies that 𝐴 is invertible (by Invertible Matrix theorem (5)). Similarly, 𝐵 is invertible
as well.

(7) If 𝐴 and 𝐵 are invertible matrices of the same size then 𝐴 + 𝐵 is invertible.
Answer. This is false. (Suppose 𝐴 is invertible. Take 𝐵 = −𝐴).

96
(8) If 𝐴 and 𝐵 are invertible matrices of the same size then 𝐴𝑇 𝐵 is invertible.
Answer. This is true.
Explanation: 𝐴 is invertible ⟺ 𝐴𝑇 is invertible.
𝐴𝑇 and 𝐵 are invertible ⇒ 𝐴𝑇 𝐵 is invertible (by part 5).

(9) If 𝐴 has a zero row then 𝐴 is not invertible.

Answer. This is true.
Explanation: If 𝐴 has a zero row then its reduced echelon form cannot be the identity
matrix 𝐼. By Invertible Matrix Theorem (4), 𝐴 is not invertible.

(10) If 𝐴 has a zero column then 𝐴 is not invertible.

Answer. This is true, just like the above one.
97
 2.4
Linear Transformations

98
SPACES ℝ AND ℝ𝟐

−2 0
ℝ

(1,2)

(0,0)
𝟐
ℝ

(−3, −2)
99
SPACE ℝ𝟑

ℝ𝟑

100
SPACE ℝ𝒏

In general, a vector 𝑋 in the space ℝn is usually written as

𝑋 = (𝑥1 , 𝑥2 , ⋯ , 𝑥𝑛 )
where 𝑥1 , 𝑥2 , … are real numbers.

𝑥1
𝑥2
We can also use the "column" notation ⋯
𝑥𝑛
or

𝑥1 𝑥2 ⋯ 𝑥𝑛 𝑇 .
101
LINEAR TRANSFORMATIONS

A linear transformation from ℝ to ℝ is a function of the form

𝑓 𝑥 = 𝑎𝑥
where 𝑎 is a constant.

For examples, 𝑓 𝑥 = 2𝑥, 𝑓 𝑥 = 3𝑥, etc.

102
LINEAR TRANSFORMATIONS (cont)

A linear transformation from ℝ2 to ℝ2 is a vector function of the form

𝑇 𝑥, 𝑦 = (𝑎𝑥 + 𝑏𝑦, 𝑐𝑥 + 𝑑𝑦).

𝑥 𝑎𝑥 + 𝑏𝑦
This can be also written as 𝑇 𝑦 = .
𝑐𝑥 + 𝑑𝑦

Example. Consider the linear transformation 𝑇: ℝ2 → ℝ2 given by

𝑥 𝑥 − 2𝑦
𝑇 𝑦 = .
2𝑥 + 𝑦

0 1
Let us calculate 𝑇 +𝑇 .
1 0
103
SOME SPECIAL LINEAR TRANSFORMATIONS IN ℝ2
1
2
𝑥 𝑥
Reflection over the 𝒙-axis: 𝑅ℎ 𝑦 = −𝑦

1 1
= 𝑅ℎ
−2 2

1 −1 1
𝑅𝑣 =
2 2 2
𝑥 −𝑥
Reflection over the 𝒚-axis: 𝑅𝑣 𝑦 = 𝑦

104
SOME SPECIAL LINEAR TRANSFORMATIONS IN ℝ2

𝑥 𝑦
Reflection over the line 𝒚 = 𝒙: 𝑇 𝑦 = .
𝑥

𝑥 −𝑦
Reflection over the line 𝒚 = −𝒙: 𝑇 𝑦 = .
−𝑥

105
SOME SPECIAL LINEAR TRANSFORMATIONS IN ℝ2
𝑥
𝑦

𝑥 𝑥
Projection on the 𝒙-axis: 𝑃ℎ 𝑦 = .
0
𝑥 𝑥
= 𝑃ℎ 𝑦
0

𝑥 𝑥
0
𝑃ℎ 𝑦 = 𝑦
𝑦
𝑥 0
Projection on the 𝒚-axis: 𝑃𝑣 𝑦 = .
𝑦

106
SOME SPECIAL LINEAR TRANSFORMATIONS ℝ2 → ℝ2 (cont)

𝑥 𝑐𝑜𝑠𝜃. 𝑥 − 𝑠𝑖𝑛𝜃. 𝑦
The rotation by 𝜽: 𝑅𝜽 𝑦 = sin𝜃. 𝑥 + 𝑐𝑜𝑠𝜃. 𝑦 .
𝑥
𝑅𝜽 𝑦

𝑥
𝜃 𝑦

Counterclockwise rotation about the origin through the angle 𝜃.

107
LINEAR TRANSFORMATIONS (cont)

A linear transformation ℝ3 → ℝ2 is a vector function of the form

𝑥1
𝑎1,1 𝑥1 + 𝑎1,2 𝑥2 + 𝑎1,3 𝑥3
𝑇 𝑥2 = 𝑎 𝑥 + 𝑎 𝑥 + 𝑎 𝑥 .
𝑥3 2,1 1 2,2 2 2,3 3

And similarly, a linear transformation ℝ3 → ℝ3 is a vector function of the form

𝑥1 𝑎1,1 𝑥1 + 𝑎1,2 𝑥2 + 𝑎1,3 𝑥3

𝑇 𝑥2 = 𝑎2,1 𝑥1 + 𝑎2,2 𝑥2 + 𝑎2,3 𝑥3 .
𝑥3 𝑎3,1 𝑥1 + 𝑎3,2 𝑥2 + 𝑎3,3 𝑥3

108
A THEOREM
Theorem. If 𝑇 ∶ ℝ𝑚 → ℝ𝑛 is a linear transformation then
𝑇 𝑎𝑋 + 𝑏𝑌 = 𝑎𝑇 𝑋 + 𝑏𝑇(𝑌)
for every 𝑋, 𝑌 ∈ ℝ𝑚 and 𝑎, 𝑏 ∈ ℝ.
1 2 0 −2
Example. Let 𝑇 ∶ ℝ2 → ℝ2 be a linear transformation such that 𝑇 = and 𝑇 = .
0 1 1 3
4
Let us compute 𝑇 .
5
4 1 0
Solution. We see that =4 + 5 . So we have
5 0 1
4 1 0
𝑇 = 4𝑇 + 5𝑇
5 0 1
2 −2 −2
=4 +5 = .
1 3 19
109
MATRIX OF A LINEAR TRANSFORMATION

Definition. Let 𝑇 ∶ ℝ𝑚 → ℝ𝑛 be a linear transformation. The matrix of 𝑻 is

𝐴 𝑇 = 𝑇 𝑒1 𝑇 𝑒2 ⋯ | 𝑇 𝑒𝑚

where 𝑒1 = [1 0 ⋯ 0]T , 𝑒2 = [0 1 ⋯ 0]T , ⋯ , 𝑒𝑚 = [0 0 ⋯ 1]T .

Remark. 𝐴 𝑇 is a matrix of size 𝑛 × 𝑚.

110
EXAMPLE

Find the matrix of the linear transformation 𝑇 ∶ ℝ3 → ℝ2 given by

𝑥1
𝑥 2𝑥1 + 𝑥2 − 𝑥3
𝑇 2 = .
𝑥3 3𝑥 1 − 𝑥2
Solution. We have

1 0 0
2 1 −1
𝑇 𝑒1 =𝑇 0 = 𝑇 𝑒2 =𝑇 1 = 𝑇 𝑒3 =𝑇 0 = .
3 −1 0
0 0 1

2 1 −1
𝐴𝑇 = .
3 −1 0

111
EXAMPLE

Find the matrix of the linear transformation 𝑇 ∶ ℝ3 → ℝ3 given by

𝑥1 𝑎1,1 𝑥1 + 𝑎1,2 𝑥2 + 𝑎1,3 𝑥3

𝑇 𝑥2 = 𝑎2,1 𝑥1 + 𝑎2,2 𝑥2 + 𝑎2,3 𝑥3 .
𝑥3 𝑎3,1 𝑥1 + 𝑎3,2 𝑥2 + 𝑎3,3 𝑥3
Solution.
𝑎1,1 𝑎1,2 𝑎1,3
𝐴 𝑇 = 𝑎2,1 𝑎2,2 𝑎2,3 .
𝑎3,1 𝑎3,2 𝑎3,3

𝑥1 𝑥1
* Note that 𝐴 𝑇 × 𝑥2 = 𝑇 𝑥2 .
𝑥3 𝑥3

112
EXAMPLE (important)
2 1 1 −1
Let 𝑇 ∶ ℝ2 → ℝ2 be a linear transformation such that 𝑇 = and 𝑇 = .
1 3 1 1
Find 𝐴 𝑇 .
1 2 1 2𝑎 + 𝑏 = 1
Solution. First, solve the equation =𝑎 +𝑏 ⇔ ቊ
0 1 1 𝑎+𝑏 =0
We find 𝑎 = 1 and 𝑏 = −1.

0 2 1
Similarly, solving the equation = 𝑎′ + 𝑏′ , we find 𝑎′ = −1 and 𝑏 ′ = 2.
1 1 1
1 2 1 2 1 2
So we have 𝑇 𝑒1 = 𝑇 =𝑇 − =𝑇 −𝑇 = ,
0 1 1 1 1 2
0 2 1 2 1 −3
𝑇 𝑒2 = 𝑇 =𝑇 − +2 = −𝑇 + 2𝑇 = .
1 1 1 1 1 −1
2 −3
Therefore, 𝐴 𝑇 = .
2 −1
113
MATRIX OF A LINEAR TRANSFORMATION (cont)

Theorem. If 𝑇 ∶ ℝ𝑚 → ℝ𝑛 is a linear transformation and 𝐴 𝑇 is the matrix of 𝑇 then

𝑇 𝑋 = 𝐴𝑇 𝑋
for every 𝑋 ∈ ℝ𝑚 .

Example (important). Suppose that the matrix of a linear transformation 𝑇 ∶ ℝ3 → ℝ2 is

2 1
4 −1 0
𝐴𝑇 = . Find 𝑇 1 and 𝑇 5 .
1 4 3
0 2

2 2
4 −1 0 7
Solution. 𝑇 1 = × 1 = .
1 4 3 6
0 0

114
COMPOSITION OF TWO LINEAR TRANSFORMATIONS

Definition. Let 𝐹 ∶ ℝ𝑚 → ℝ𝑛 and 𝑇 ∶ ℝ𝑛 → ℝ𝑝 be two linear transformations.

The composition 𝑇 ∘ 𝐹: ℝ𝑚 → ℝ𝑝 is the linear transformation given by
𝑇∘𝐹 𝑋 =𝑇 𝐹 𝑋
for every 𝑋 ∈ ℝ𝑚 .

Theorem. If 𝐴𝐹 is the matrix of 𝐹 and 𝐴 𝑇 is the matrix of 𝑇, then the matrix of 𝑇 ∘ 𝐹 is

𝐴 𝑇∘𝐹 = 𝐴 𝑇 𝐴𝐹 .

115
INVERSE OF A LINEAR TRANSFORMATION

Definition. An inverse (if any) of a linear transformation 𝑇 ∶ ℝ𝑛 → ℝ𝑛 is a linear

transformation 𝑈 ∶ ℝ𝑛 → ℝ𝑛 such that
𝑈 𝑇 𝑋 =𝑋=𝑇 𝑈 𝑋
for every 𝑋 ∈ ℝ𝑛 .

2 2
𝑥1 𝑥1 + 𝑥2 2 2
𝑥1 𝑥2
Example. 𝑇 ∶ ℝ → ℝ , 𝑇 𝑥 = and 𝑈: ℝ → ℝ , 𝑈 𝑥 = 𝑥 − 𝑥 .
2 𝑥1 2 1 2

Show that 𝑈 and 𝑇 are inverses of each other.

116
Convention. When 𝑇 has an inverse, we will say that 𝑇 is invertible and denote its
inverse by 𝑇 −1 .

Theorem. A linear transformation 𝑇 ∶ ℝ𝑛 → ℝ𝑛 is invertible iff its matrix 𝐴 𝑇 is

invertible and then,
𝑨𝑻−𝟏 = (𝑨𝑻 )−𝟏 .

117
EXAMPLE (important)

Consider the linear transformations 𝑇 ∶ ℝ3 → ℝ2 and 𝐹 ∶ ℝ2 → ℝ3 provided that

1 −1
0 2 1
𝐴𝑇 = and 𝐴𝐹 = 2 0 .
−1 3 2
4 2
(a) Determine the matrices of 𝐹 ∘ 𝑇 and 𝑇 ∘ 𝐹.

−1
4
(b) Calculate (𝐹 ∘ 𝑇) 3 and (𝑇 ∘ 𝐹) .
5
4
(c) Find the matrix of 𝑇 ∘ 𝐹 −1 .

−1 2
(d) Calculate 𝑇 ∘ 𝐹 .
3

118
 1 −1 1 −1 −1
0 2 1
(a) 𝐴𝐹∘𝑇 = 𝐴𝐹 𝐴 𝑇 = 2 0 × = 0 4 2 .
−1 3 2
4 2 −2 14 8

1 −1
0 2 1 8 2
𝐴 𝑇 ∘𝐹 = 𝐴 𝑇 𝐴𝐹 = × 2 0 = .
−1 3 2 13 5
4 2

−1 −1 1 −1 −1 −1 −8
(b) 𝐹 ∘ 𝑇 3 = 𝐴𝐹∘𝑇 3 = 0 4 2 × 3 = 20 .
4 4 −2 14 8 4 76

4 4 8 2 4 42
𝑇∘𝐹 = 𝐴 𝑇∘𝐹 = × = .
5 5 13 5 5 77

−1 5/14 −2/14
8 2
(c) 𝐴 𝑇∘𝐹 −1 = (𝐴 𝑇 ∘𝐹 )−1 = = .
13 5 −13/14 8/14

119
MATRICES OF SOME SPECIAL LINEAR TRANSFORMATIONS
𝑥 𝑥 1 0
1. Reflection over the 𝒙-axis: 𝑅ℎ 𝑦 = −𝑦 . A 𝑅ℎ = .
0 −1

𝑥 −𝑥 −1 0
2. Reflection over the 𝒚-axis: 𝑅𝑣 𝑦 = 𝑦 . A𝑅𝑣 = .
0 1

𝑥 𝑥 1 0
3. Projection on the 𝒙-axis: 𝑃ℎ 𝑦 = . A𝑃ℎ = .
0 0 0

𝑥 0 0 0
4. Projection on the 𝒚-axis: 𝑃𝑣 𝑦 = . A𝑃𝑣 = .
𝑦 0 1
120
MATRICES OF SOME SPECIAL LINEAR TRANSFORMATIONS (cont)
𝑥 𝑦 0 1
5. Reflection over the line 𝒚 = 𝒙: 𝑇 𝑦 = . Matrix: .
𝑥 1 0

𝑥 −𝑦 0 −1
6. Reflection over the line 𝒚 = −𝒙: 𝑇 𝑦 = . Matrix: .
−𝑥 −1 0

𝑥 𝑐𝑜𝑠𝜃. 𝑥 − 𝑠𝑖𝑛𝜃. 𝑦
7. The rotation by 𝜽: 𝑅𝜽 𝑦 = .
sin𝜃. 𝑥 + 𝑐𝑜𝑠𝜃. 𝑦

𝜋 0 −1 𝜋 2/2 − 2/2
𝜃= : . 𝜃= : .
2 1 0 4 2/2 2/2
121
PRACTICE
1. Find conditions on 𝑎 and 𝑏 such that the following system
−𝑥 + 3𝑦 + 2𝑧 = −8
ቐ 𝑥+𝑧 =2
3𝑥 + 3𝑦 + 𝑎𝑧 = 𝑏
(i) has no solution, (ii) has infinitely many solutions, (iii) has only one solution.
Solution. We carry the augmented matrix (closed) to the echelon form.

−1 3 2 −8 Row2 ↔ Row1 1 0 1 2 Row2 + Row1 1 0 1 2

1 0 1 2
→ −1 3 2 −8 → 0 3 3 −6
3 3 𝑎 𝑏 3 3 𝑎 𝑏 Row3 −3 ×Row1 0 3 𝑎−3 𝑏−6
1
3
×Row2 1 0 1 2 Row3 −3 ×Row2 1 0 1 2
0 1 1 −2 0 1 1 −2
→ 0 3 𝑎−3 𝑏−6
→ 0 0 𝑎−6 𝑏
122
 1 0 1 2
0 1 1 −2
0 0 𝑎−6 𝑏

So the original system is equivalent to the system

𝑥+𝑧 =2
ቐ 𝑦 + 𝑧 = −2
(𝑎 − 6)𝑧 = 𝑏
(i) The system has no solution ⇔ 𝑎 = 6 and 𝑏 ≠ 0.
(ii) The system has infinitely many solutions ⇔ 𝑎 = 6 and 𝑏 = 0.
(iii) Otherwise (𝑎 ≠ 6), the system has only one solution.

123
2. Find condition on 𝑎 such that the homogeneous sys tem
𝑥 − 2𝑦 + 𝑧 = 0 (1)
ቐ𝑥 − 𝑦 + 3𝑧 = 0 (2)
2𝑥 + 𝑎𝑦 + 4𝑧 = 0
has only trivial solution (𝑥 = 𝑦 = 𝑧 = 0).
Note: A homogeneous system has either only trivial solution or infinitely many solutions.

𝑥 − 2𝑦 + 𝑧 = 0
Solution. Taking (1) + (2), the system is equivalent to ቐ2𝑥 − 3𝑦 + 4𝑧 = 0
2𝑥 + 𝑎𝑦 + 4𝑧 = 0
Observe that the system has infinitely many solutions if and only if 𝑎 = −3.
Thus, the system has only trivial solution if and only if 𝑎 ≠ −3.

124
 1 1 2
3. Let 𝑇 ∶ ℝ2 → ℝ2 be the matrix transformation induced by 𝐴 = . Find (𝑇 ∘ 𝑇) .
0 1 2

1 1
Note: "𝑇 is the matrix transformation induced by 𝐴 = "
0 1
𝑚𝑒𝑎𝑛𝑠
1 1
"𝑇 is the linear transformation whose matrix is 𝐴 = ".
0 1
Solution. The matrix of 𝑇 ∘ 𝑇 is

1 1 1 1 1 2
𝐴×𝐴= × = .
0 1 0 1 0 1

2 1 2 2 6
𝑇∘𝑇 = × = .
2 0 1 2 2

125
 2 1 −1 4
4. Let 𝑇 ∶ ℝ2 → ℝ2 be a linear transformation such that 𝑇 = and 𝑇 = .
1 7 1 3
4
Calculate 𝑇 .
1

4 2 −1
Solution. Solving the equation =𝑎 +𝑏
1 1 1
2𝑎 − 𝑏 = 4
⇔ቊ
𝑎+𝑏 =1
5 2
We find 𝑎 = and 𝑏 = − .
3 3
So we have

4 5 2 2 −1 −1
𝑇 = 𝑇 − 𝑇 = .
1 3 1 3 1 29/3

126
 𝑥 2
Solution. 𝑇 𝑦 =
3
1 1 𝑥 2
⇔ × 𝑦 =
0 1 3
𝑥+𝑦 2
⇔ 𝑦 =
3
⇔ 𝑥 = −1 and 𝑦 = 3.

127