Chapter 2: Matrices - Determinants - Systems of linear equations

Lecturer: Assoc. Prof. Nguyễn Duy Tân

email: [email protected]

Falculty of Mathematics and Informatics, HUST

October 2024

1 / 99
 Contents

Contents
1 2.1. Matrices
2.1.1. Definitions
2.1.2. Operations with matrices
2 2.2. Determinants
2.2.1. Definition
2.2.2. Some properties of determinants
2.2.3. Evaluation of a determinant using elementary row operations
3 2.3. Rank of a matrix
2.3.1. Rank of a matrix
2.3.2. Evaluation of the rank of a matrix
2.3.3. Inverse of a matrix
2.3.4. Finding the inverse of a matrix
4 2.4. Systems of linear equations
2.4.1. Definition
2.4.2. Cramer’s system
2.4.3. General system of linear equations
2.4.4. Homogeneous systems
2 / 99
 2.1. Matrices 2.1.1. Definitions

2.1.1. Definitions
Let K be the field of real numbers or the field of complex numbers.
A matrix (over K ) of size m × n is a rectangular array of numbers (in K ), which has m row and n
column:  
a11 a12 . . . a1n
 a21 a22 . . . a2n 
A= . .. .. 
 
 .. ..
. . . 
am1 am2 ... amn
vhere aij ∈ K (for i = 1, . . . , m, j = 1, . . . , n).
If m = n then A is called a square matrix (of ordern). The entries a11 , a22 , . . . , ann đare call
diagonal entries. Diagonal entries form the main diagonal of A.
Matrices are usually written in square brackets as above, or parentheses, and may be abberivated
by writing only single generic term, such as A = [aij ]m×n or A = (aij )m×n .
The set of matrices of size m × n with entries in K is denoted by Mm×n (K ), or Mm,n (K ), or
M(m × n, K ). In the case m = n, we also use the notation Mn (K ) to denote the set of square
matrices of order n (with entries in K ).
3 / 99
 2.1. Matrices 2.1.1. Definitions

Some special matrices

A matrix of size1 × n is called a row matrix.

A matrix of size m × 1 is called a column matrix.
A matrix A = [aij ]m×n where all entries aij = 0 (∀i, j) is called the zero matrix, usually denoted by
O, or Om×n .  
0 0 ... 0
0 0 . . . 0
O = . . . .. 
 
 .. .. .. .
0 0 ... 0

4 / 99
 2.1. Matrices 2.1.1. Definitions

A square matrix A = [aij ]n×n is called an upper triangular matrix if aij = 0, for all i > j.
 
a11 a12 . . . a1n
0 a22 . . . a2n 
A= . .. .. 
 
 .. ..
. . . 
0 0 ... ann

A square matrix A = [aij ]n×n is called a lower triangular matrix if aij = 0, for all i < j.
 
a11 0 . . . 0
a21 a22 . . . 0 
A= . .. .. 
 
 .. ..
. . . 
an1 an2 . . . ann

5 / 99
 2.1. Matrices 2.1.1. Definitions

A square matrix A = [aij ]n×n is called a diagonal matrix if aij = 0, for all i ̸= j.
 
a11 0 . . . 0
 0 a22 . . . 0 
A= . .. .. 
 
 .. ..
. . . 
0 0 ... ann

A square matrix A = [aij ]n×n is called the ịdentity matrix (of order n) if it is a diagonal matrix and
aii = 1, for all i. The identity matrix of order n is usually denoted by In (or I ), or En (or E ).
 
1 0 ... 0
0 1 . . . 0
In =  . . . .. 
 
 .. .. .. .
0 0 ... 1

6 / 99
 2.1. Matrices 2.1.1. Definitions

Equality of matrices

Defintion
Two matrices A = [aij ]m×n and B = [bij ]p×q are equal, written A = B if
they have the same size:: m = p and n = q;
aij = bij for all i = 1, . . . , m, j = 1, . . . , n.

7 / 99
 2.1. Matrices 2.1.2. Operations with matrices

2.1.2. Operations with matrices

Phép cộng ma trận
Consider two matrices of the same size m × n: A = [aij ]m×n , B = [bij ]m×n . Their sum A + B is the
m × n matrix given by:
A + B = [aij + bij ]m×n .
The sume of two matrices of different size is undefined.
To add two matrices of same size, we add their corresponding entries.
Example:
−1 −3 −5
ï ò ï ò ï ò
1 1 −2 2 3 0
+ = .
2 1 −1 0 1 2 2 2 1
Example:
 
ï ò 1
1 2
+ 2
3 4
3
is undefined.
8 / 99
 2.1. Matrices 2.1.2. Operations with matrices

For a matrix A = [aij ]m×n , the negative of A, writtend as −A, is defined by −A = [−aij ]m×n .
We also define A − B = A + (−B).

Properties
On the set ofm × n (over K ), we have
(A + B) + C = A + (B + C ),
A + O = O + A = A,
A + (−A) = (−A) + A = O,
A + B = B + A.

In other words, the set Mm×n (K ) together with matrix addition is an abelian group.

9 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Scalar multiplication

Definition
The product of a number (scalar) k and an m × n matrix A = [aij ]m×n is the matrix kA of size m × n,
which is given by
kA = [kaij ]m×n .

To multiply a matrix A by a scalar k, we multiplying each entry in A by k.

Example: ï ò ï ò
1 2 3 2 4 6
2 = .
4 5 6 8 10 12

Remark: We hve (−1)A = −A.

10 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Properties

Properties
Let A, B be matrices in Mm×n (K ) and c, d ∈ K . We have
(cd)A = c(dA),
1A = A,
c(A + B) = cA + cB,
(c + d)A = cA + dA.

Extra property: If A lis a matrix of size m × n and O is the zero matrix of size m × n then
ñ
c =0
cA = O ⇔ .
A=O

11 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Matrix multplication

Definition
Let A = [aij ]m×n be a matrix of sizem × n and B = [bij ]n×p a matrix of size n × p. The product AB is
the matrix C = [cij ]m×p of size m × p given by
n
X
cij = ai1 b1j + ai2 b2j + · · · + ain bnj = aik bkj (∀i = 1, . . . , m, j = 1, . . . , p).
k=1

12 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Remarks

The product AB is defined only in the case that the number of columns of A is equal to the
number of rows of B.
To obtain the entry cij of the product AB we multiply the entries in the ith row of A by the
corresponding entries in the jth column of B and then add the results.
1 b1j

  : b



 2j

  .
ai1 ai2 . . . ain . =⇒ cij = ai1 b1j + ai2 b2j + · · · + ain bnj
XX .
the ith row of A X
z
X
bnj
the jth column of B

he product AB could be defined meanwhile the product BA is not defined. Even in the case that
both AB and BA are defined, in general we still have AB ̸= BA.

13 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Remarks

In general, AB = O does not imply that A = O or B = O.

In general, AC = BC (or CA = CB) and C ̸= O do not imply that A = B.

Example:
ï ò ï ò
1 1 1 −2
A= ,B= : A ̸= O, B ̸= O but AB = O.
2 2 −1 2
ï ò ï ò ï ò
1 0 0 −1 1 −2
A= ,B= ,C= :
2 0 0 −2 −1 2
ï ò
1 −2
AC = BC = but A ̸= B.
2 −4

14 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Example
 
1 2
−1
ï ò
1 2
Let A = , B = 2 −1. Compute AB.
0 1 −2
3 1

ï ò
c c12
C = AB is of size 2 × 2, C = 11 .
c21 c22
c11 = 1 · 1 + (−1) · 2 + 2 · 3 = 5.
c12 = 1 · 2 + (−1) · (−1) + 2 · 1 = 5.
c21 = 0 · 1 + 1 · 2 + (−2) · 3 = −4.
c22 = 0 · 2 + 1 · (−1) + (−2) · 1 = −3.
ï ò
5 5
AB = C = .
−4 −3

15 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Properties
Let A, B, C be matrices (with sizes such the given operations are defined) and c ∈ K . Then we have
following properties.
(AB)C = A(BC )
A(B + C ) = AB + AC , (B + C )A = BA + CA
(cA)B = A(cB) = c(AB)
If A is of size m × n then AIn = A and Im A = A.

Remark: The set Mn (K ) of square matrices of order n together with matrix addition and multiplication
is a ring (with unit).

16 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Powers of a matrix

Let A be a square matrix of order n.

For k ≥ 1 being a positive interger, we define

Ak = A
| · A{z· · · A} .
k lần

Properties: Ak+l = Ak Al , Akl = (Ak )l , for all positive integers k, l.

If f (x) = ak x k + · · · + a1 x + a0 is a polynomial of degree k, we define

f (A) = ak Ak + · · · + a1 A + a0 In .

17 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Example (GK20161)
 
−1 −1 0
Let A =  0 −1 2  and P(x) = x 2 + 2x + 1. Compute P(A).
1 −1 −1

Solution 1:    
−1 −1 0 −1 −1 0 1 2 −2
A2 =  0 −1 2  0 −1 2  =  2 −1 −4.
1 −1 −1 1 −1 −1 −2 1 −1
       
1 2 −2 −1 −1 0 1 0 0 0 0 −2
P(A) = A2 + 2A + I3 = 2 −1 −4 + 2 0 −1 2  + 0 1 0 = 2 −2 0 .
0 1 −1 1 −1 −1 0 0 1 2 −1 −2
    
0 −1 0 0 −1 0 0 0 −2
Solution 2: P(A) = A2 + 2A + I3 = (A + I3 )2 = 0 0 20 0 2 = 2 −2 0 .
1 −1 0 1 −1 0 2 −1 −2

18 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Example (GK20161)
i −1
ï ò
Cho A = . Compute A3 and A27 .
1 i

−1 i −1 −2 −2i −1
ï òï ò ï ò ï ò
2 i i
A = = = 2i = (2i)A.
1 i 1 i 2i −2 1 i
−4i
ï ò
3 2 4
A = (2i)A · A = (2i)A = (2i) · (2i)A = −4A = .
−4 −4i
By induction on k: Ak = (2i)k−1 A, for every natural number k ≥ 1.
A27 = (2i)26 A = −226 A.

19 / 99
 2.1. Matrices 2.1.2. Operations with matrices

The transpose of a matrix

The transpose of a matrix

Let A = [aij ]m×n be an m × n matrix. The transpose of A, denoted by AT = [bij ] is the matrix of size
n × m given by
bij = aji , ∀i = 1, . . . n, j = 1, . . . , m.

Thus, the columns of AT are the rows of A, the rows of AT are the columns of A.
 
ï ò 1 4
1 2 3
Example: If A = then AT = 2 5.
4 5 6
3 6

20 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Symmetric and skew-symmetric matrices

Definition
A matrix A is said to be symmetric if AT = A.
A matrix A is said to be skew-symmetric if AT = −A.

Thus
Matrix A = [aij ] is symmetric if and only if A is a square matrix and aij = aji , ∀i, j.
Matrix A = [aij ]is skew-symmetric if and only if A is a square matrix and aij = −aji , ∀i, j.
   
1 2 3 0 2 3
Example: Matrix 2 5 6 is symmetric, and −2 0 6 is skew-symmetric.
3 6 4 −3 −6 0

21 / 99
 2.1. Matrices 2.1.2. Operations with matrices

Properties

Let A, B be matrices (of suitable sizes) and c a scalar. We have:

1 (AT )T = A,
2 (A + B)T = AT + B T ,
3 (cA)T = cAT ,
4 (AB)T = B T AT ,
5 AAT and AT Aare symmetric, for every matrix A of arbitary size m × n.

22 / 99
 2.2. Determinants 2.2.1. Definition

2.2.1. Definition

Let A = [aij ]n×n be a square matrix of order n. We shall define recursively the determinant of A,
denoted by det(A) or |A|.

The determinant of a square matrix of order 1 or 2

If A = [a11 ] is a square matrix of order 1, then det(A) = a11 .
ï ò
a11 a12
If A = then
a21 a22
a a12
det(A) = 11 = a11 a22 − a12 a21 .
a21 a22

23 / 99
 2.2. Determinants 2.2.1. Definition

Determinant of square matrices of order n ≥ 3

Suppose we have defined the determinant of any square matrix of order n − 1.

Consider a square matrix A = [aij ]n×n of order n.
For each i, j, let Mij be the matrix obtained from A by deleting the ith row and the jth column.
The matrix Mij is a square matrix of order n − 1.
Let Cij = (−1)i+j det(Mij ), and Cij is called the cofactor of aij (det(Mij ) is called the minor of aij ).

Definition
The determinant ofA = [aij ]n×n is

det(A) = |A| = a11 C11 + a12 C12 + · · · + a1n C1n .

24 / 99
 2.2. Determinants 2.2.1. Definition

Example
 
1 2 −1
Evaluate the determinant of A = 2 −1 2 .
3 1 2

ï ò
−1 2 −1 2
M11 = ⇒ C11 = + det(M11 ) = = −4.
1 2 1 2
ï ò
2 2 2 2
M12 = ⇒ C12 = − det(M12 ) = − = −(−2) = 2.
3 2 3 2
ï ò
2 −1 2 −1
M13 = ⇒ C13 = + det(M13 ) = = 5.
3 1 3 1
|A| = a11 C11 + a12 C12 + a13 C13 = 1 · (−4) + 2 · 2 + (−1) · 5 = −5.
In short, we have:
|A| = a11 det(M11 ) − a12 det(M12 ) + a13 det(M13 )
−1 2 2 2 2 −1
=1· −2· + (−1) ·
1 2 3 2 3 1
= 1 · (−4) + 2 · 2 + (−1) · 5 = −5.
25 / 99
 2.2. Determinants 2.2.1. Definition

The Laplace expansion

The Laplace expansion

Let A = [aij ] be a square matrix of order n. For any i and j, we have:
The Laplace expansion along the ith row:
n
X
det(A) = aij Cij = ai1 Ci1 + ai2 Ci2 + · · · + ain Cin .
j=1

The Laplace expansion along the jth column:

n
X
det(A) = aij Cij = a1j C1j + a2j C2j + · · · + anj Cnj .
i=1

Remark: We usually use the Laplace expansion along a row or column which has many zeroes.

26 / 99
 2.2. Determinants 2.2.1. Definition

Example
 
1 2 −1
Consider the matrix A = 2 −1 2 .
3 1 2
y the Laplace expansion along the 2nd row, we have
|A| = −a21 det(M21 ) + a22 det(M22 ) − a23 det(M23 )
2 −1 1 −1 1 2
= (−2) · + (−1) · −2·
1 2 3 2 3 1
= (−2) · 5 + (−1) · 5 − 2 · (−5) = −5.
y the Laplace expansion along the 3rd column, we have
|A| = a13 det(M13 ) − a23 det(M23 ) + a33 det(M33 )
2 −1 1 2 1 2
= (−1) · −2· +2·
3 1 3 1 2 −1
= (−1) · 5 − 2 · (−5) + 2 · (−5) = −5.

27 / 99
 2.2. Determinants 2.2.1. Definition

The determinant of a 3 × 3 matrix

he determinant of a  
a11 a12 a13
A =  a21 a22 a23  ?
a31 a32 a33

Method 1: Expansion along a row or column.

Method 2:
a11 a12 a13 a11 a12
a21 a22 a23 a21 a22
a31 a32 a33 a31 a32

det(A) = a11 a22 a33 + a12 a23 a31 + a13 a21 a32
− a31 a22 a13 − a32 a23 a11 − a33 a21 a12 .

28 / 99
 2.2. Determinants 2.2.1. Definition

Example

 
1 2 −1
Evaluate the determinant of A = 2 −1 2 .
3 1 2

1 2 −1 1 2
2 −1 2 2 −1
3 1 2 3 1
|A| = 1 · (−1) · 2 + 2 · 2 · 3 + (−1) · 2 · 1 − 3 · (−1) · (−1) − 1 · 2 · 1 − 2 · 2 · 2
= (−2) + 12 + (−2) − 3 − 2 − 8
= −5.

29 / 99
 2.2. Determinants 2.2.1. Definition

Some exercises

1 1 1
(GK20161) Find x such that 2 x −3 = 0.
4 x2 9
3−x 2 2
(GK20171) b) Solve for 2 3−x 2 = 0.
2 2 3−x
 
2 −1 1
(GK20191) Let A = 1 0 3. Find λ ∈ R such that det(A − λE ) = 0, where E the identity
0 1 1
matrix of order 3.
1 1 1
(GK20201) Find a condition on a, b, c to ensure that a b c = 0.
a2 b2 c2

30 / 99
 2.2. Determinants 2.2.1. Definition

The determinant of a triangular matrix

Consider triangular matrices

   
a11 a12 ... a1n b11 0 ... 0
 0 a22 ... a2n   b21 b22 ... 0 
A= . .. .. , B= .. .. .. .
   
 .. .. ..
. . .   . . . . 
0 0 ... ann bn1 bn2 ... bnn

We have
det(A) = a11 a22 . . . ann , det(B) = b11 b22 . . . bnn .

Particular case: This formular can be applied for diagonal matrices. In particular, det(In ) = 1.

31 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

2.2.2. Tính chất cơ bản

Determinant of transpose

det(AT ) = det(A)
.
Remark: In the following, we only state properties of determinants in terms of "rows". But these
properties still holds true if we replace "rows" by "column".

32 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

Interchange two rows

If we interchange two rows of A to obtain B then det(B) = − det(A).

Example:
1 2 3 4 5 6
4 5 6 =−1 2 3 (H2 ↔ H1 )
7 8 10 7 8 10

Corollary
If A has two equal rows then det(A) = 0.

33 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

Multiply a row by a scalar

If B is obtained from A by multiplying a row of A by a scalar k, thendet(B) = k det(A).

Example:
2 4 6 1 2 3 (H1 ← 12 H1 )
4 5 6 =24 5 6
7 8 10 7 8 10

Corollary
If one row of A is a multiple of another row then det(A) = 0.
If A has a zero row then det(A) = 0.
If A is a square matrix of order n and k is a scalar then det(kA) = k n det(A).

34 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

Property
If a square matrix A = [aij ]n×n has some ith row such that aij = bj + cj (for all j = 1, . . . , n) then
det(A) the sum of two determinants

a11 a12 ··· a1n a11 a12 ··· a1n a11 a12 ··· a1n
.. .. .. .. .. .. .. .. ..
. . . . . . . . .
b1 + c1 b2 + c 2 ··· bn + cn = b1 b2 ··· bn + c1 c2 ··· cn
.. .. .. .. .. .. .. .. ..
. . . . . . . . .
an1 an2 ··· ann an1 an2 ··· ann an1 an2 ··· ann

Example:
1 2 3 1 2 3 1 2 3 1 2 3
4 5 6 = 1+3 (−1) + 6 2+4 = 1 −1 2 + 3 6 4
7 8 10 7 8 10 7 8 10 7 8 10

35 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

Corollary
If B is obtained from A by adding a multiple of a row A to another row of A then det(B) = det(A).

Example:
1 2 3 1 2 3
4 5 6 = 6 9 12 (H2 ← H2 + 2H1)
7 8 10 7 8 10

1 2 3 1 2 3
4 5 6 = 4 5 6
7 8 10 4 2 1 (H3 ← H3 − 3H1)

36 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

The determiant of the product

det(AB) = det(A) det(B).

37 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

Example (GK20201-N3)
Show there is no real square matrix A of order 2019 such that A2020 + E = O, where E is the identity
matrix of order 2019.

Suppose there is a real matrix A of order 2019 such that A2020 + E = O. Then A2020 = −E .
Hence det(A2020 ) = det(A)2020 = det(−E ) = (−1)2019 det(E ) = −1.
This is impossible since. . . det(A) ∈ R and hence det(A)2020 ≥ 0.

38 / 99
 2.2. Determinants 2.2.2. Some properties of determinants

Some exercises*
(GK20213) Let A, B be real square matrices of order 2023. Show that there is no real matrix X
such that (AX )T B 2022 XA + 3E = O.
(GK20191) Let A, B be two square matrices of the same orders such that A2019 = 0 and
AB = A + B. Show thadet(B) = 0. [Hint: det(A) = 0 and A(B − I ) = B, where I is the identity
matrix.]
(GK20181) Let A ̸= O and n ∈ N, n ≥ 2 such that An = O. Show that det(A − E ) ̸= 0, where E
the identity matrix. [Hint: (A − E )(An−1 + · · · + A + E ) = An − E .]
(GK20181) Let A be a square matrix and λ ∈ R such that det(A − λE ) = 0, where E is the
identity matrix. Show that
det(A2 + 2A − (λ2 + 2λ)E ) = 0.
(GK2017-N1) Let A, B be real square matrices of order n, n ≥ 2, such that AB = BA. Show that
det(A2 + B 2 ) ≥ 0. [Hint: since AB = BA, A2 + B 2 = (A + iB)(A − iB).]
(CK20161) Let A be a real square matrix of order 2017. Show that

det(A − AT )2017 = 2017(det A − det AT ).

39 / 99
 2.2. Determinants 2.2.3. Evaluation of a determinant using elementary row operations

2.2.3. Evaluation of a determinant using elementary row operations

Elementary row operations

The following operations on rows of matrices are called elementary row operations.
Interchange two rows. (Hi ↔ Hj .)
Multiply a row by a nonzero constant. (Hi ← kHi , k ̸= 0.)
Add a multiple of a row to another row. (Hi ← Hi + kHj .)

Similarly, replacing "rows" by "columns" we obtain elementary column operations.

40 / 99
 2.2. Determinants 2.2.3. Evaluation of a determinant using elementary row operations

Effects of elementary row operations on a determinant

Elementary row operation The effect on a determinant

Interchange two rows Change sign
Multiply a row by a nonzero constan k ̸= 0 Multiply by k
Add a multiple of a row to another row Unchanged

We have a similar table, if we replace "row(s)" by "column(s)".

Evaluating of a determinant using elementary row operations

Use elementary (row, column) operations to transform a given determinant
to a determinant of a triangular matrix,
or to a determiant of a simpler matrix (for example, a matrix that has a column with all zeroes
except at one entry, we can use the Laplace expansion along that row to reduce the determinant to
a determinant of smaller order).

41 / 99
 2.2. Determinants 2.2.3. Evaluation of a determinant using elementary row operations

Example
1 2 −1
Evaluate 2 −1 2 .
3 1 2

We have
1 2 −1 1 2 −1
2 −1 2 = 0 −5 4 (H2 ← H2 − 2H1 )
3 1 2 3 1 2
1 2 −1
= 0 −5 4
0 −5 5 (H3 ← H3 − 3H1 )
1 2 −1
= 0 −5 4
0 0 1 (H3 ← H3 − H2 )
= −5.

42 / 99
 2.3. Rank of a matrix 2.3.1. Rank of a matrix

2.3.1. Rank of a matrix

Let A = [aij ]m×n and p ≤ min{m, n}.

Submatrices
Selecting p columns and and p rows of A, the entries that appear in both the rows and columnns of
those selected form a new square matrix of order p. Such a matrix is called a submatrix of order p. Its
determinant is called a minor (or minor determinant) of order p (in A).
 
ai1 j1 ai1 j2 . . . ai1 jp
ai2 j1 ai2 j2 . . . ai2 jp 
j ,...,j
Notation Ai11,...,ipp =  . .. .. .
 
 .. ..
. . . 
aip j1 aip j2 . . . aip jp
 
1 2 3 4 ï ò ï ò
1,2 1 2 1,4 5 8
Example: A = 5 6 7 8 . We have A1,2 =
 , A2,3 =
5 6 9 12
9 10 11 12

43 / 99
 2.3. Rank of a matrix 2.3.1. Rank of a matrix

Rank of a matrix
Rank of a matrix A is the largest order of any non-zero minor in A. The rank of A is denoted by
rank(A), or r (A) (or ρ(A)).
Remark: The rank of the zero matrix A = O is 0.
Thus, rank(A) = r if and only if A has at least a non-zero minor of order r and every minor of order
greater than r (if exist) is equal to 0.
 
1 2 3 −1
Example: Consider matrix A = 2 1 1 0 .
1 −1 −2 1
The minors of order 3 are
1 2 3 1 2 −1 1 3 −1 2 3 −1
2 1 1 = 0, 2 1 0 = 0, 2 1 0 = 0, 1 1 0 = 0,
1 −1 −2 1 −1 1 1 −2 1 −1 −2 1

1 2
A nonzro minor of order 2 = −3 ̸= 0.
2 1
Hence r (A) = 2.
44 / 99
 2.3. Rank of a matrix 2.3.1. Rank of a matrix

Some properties

Let A be a matrix of size m × n.

1 0 ≤ r (A) ≤ min{m, n}.
2 r (AT ) = r (A).

Rank of a square matrix

Let A be a square matrix of order n.
1 If det(A) ̸= 0 then r (A) = n.
2 If det(A) = 0 then r (A) < n.
Hence, r (A) ≤ n and the equality holds if and only if det(A) ̸= 0.

45 / 99
 2.3. Rank of a matrix 2.3.1. Rank of a matrix

Example (GK20201-N3)
 
0 1 −3
Consider A = m − 1 m − 2 1 , where m is a parameter.
1 m−2 3−m
(a) Find m such that the rank of A is equal to 2.
(b) Find m such that the rank of A3 is less than3.

Part a):
r (A) = 2 ⇒ det(A) = 0.
det(A) = −2m2 + 8m − 8 = −4(m − 2)2 .
det(A) = −4(m − 2)2 = 0 ⇔ m = 2.
Check: if m = 2 then r (A) = 2.
Part b):
r (A3 ) < 3 ⇔ det(A3 ) = 0 ⇔ det(A)3 = 0 ⇔ det(A) = 0.
By part a, m = 2.
46 / 99
 2.3. Rank of a matrix 2.3.1. Rank of a matrix

Definition (Echelon form)

A matrix is in row echelon form if it has the following properties:
The zero rows (if any) are at the bottom of the matrix;
In any non-zero row, the leading entry (that is the first nonzero entry from the left) is to the right
of the leading entry of the row above it.

Here, a zero row is a row that consists entirely of zeros.

An entry a is said to be to the right of the entry b if the column containing a is to the right of the
column containing b.

Example:
   
1 ∗ ∗ ∗ ∗ ∗ ∗ 1 ∗ ∗ ∗ ∗ ∗ ∗
0 0 2 ∗ ∗ ∗ ∗ 0 0 0 3 ∗ ∗ ∗
Matrix 
0 0 0 3 ∗ ∗ ∗ is in row echelon form. Matrix 0 0 2 ∗ ∗ ∗ ∗ is not in row
  

0 0 0 0 0 0 0 0 0 0 0 0 0 0
echelon form.

47 / 99
 2.3. Rank of a matrix 2.3.1. Rank of a matrix

Remark
By using repeatly elementary row operations, we can bring any matrix to a matrix in row echelon form.B.

Example:    
1 2 3 −1 1 2 3 −1
A := 2 1 1 0 −→ 0
  −3 −5 2(H2 ← H2 − 2H1 )
1 −1 −2 1 1 −1 −2 1
 
1 2 3 −1
−→ 0 −3 −5 2 
0 −3 −5 2 (H3 ← H3 − H1 )
 
1 2 3 −1
−→ 0 −3 −5 2 
0 0 0 0 (H3 ← H3 − H2 )

48 / 99
 2.3. Rank of a matrix 2.3.1. Rank of a matrix

Rank of a matrix in row echelon form

Proposition
The rank of a matrix in row echelon form is equal to the number of non-zero row of this matrix.

 
1 ∗ ∗ ∗ ∗ ∗ ∗
0 0 2 ∗ ∗ ∗ ∗
Example: Matrix of form 
0 0 0 3 ∗ ∗ ∗ has rank 3.


0 0 0 0 0 0 0

49 / 99
 2.3. Rank of a matrix 2.3.2. Evaluation of the rank of a matrix

2.3.2. Evaluation of the rank of a matrix

Proposition (Rank and elementary operations)

If B is a matrix obtained from A by using elementary operations then A, thì r (A) = r (B).

(Elementary operations do not change the rank of a matrix.)

Evaluation of the rank of a matrix

Let A be a matrix of size m × n.
1 Using elementary row operations bring A to a matrix B in row echelon form.
2 r (A) = r (B)= the number of nonzero rows in B.

Remarks:
To compute rank, it is enough to use only elementary row operations
In practice, to compute rank, we can also use elementary column operations.

50 / 99
 2.3. Rank of a matrix 2.3.2. Evaluation of the rank of a matrix

Example
 
1 2 3 −1
Evaluate the rank of A = 2 1 1 0 .
1 −1 −2 1

Using elementary row operations

   
1 2 3 −1 1 2 3 −1
A = 2 1 1 0  −→ · · · −→ 0 −3 −5 2 =B
1 −1 −2 1 0 0 0 0

B is in row echelon form and r (A) = r (B) = 2.

51 / 99
 2.3. Rank of a matrix 2.3.2. Evaluation of the rank of a matrix

Example (GK20213)
 
1 −2 1 1
2 0 −2 3
Find the rank of A =  . Find m such that r (A) = 3.
2 0 1 −3
1 2 −2 2

 
1 −2 1 1
0 4 −4 1 
(H2 ← H2 − 2H1 )
A −→ 
0 4 −1 −5(H3 ← H3 − 2H1 )
0 4 −3 1 (H4 ← H4 − H1 )
   
1 −2 1 1 1 −2 1 1
0 4 −4 1  0 4 −4 1
−→   −→   = B.
0 0 3 −6(H3 ← H3 − H2 ) 0 0 1 0
0 0 1 0 (H4 ← H4 − H2 ) 0 0 0 −6

B is in row echelon form and r (A) = r (B) = 4.

Another solution: det(A) = 24 ̸= 0 ⇒ r (A) = 4.
52 / 99
 2.3. Rank of a matrix 2.3.2. Evaluation of the rank of a matrix

Example (GK20191)
 
1 0 −1 m
Consider A = 1 −1 3 5 . Find m such that r (A) = 3.
0 −1 4 3

 
1 0 −1 m
A −→ 0 −1 4 5 − m(H2 ← H2 − H1 )
0 −1 4 3
 
1 0 −1 m
−→ 0 −1 4 5 − m = B.
0 0 0 m − 2 (H3 ← H3 − H2 )

r (A) = 3 ⇔ r (B) = 3 ⇔ m − 2 ̸= 0 ⇔ m ̸= 2.

53 / 99
 2.3. Rank of a matrix 2.3.2. Evaluation of the rank of a matrix

Some exercises
 
2 m 4 3
(GK20201) Find m such that the rank of A = 1 1 2 2 là lớn nhất.
1 1 2 1
 
1 −1 2 b
(GK20191-N3) Find the rank of −2 a −4 −1 − b , where a, b are real numbers.
−3 a + 1 −6 −b
 
1 m −1 2
(GK20181) Find m such that the rank of A = 2 −1 m 5 nhỏ nhất.
1 10 −6 1
 
1 1 2 b
(GK20181-N2) Find the rank of A = 2 1 −1 2, where a, b are real numbers..
4 3 a 5
 
1 m 1 m−2
(GK20161) Consider A = 2 2m + 2 m + 1 2m − 1. Find m such that r (A) = 2.
1 m+2 m 2m + 4
54 / 99
 2.3. Rank of a matrix 2.3.3. Inverse of a matrix

2.3.3. Inverse of a matrix

Let A be a square matrix of orde n.

Định nghĩa
A is said to be invertible if there exists a square matrix B of order n such that n such that

AB = BA = In .

In this case, B is called the inverse of A, denoted by A−1 . Thus,

AA−1 = A−1 A = In .
ï ò ï ò
1 2 3 −2
Example: Matrix A = is invertible and A−1 = .
1 3 −1 1

55 / 99
 2.3. Rank of a matrix 2.3.3. Inverse of a matrix

Remarks

Nonsquare matrices do not have inverses.

A square matrix may be invertible may be not invertible (=noninvertible=singular).
The identity matrix In is invertible andIn−1 = In .
he zero matrix O of order n is not invertible.
The set of invertible square matrix of order n over K is denoted by GLn (K ).

56 / 99
 2.3. Rank of a matrix 2.3.3. Inverse of a matrix

Uniqueness of an inverse matrix

If A is invertible then its inverse is unique.

Existence of an inverese matrix

A square matrix A is invertible if and only if det(A) ̸= 0.

Definition (nonsingular matrix)

A square matrix A is said to be nonsingular if det(A) ̸= 0.

Corollary
Let A and B be square matrices of the same size n. n.
If AB = In then A is invertible, A−1 = B and BA = In .
If BA = In then A is invertible, A−1 = B and AB = In .

57 / 99
 2.3. Rank of a matrix 2.3.3. Inverse of a matrix

Some exercises
 
m+1 m 5
(GK20191-N2) Find m such that matrix A =  1 m 3 − m is nonsingular.
m 0 m+1
 
−a + 4 −a 3
(GK20161) Find a such that A =  4 1 2  is invertible.
3 −a a + 2
(GK20201*) Let A; B be real square matrices of order n (n ≥ 2) satisfying that AB − nA − B = 0.
Show that A2n+1 B = BA2n+1 . [Hint: AB − nA − B = 0⇒ (A − In )(B − nIn ) = nIn ⇒
(B − nIn )(A − In ) = nIn ⇒ BA − nA − B = 0. Hence AB = BA. ]
(GK20171-N2*) Let n be a positive integer such that there are two square invertible matrices A, B
of order n satisfying that AB + BA = O. Show that n ̸= 2017. [Hint: AB = −BA and use det.]
(GK20161*) Let A, B be square matrices of the same order n ≥ 2 such that AB + A + B = 0.
Show that if A is invertible then B is invertible. [Hint: AB + A + B = 0 ⇒ (A + In )B = −A.
Taking det both sides.]

58 / 99
 2.3. Rank of a matrix 2.3.3. Inverse of a matrix

Properties
Let A be an invertible matrix, k ≥ 0 a natural number, 0 ̸= c ∈ K (= R or C). Then A−1 , Ak , cA, AT
are invertible and

−1 1 −1
A−1 = A, (Ak )−1 = (A−1 )k , (cA)−1 = A , (AT )−1 = (A−1 )T .
c

Proposition
Let A and B be invertible matrices of order n. Then AB is also invertible and

(AB)−1 = B −1 A−1 .

Remark: The set GLn (K ) of invertible matrices of order n together with matrix multiplication is a
group.

59 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

2.3.4. Finding the inverse of a matrix

Adjugate matrix
Consider A = [aij ]n×n . Let Cij be the cofactor of aij . Put
   
C11 C12 . . . C1n C11 C21 ... Cn1
C21 C22 . . . C2n  C12 C22 ... Cn2 
T
C = . .. ..  và C =  .. .. .. 
   
 .. .. ..
. . .   . . . . 
Cn1 Cn2 . . . Cnn C1n C2n ... Cnn

The matrix adj(A) := C T is called the adjugate (or adjoint) of A.

60 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Example
 
1 2 −1
Find the adjugate ofA = 2 −1 2 .
3 1 2

−1 2 2 2 2 −1
C11 = + = −4, C12 = − = 2, C13 = + =5
1 2 3 2 3 1
2 −1 1 −1 1 2
C21 = − = −5, C22 = = 5, C23 = − =5
1 2 3 2 3 1
2 −1 1 −1 1 2
C31 = + = 3, C32 = − = −4, C33 = + = −5.
−1 2 2 2 2 −1
The adjugate of A:    
C11 C21 C31 −4 −5 3
adj(A) = C12 C22 C32  =  2 5 −4 .
C13 C23 C33 5 5 −5

61 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

TFinding the inverse of a matrix using minors

Theorem

A · adj(A) = adj(A) · A = det(A)In .

   
1 2 −1 −4 −5 3
Example: A = 2 −1 2  ⇒ det(A) = −5, adj(A) =  2 5 −4 and
3 1 2 5 5 −5
    
1 2 −1 −4 −5 3 −5 0 0
A · adj(A) = 2 −1 2  2 5 −4 =  0 −5 0
3 1 2 5 5 −5 0 0 −5
    
−4 −5 3 1 2 −1 −5 0 0
adj(A) · A =  2 5 −4 2 −1 2 = 0 −5 0
5 5 −5 3 1 2 0 0 −5

62 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Corollary
If det(A) ̸= 0 then A is invertible and
1
A−1 = adj(A).
det(A)

   
1 2 −1 −4 −5 3
Example: A = 2 −1 2  ⇒ det(A) = −5, adj(A) =  2 5 −4 and
3 1 2 5 5 −5
   
−4 −5 3 4/5 1 −3/5
1 1
A−1 = adj(A) = −  2 5 −4 = −2/5 −1 4/5  .
det(A) 5
5 5 −5 −1 −1 1
ï ò ï ò
a b d −b
Example: A = , adj(A) = and if det(A) ̸= 0 then
c d −c a
1
ï ò
−1 d −b
A = .
det(A) −c a
63 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Finding the inverse of a matrix using Gauss-Jordan elimination method

Gauss-Jordan method
Let A be a square matrix of order n.
.
1 Write the identity matrix In after the matrix A to form the matrix [A .. In ] of size n × 2n.
. .
2 Use elementary row operations on the entire matrix [A .. In ] về dạng [In .. B]. (If this is not possible,
then A is not invertible.)
3 We have A−1 = B.
. .
[A .. In ] −→ · · · −→ [In .. A−1 ]

Remark: In Gauss-Jordan method, we only use elementary row operations

64 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Example
ï ò
1 2
Use Gauss-Jordan method to find the inverse of A = .
1 3

ï ò ï ò
. 1 2 1 0 1 2 1 0
[A .. I2 ] = −→
1 3 0 1 0 1 −1 1 (H2 ← H2 − H1 )
3 −2 (H1 ← H1 − 2H2 )
ï ò
1 0
−→
0 1 −1 1
ï ò
3 −2
A−1 = .
−1 1

65 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix
 
1 2 −1
Example: Using Gauss-Jordan method to find the inverse of A = 2 −1 2 .
3 1 2
   
1 2 −1 1 0 0 1 2 −1 1 0 0
..
[A . I3 ] = 2 −1 2 0 1 0 → 0 −5 4 −2 1 0(H2 ← H2 − 2H1 )
3 1 2 0 0 1 0 −5 5 −3 0 1 (H3 ← H3 − 3H1 )
 
1 2 −1 1 0 0
−→ 0 −5 4 −2 1 0
0 0 1 −1 −1 1 (H3 ← H3 − H2 )
 
1 2 0 0 −1 1 (H1 ← H1 + H3)
−→ 0 −5 0 2 5 −4(H2 ← H2 − 4H3)
0 0 1 −1 −1 1
 
1 2 0 0 −1 1
−→ 0 1 0 −2/5 −1 4/5(H2 ← −15
H2 )
0 0 1 −1 −1 1
 
1 0 0 4/5 1 −3/5 (H1 ← H1 − 2H2)
.
−→ 0 1 0 −2/5 −1 4/5 = [I3 .. A−1 ].
0 0 1 −1 −1 1

66 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Invertible matrices and matrix equations)

Basic problem 1
Let A be a square matrix of order n, B a matrix of size n × p. Find a matrix X (of size n × p) such that

AX = B.
If A is invertible then the above equation always has a unique solution:
AX = B ⇔ A−1 AX = A−1 B ⇔ In X = A−1 B ⇔ X = A−1 B.

Basic problem 2
Let A be a square matrix of order n, B matrix of size m × n. Find a matrix X (of sizr m × n) such that

XA = B.
If A is invertible then the above equation always has a unique solution:
XA = B ⇔ XAA−1 = BA−1 ⇔ XIn = BA−1 ⇔ X = BA−1 .
67 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Some variants
Given invertible matrices A and B, find X such that AXB = C . Answer: X = A−1 CB −1
Given an invertible matrix A, find X such that AX + B = C . Answer: X = A−1 (C − B).
Given an invertible matrix A, find X such that XA + B = C . Answer: X = (C − B)A−1 .

68 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Example (GK20201)
ï ò ï òT
4 1 −1 3
Find matrix X such that X = 2X + .
5 5 2 1

The above equation is equivalent to

Åï ò ã ï ò
4 1 −1 2
− 2I2 X =
5 5 3 1
ï ò ï ò
2 1 −1 2
⇔ X =
5 3 3 1
ï ò−1 ï ò ï òï ò ï ò
2 1 −1 2 3 −1 −1 2 −6 5
⇔X = = = .
5 3 3 1 −5 2 3 1 11 −8

69 / 99
 2.3. Rank of a matrix 2.3.4. Finding the inverse of a matrix

Some exercises
ï ò ï ò
1 3 0 4
(GK20201) Consider A = ,B= .
1 2 2 −1
a) Let P(x) = x 2 − 3x − 2. Find P(A).
b) Find matrix X such that (A2 − 3A2 − 2A)XA = AB T .
ï ò ï ò
1 1 2 1
(GK20201-N3) Let A = ,B= . Find matrix X such that A2 XA = AB.
2 1 3 4
   
ï ò 0 2 1 3
1 −2
(GK20191) Find matrix X such that X − 3 −4 1  = 2 5 1 .
−2 5
3 −2 −3 1
 
ï ò 1 −1
1 −1
(GK20191-N3) Consider A = , B = 3 0 . Find X such that 2X = B T + AX .
2 5
1 2
   
1 −1 1 1 3
(GK20181) Find matrix X such that  0 1 2  X = 4 −1.
−1 1 −2 5 7
ï ò ï òT ï ò
2 1 1 3 0 −1
(GK20171) Find matrix X such that X −2 = .
1 1 2 5 1 0
70 / 99
 2.4. Systems of linear equations 2.4.1. Definition

2.4.1. Definition

Let K be a field, K = R or K = C.
Linear equation
A linear equation in n variables x1 , . . . , xn is an equation of the form

a1 x1 + a2 x2 + · · · + an xn = b,

where the coefficients a1 , . . . an and the constant term b are in K .

A sequence of n numbers s1 , s2 , . . . , sn ∈ K satisfying that

a1 s1 + a2 s2 + · · · + an sn = b

is called a solution to the equation.

71 / 99
 2.4. Systems of linear equations 2.4.1. Definition

System of linear equations

A system of linear equations of m equations, in n variables x1 , x2 , . . . , xn is a set of m equation of
the form


 a11 x1 + a12 x2 + · · · + a1n xn = b1

a x + a x + · · · + a x
21 1 22 2 2n n = b2
(1)


 . . . . . . . . . . . .
am1 x1 + am2 x2 + · · · + amn xn = bm


where the coefficients aij and the constant terms bi (1 ≤ i ≤ m, 1 ≤ j ≤ m) are in K .

A sequence of n numbers s1 , s2 , . . . , sn ∈ K satisfying that

ai1 s1 + ai2 s2 + · · · + ain sn = bi

for all i = 1, 2, . . . , m is called a solution to the system (1).

In the system (1) f all constant terms bi = 0 (for i = 1, 2, . . . , n) then the system is called a
homogeneous system.

72 / 99
 2.4. Systems of linear equations 2.4.1. Definition

Matrix form of a system of linear equations

Consider the linear system (1).

1 The matrix A = [aij ]m×n is called the coefficient matrix of (1).
   
x1 b1
 x2   b2 
2 Let x = 
. . . and b = . . ..
  

xn bm
3 The system (1) can be presented in the form

Ax = b.

73 / 99
 2.4. Systems of linear equations 2.4.2. Cramer’s system

2.4.2. Cramer’s system

Definition (Cramer’s system)
A system of n linearequations in n variable whose coefficient matrix has nonzero determinant, is called

 a11 x1 + a12 x2 + · · · + a1n xn = b1

a x + a x + · · · + a x
21 1 22 2 2n n = b2
a Cramer’s system and det([aij ]n×n ) ̸= 0.
. . . . . . . . . . . .


an1 x1 + an2 x2 + · · · + ann xn = bn


In matrix form: Ax = b. (Here, det(A) ̸= 0.)

Theorem (Cramer’s rule)
Any Cramer’s system as above has a unique solution, whose individual values for variables are given by

det(A1 ) det(An )
x1 = , . . . , xn = ,
det(A) det(A)

Here Aj is the matrix obtained from A by replacing the j-th column of A by the column b.
74 / 99
 2.4. Systems of linear equations 2.4.2. Cramer’s system

Example

x + 2y − z
 =1
Consider the system 2x − y + 2z = 1 . Find y .

3x + y + 2z =1


1 2 −1
The determinant of the coefficient matrix of the system 2 −1 2 = −5 is different from 0.
3 1 2
The system is a Cramer’s system.
1 1 −1
|A2 | 3
|A2 | = 2 1 2 = 3 and y = =− .
|A| 5
3 1 2
6
(x = , z = −1.)
5

75 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

2.4.3. General system of linear equations

Consider the system 


 a11 x1 + a12 x2 + · · · + a1n xn = b1

a x + a x + · · · + a x
21 1 22 2 2n n = b2


 . . . . . . . . . . . .
am1 x1 + am2 x2 + · · · + amn xn

= bm

Augmented matrix
The augmented matrix of the above system is
 
a11 a12 . . . a1n b1
 a21 a22 . . . a2n b2 
Ā = [A | b] =  .. .. . . .. ..
 
 . . . . .
am1 am2 . . . amn bm

76 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Echelon system
A system of linear equations is said to be in echelon form if its augmented matrix is in row echelon form.

Example: System in echelon form:

 
x + 2y + z = 2 
 x + 2y + z =2
y + z = 2, y +z = 2,
 
0·z =2 2z =2
 

®
x + 2y − z + t =2
z +t =1

77 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Remark
We can solve easily a system in echelon form by using back-substitution.

x + 2y + z = 2

Hệ y + z = 2 has no solution..

0·z =2

 
x + 2y + z = 2
 x = 2 − 2y − z = 1 − 2 · 1 − 1 = −1

System y +z =2 ⇔ y =2−z =1 has a unique solution
 
2z =2 z =1
 
(x, y , z) = (−1, 1, 1).


 x = 2 − 2y + z − t = 2 − 2a + (1 − b) − b = 3 − 2a − 2b
® 
x + 2y − z + t = 2  y = a (a ∈ R)
System ⇔
z +t =1 z = 1 − b


t = b (b ∈ R)


This system has infinitely many solutions, (solutions can be described using 2 parameters)
(x, y , z, t) = (3 − 2a − 2b, a, 1 − b, b), a, b ∈ R.

78 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Solving systems in echelon form

Consider a system Ax = b in echelon form, with n variables. Let Ā = [A | b] be the augmented matrix of
the system.
Case 1: r (A) < r (Ā). The system has no solution.
Case 2: r (A) = r (Ā) = n. The system has a unique solution.
We first find xn (from the last equation) and subsitute xn into the previous equation to find xn−1 ,
etc.
Case 3: r (A) = r (Ā) < n. The system has infinitely many solutions. The solutions can be described
using n − r parameter, where r := r (A).
We keep r variables (called basic variables) corresponding to r columns containing leading entries
in r non-zero rows. We consider n − r remaining variables as parameters (free variables). We solve
r basic variables (from the last equation backward) in term n − r free variables.

79 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Equivalent systems
Two systems are equivalent if they have the same set of solutions.

Elementary operations
The following operations bring a system to an equivalent system:
1 Interchange two equations,
2 Multiply an equation by a nonzero constant,
3 Add a multiple of an equation to another equation.

80 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Example
 
x + 2y + z
 =2 x + 2y + z = 2

x + 3y + 2z =4 ⇔ y +z = 2 (Eq2 ← Eq2 − Eq1 )
 
2x + 5y + 5z =8 y + 3z = 4 (Eq3 ← Eq3 − 2Eq1 )
 

x + 2y + z
 =2
⇔ y +z =2

2z = 2 (Eq3 ← Eq3 − 2Eq2 )


Example: Elementary operations on systems correspond to elementary row operations on augmented

matrices
   
1 2 1 2 1 2 1 2
1 3 2 4 −→ 0 1 1 2(H2 ← H2 − H1)
2 5 5 8 0 1 3 4 (H3 ← H3 − 2H1)
 
1 2 1 2
−→ 0 1 1 2
0 0 2 2 (H3 ← H3 − H2 )
81 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Gaussian elimination method to solve a system (matrix form)

1 Write the augmented matrix of the system.

2 Use elementary row operations to bring the augmented matrix to a matrix in row echelon form.
3 Write the system associated to the matrix obtained in Step 2. This system is in echelon form. Solve
this system.
Remark: Gaussian elemination use elementary row operations..

82 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Example (GK20191-N2)

x + y − 2z
 =1
Solve the system 2x +z = 3 by Gaussian elimination.

3x + y − z =4


The augmented matrix of the system

     
1 1 −2 1 1 1 −2 1 1 1 −2 1
Ā = 2 0 1 3 −→ 0 −2 5 1 −→ 0 −2 5 1
3 1 −1 4 0 −2 5 1 0 0 0 0
®
x + y − 2z =1
The corresponding system in echelon form
− 2y + 5z = 1
HThis system has infinitely many solutions. Set z = 2t + 1, (t ∈ R). Then y = 5t + 2 and
x = 1 − t. Solutions are
(x, y , z) = (1 − t, 5t + 2, 1 + 2t) (t ∈ R)
Remark: There are many ways to describe the set of solutions of the above system using one
parameter 83 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Some exercises


x1 + x2 + x4
 =3

x − 2x + x
1 3 4 =5
(GK20181) Solve the system


 x2 + x3 + 2x4 =3
3x1 + x2 − 3x3 + x4

=8
Solve the following system using Gaussian elimination:
 

 x + 2y − z + 3t = 12 
 x + 2y − z + 3t =2

2x + 5y − z + 11t 
2x + 2y − 5z + 8t
= 49 =2
(GK20171) , (GK20161)


 3x + 6y − 4z + 13t = 49 

 x + 4y + 4z + 6t = 13
x + 2y − 2z + 9t 4x + 8y − 2z + 14t
 
= 33 = 14

x − 2y + z
 =1
(GK20171) Solve the system 2x − 3y + z = 4.

3x − 5y + 2z = 5


84 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Kronecker-Capelli theorem
Consider a system of linear equations Ax = b. he system has a solution if and only if r (A) = r (Ā). (Here
A = [A | b] is the augmented matrix of the system.)

Cụ thể hơIn fact, we have the following

Let n be the number of variables in the system
r (A) ̸= r (Ā) ⇔ the system has no solution
r (A) = r (Ā) = n ⇔ the system has a unique solution.
r (A) = r (Ā) < n ⇔ has infinitely many solutions (which can be described by using (n − r )
parameters)

85 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations


x1 + 2x2 − x3 + 2x4
 =b
Example (GK20201) Consider 2x1 + 5x2 + (a + 3)x4 = 1 + 2b , where a, b are constants.

x1 + x2 − 3x3 + x4 = 2b

1 Solve the system when a = b = −1.
2 Find the values of a, b such that the system has infinitely many solutions.
     
1 2 −1 2 b 1 2 −1 2 b 1 2 −1 2 b
Ā = 2 5 0 a+3 1 + 2b  → 0 1 2 a−1 1 → 0 1 2 a−1 1.
1 1 −3 1 2b 0 −1 −2 −1 b 0 0 0 a−2 b+1
If a = b = −1 then r (A) = r (Ā) = 3 < 4 (the number of variables). The system has infinitely many
solutions. Solutions are

(x1 , x2 , x3 , x4 ) = (−3 + 5t, 1 − 2t, t, 0),

( với t ∈ R).
ñ
a ̸= 2
HThe system has infinitely many solutions ⇔ r (A) = r (Ā) < 4 ⇔ .
a = 2, b = −1

86 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Particular case: square systems

Let A be a square matrix of order n. The following statements are equivalent.

A khả nghịch.
det(A) ̸= 0.
r (A) = n.
For any column matrixb, the system Ax = b has a unique solution.
TFor some column matrix b, the system Ax = b has a unique solution.
The (homogeneous) system Ax = 0 has a unique solution
*For any column matrix b, the system Ax = b has a solution.

87 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

Some exercises

x1 + 2x2 + mx3 + (m + 1)x4
 =0
(GK20171) Find m such that the system 2x1 + (m + 2)x2 + (2m + 1)x3 + (2m + 4)x4 = 0 has

x1 + (4 − m)x2 + (m − 1)x3 + (2m − 4)x4 =0

infinitely many solutions which can
 be described using 2 parameters.
x
 1
 + 2x 2 + mx 3 + x4 = −1
(GK20171) Consider the system 2x1 + 5x2 − 2x3 + x4 = −2

x1 + 2x2 − x3 + mx4 = 3

When m = −1, find the rank of the augmented matrix of the system and solve the system
Solve the system (in terms of m).
(CK202161) Find a, b for which the following system has infinitely many solutions with 1 arbitrary
parameter

x1 − x3 + ax4
 =2
2x1 − 2x2 + 3x3 + (3a + 1)x4 =b+6

3x1 + 4x2 − 13x3 + (2a − 2)x4 = −b − 1


88 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

   
−2 1 −3 −6
1 0 5 6
(GK20191) Find m such that there exists a matrix X satisfying that 
  X = 
 .
−1 2 −1 m
0 1 3 2
(GK20191) Findm such that
  the following matrix equation has infinitely many solutions
1 2 m −1
2 7 2m + 1 X =  2 .
3 9 4m 1

mx + 4y − z = 4

(GK20191-N3) Determine the value of m for which the system x − y = 1 has a unique

2x + y + mz = 5

solution. 
2x1 + x2 − 3x3
 =2
(GK20201-N3) Find a, b such that the following system has a solution 5x1 + 2x2 + x3 = 4.

3x1 + x2 + ax3 =b


89 / 99
 2.4. Systems of linear equations 2.4.3. General system of linear equations

(GK20201) Let f : R3 → R3 be a map given by f (x, y , z) = (x + y − z, 2x − y , x − my + z), where

m is a constant. Find the value of m for which f is bijective.
(GK20181) Let f : R3 → R3 be a map given by f (x, y , z) = (2x − y + z, x − z, x + my ). Find m
such that f is surjective.
(GK20181) Let f : R3 → R3 be a map given by f (x, y , z) = (2x − 3y + z, x + y , x − mz). Find m
such that f is injective.
(GK20191) Let M1×2 be the set ò matrices of size 1 × 2. Find m such that the map
ï of real
2 −1
f : M1×2 → M1×2 , f (X ) = X is injective.
4 m
ò matrices of size 2 × 1. Find m such that the map
(GK20191) Let M1×2 be theï set of real
1 −2
f : M2×1 → M2×1 , f (X ) = X is surjective.
m 4

90 / 99
 2.4. Systems of linear equations 2.4.4. Homogeneous systems

2.4.4. Homogeneous systemst

Homogeneous systems
A homogeneous system is a system of the form

a11 x1 + a12 x2 + · · · + a1n xn


=0
a x + a x + · · · + a x =0
21 1 22 2 2n n
. . . . . . . . . . . .


am1 x1 + am2 x2 + · · · + amn xn =0


Matrix form:
Ax = 0.

91 / 99
 2.4. Systems of linear equations 2.4.4. Homogeneous systems

Solutions to homogeneous systems

Consider a homogeneous system Ax = 0 in n variables.

The system always has a solution. In fact, (0, 0, . . . , 0)is always a solution to the system. This
solution is called the trivial solution to the system
A solution (c1 , . . . , cn ) ̸= (0, 0, . . . , 0), if exist, is a called a nontrivial solution.

Homogeneous systems
onsider a homogenous square system t Ax = 0 in n variables. Only two following cases can occur.
1 Case 1: The system has only the trivial solution. This happens if only if r (A) = n (=the number of
variables). In this case, the system has a unique solution.
2 Case 2: The solution has nontrivial solutions. This happens if only if r (A) < n. In this case, the
system has infinitely many solutions.

92 / 99
 2.4. Systems of linear equations 2.4.4. Homogeneous systems

Homogeneous square systems

Homogeneous square systems

Consider a homogeneous square system Ax = 0 where A is a square matrix. Then
1 The system has only the trivial solution ⇔ det(A) ̸= 0,
2 The solution has nontrivial solutions ⇔ det(A) = 0.

93 / 99
 2.4. Systems of linear equations 2.4.4. Homogeneous systems

Example (GK20201)

x + 2y + z
 =0
Find the value(s) of m for which the system −x + my − 2z = 0 has nontrivial solutions.

(m + 1)x + y − z =0


1 2 1
The homogeneous square system has nontrivial solutions⇔ det(A) = −1 m −2 = 0.
m+1 1 −1
det(A) = −m2 − 6m − 5.
det(A) = 0 ⇔ m = −1 or m = −5.

94 / 99
 2.4. Systems of linear equations 2.4.4. Homogeneous systems

Some exercises

(GK20181-N2)
 Find m such that the following system has infinitely many solutions

 mx + 2y + z =0
3x − y + 2z = 0.

7x + y + 2mz = 0


(a + 5)x + 3y + (2a + 1)z = 0

(GK20171) Find a such that the system ax + (a − 1)y + 4z = 0 has nontrivial solutions.

(a + 5)x + (a + 2)y + 5z =0


x1 + 2x2 + mx3 + (m + 1)x4
 =0
(GK20171) Find m such that the system 2x1 + (m + 2)x2 + (2m + 1)x3 + (2m + 4)x4 = 0 has

x1 + (4 − m)x2 + (m − 1)x3 + (2m − 4)x4 =0

infinitely many solutions which can be described using 2 parameters.

95 / 99