0% found this document useful (0 votes)
8 views64 pages

Chapter 2 Matrix Algebra

Đại số tuyến tính chương 2

Uploaded by

oanhtran020906
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
8 views64 pages

Chapter 2 Matrix Algebra

Đại số tuyến tính chương 2

Uploaded by

oanhtran020906
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 64

• 2.1 Matrix Addition, Scalar • 2.

4 Matrix Inverses
Multiplication, and Transposition • 2.5 Element Matrices
• 2.2 Equations, Matrices, and • 2.7 LU-Factorization
Transformations • 2.8 An Application to Input-Output
• 2.3 Matrix Multiplication Economic Models
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition

• A rectangular array of numbers is called a matrix (the plural is matrices),


and the numbers are called the entries of the matrix. Matrices are usually
denoted by uppercase letters: 𝐴, 𝐵, 𝐶, and so on.
• In general, a matrix with 𝑚 rows and 𝑛 columns is referred to as an 𝒎 × 𝒏
matrix or as having size 𝒎 × 𝒏.
𝑎11 𝑎12 𝑎13 ⋯ 𝑎1𝑛
𝑎21 𝑎22 𝑎23 ⋯ 𝑎2𝑛
𝐴= ⋮ ⋮ ⋮ ⋮
𝑎𝑚1 𝑎𝑚2 𝑎𝑚3 ⋯ 𝑎𝑚𝑛
• A matrix of size 1 × 𝑛 is called a row matrix, whereas one of size 𝑚 × 1 is
called a column matrix. Matrices of size 𝑛 × 𝑛 for some 𝑛 are called
square matrices.

2
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition

Two matrices A and B are called equal (written A = B) if and only if:
1. They have the same size.
2. Corresponding entries are equal.

EXAMPLE 1
𝑎 𝑏 1 2 −1 1 0
Given 𝐴 = ,𝐵 = , and 𝐶 = , discuss the
𝑐 𝑑 3 0 1 −1 2
possibility that 𝐴 = 𝐵, 𝐵 = 𝐶, 𝐴 = 𝐶

3
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition
Matrix Addition
Definition 1.1
If 𝐴 and 𝐵 are matrices of the same size, their sum 𝐴 + 𝐵 is the matrix formed
by adding corresponding entries.

If 𝐴 = 𝑎𝑖𝑗 and 𝐵 = 𝑏𝑖𝑗 , this takes the form


𝐴 + 𝐵 = 𝑎𝑖𝑗 + 𝑏𝑖𝑗
EXAMPLE 2
2 1 3 1 1 −1
If 𝐴 = and 𝐵 = , compute 𝐴 + 𝐵.
−1 2 0 2 0 6

EXAMPLE 3
Find 𝑎, 𝑏, and 𝑐 if 𝑎 𝑏 𝑐 + 𝑐 𝑎 𝑏 = 3 2 −1 .
4
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition

• The negative of an 𝑚 × 𝑛 matrix 𝐴 (written −𝐴) is defined by


−𝐴 = −𝑎𝑖𝑗 , where 𝐴 = 𝑎𝑖𝑗
Hence, 𝐴 + −𝐴 = 0, where 0 is the zero matrix of the same size as 𝐴.
• If 𝐴 and 𝐵 are two 𝑚 × 𝑛, matrices, their difference 𝐴 − 𝐵 is defined by
𝐴 − 𝐵 = 𝐴 + −𝐵 .

EXAMPLE 4
3 −1 0 1 −1 1 1 0 −2
Let 𝐴 = ,𝐵 = , and 𝐶 = .
1 2 −4 −2 0 6 3 1 1
Compute −𝐴, 𝐴 − B, and 𝐴 + 𝐵 − 𝐶.
5
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition

EXAMPLE 2
3 2 1 0
Solve +𝑋 = , where 𝑋 is a matrix.
−1 1 −1 2
Scalar Multiplication
Definition 2.2
If A is any matrix and k is any number, the scalar multiple kA is the matrix
obtained from A by multiplying each entry of A by k.
If 𝐴 = [𝑎𝑖𝑗 ], then 𝑘𝐴 = 𝑘𝑎𝑖𝑗 .
EXAMPLE 6
3 −1 4 1 2 −1 1
If 𝐴 = and 𝐵 = , compute 5𝐴, 𝐵, and 3𝐴 − 2𝐵.
2 0 1 0 3 2 2 6
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition

Theorem 1
Let 𝐴, 𝐵, and 𝐶 denote arbitrary 𝑚 × 𝑛 matrices where 𝑚 and 𝑛 are
fixed. Let 𝑘 and 𝑝 denote arbitrary real numbers. Then
1. 𝐴 + 𝐵 = 𝐵 + 𝐴.
2. 𝐴 + (𝐵 + 𝐶) = (𝐴 + 𝐵) + 𝐶.
3. There is an 𝑚 × 𝑛 matrix 0, such that 0 + 𝐴 = 𝐴 for each 𝐴.
4. For each 𝐴 there is an 𝑚 × 𝑛 matrix, −𝐴, such that 𝐴 + (−𝐴) = 0.
5. 𝑘(𝐴 + 𝐵) = 𝑘𝐴 + 𝑘𝐵.
6. (𝑘 + 𝑝)𝐴 = 𝑘𝐴 + 𝑝𝐴.
7. (𝑘𝑝)𝐴 = 𝑘(𝑝𝐴).
8. 1𝐴 = 𝐴.
7
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition
Transpose of a Matrix
Definition 1.1
If 𝐴 is an 𝑚 × 𝑛 matrix, the transpose of 𝐴, written 𝐴𝑇 , is the 𝑛 × 𝑚 matrix
whose rows are just the columns of 𝐴 in the same order.

EXAMPLE 9
Write down the transpose of each of the following matrices:
1 1 2 3 1 −1
𝐴= 3 𝐵= 5 2 6 𝐶= 3 4 𝐷= 1 3 2 .
2 5 6 −1 2 1

8
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition

Theorem 1
Let 𝐴 and 𝐵 denote matrices of the same size, and let 𝑘 denote a scalar.
1. If 𝐴 is an 𝑚 × 𝑛 matrix, then 𝐴𝑇 is an 𝑛 × 𝑚 matrix.
2. 𝐴𝑇 𝑇 = 𝐴.
3. (𝑘𝐴)𝑇 = 𝑘𝐴𝑇 .
4. (𝐴 + 𝐵)𝑇 = 𝐴𝑇 + 𝐵𝑇 .

EXAMPLE 10
𝑇
𝑇 1 2 2 3
Solve for 𝐴 if 2𝐴 − 3 = .
−1 1 −1 2

9
SECTION 2 .1 Matrix Addition, Scalar Multiplication, and Transposition

EXAMPLE 11
If 𝐴 and 𝐵 are symmetric 𝑛 × 𝑛 matrices, show that 𝐴 + 𝐵 is symmetric.

EXAMPLE 12
Suppose a square matrix 𝐴 satisfies 𝐴 = 2𝐴𝑇 . Show that necessarily 𝐴 = 0.

10
SECTION 2 .2 Equations, Matrices
Vectors
Definition 2.4
Let ℝ denote the set of all real numbers. The set of all ordered n-tuples from ℝ
has a special notation:
ℝ𝑛 denotes the set of all ordered n-tuples of real numbers.
Matrix-Vector Multiplication
Now consider any system of linear equations with 𝑚 × 𝑛 coefficient matrix
𝑥1
𝑥2
and augmented matrix: 𝐴 = (𝒂𝟏 𝒂𝟐 … 𝒂𝒏 |𝒃), and if 𝐱 = ⋮ is the matrix of
𝑥Ǘ 𝑛
variables then, the system can be written as a single vector equation
𝑥1 𝒂𝟏 + 𝑥2 𝒂𝟐 + ⋯ + 𝑥𝑛 𝒂𝒏 = 𝒃 (∗∗∗) 11
SECTION 2 .2 Equations, Matrices
EXAMPLE 1

3𝑥1 + 2𝑥2 − 4𝑥3 =0


Write the system ቐ 𝑥1 − 3𝑥2 + 𝑥3 = 3 in the form given in (∗∗∗).
𝑥2 − 5𝑥3 = −1

Definition 2.5 Matrix-Vector Products


Let 𝐴 = 𝒂𝟏 𝒂𝟐 ⋯ 𝒂𝒏 be an 𝑚 × 𝑛 matrix, written in terms of its columns
𝑥1
𝑥2
𝒂𝟏 𝒂𝟐 ⋯ 𝒂𝒏 . If 𝐱 = ⋮ is any n-vector, the product 𝐴𝒙 is defined to be the
𝑥Ǘ 𝑛
m-vector given by:
𝐴𝒙 = 𝑥1 𝑎1 + 𝑥2 𝑎2 + ⋯ + 𝑥𝑛 𝑎𝑛
12
SECTION 2 .2 Equations, Matrices
Theorem 1
1) Every system of linear equations has the form 𝐴𝒙 = 𝒃 where 𝐴 is the
coefficient matrix, 𝒃 is the constant matrix, and 𝒙 is the matrix of
variables.
2) The system 𝐴𝒙 = 𝒃 is consistent if and only if 𝒃 is a linear combination of
the columns of 𝐴.
𝑥1
𝑥2
3) If 𝒂1 , 𝒂2 , … , 𝒂𝑛 are the columns of 𝐴 and if 𝒙 = ⋮ , then 𝒙 is a solution
𝑥𝑛
to the linear system 𝐴𝒙 = 𝒃 if and only if 𝑥1 , 𝑥2 , … , 𝑥𝑛 are a solution of the
vector equation 𝑥1 𝒂1 + 𝑥2 𝒂2 + ⋯ + 𝑥𝑛 𝒂𝑛 = 𝒃.
• A system of linear equations in the form Ax = b as in (1) of Theorem 1 is
13
said to be written in matrix form.
SECTION 2 .2 Equations, Matrices
EXAMPLE 2
2
2 −1 3 5
1
If 𝐴 = 0 2 −3 1 and 𝐱 = , compute 𝐴𝐱.
0
−3 4 1 2
−2

EXAMPLE 3
Given columns 𝐚1 , 𝐚2 , 𝐚3 , and 𝐚4 in ℝ3 , write 2𝐚1 − 3𝐚2 + 5𝐚3 + 𝐚4 in the
form 𝐴x where 𝐴 is a matrix and x is a vector.

14
SECTION 2 .2 Equations, Matrices

EXAMPLE 4
Let 𝐴 = 𝐚1 𝐚2 𝐚3 𝐚4 be the 3 × 4 matrix given in terms of its columns
2 1 3 3
𝐚1 = 0 , 𝐚2 = 1 , 𝐚3 = −1 , and 𝐚4 = 1 . In each case below, either
−1 1 −3 0
express b as a linear combination of 𝐚1 , 𝐚2 , 𝐚3 , and 𝐚4 , or show that it is not
such a linear combination. Explain what your answer means for the
corresponding system 𝐴𝐱 = 𝐛 of linear equations.

1 4
(a) 𝐛 = 2 (b) 𝐛 = 2 .
3 1
15
SECTION 2 .2 Equations, Matrices

EXAMPLE 5
Taking 𝐴 to be the zero matrix, we have 0𝐱 = 0 for all vectors 𝐱 by Definition
2.5 because every column of the zero matrix is zero. Similarly, 𝐴0 = 0 for all
matrices 𝐴 because every entry of the zero vector is zero.

EXAMPLE 6
1 0 0
If 𝐼 = 0 1 0 , show that 𝐼𝐱 = 𝐱 for any vector 𝐱 in ℝ3 .
0 0 1

16
SECTION 2 .2 Equations, Matrices
Theorem 2
Let 𝐴 and 𝐵 be 𝑚 × 𝑛 matrices, and let 𝒙 and 𝒚 be 𝑛-vectors in ℝ𝑛 . Then:
(1) 𝐴(𝒙 + 𝒚) = 𝐴𝒙 + 𝐴𝒚
(2) 𝐴(𝑎𝒙) = 𝑎(𝐴𝒙) = 𝑎𝐴 𝒙 for all scalars 𝑎.
(3) (𝐴 + 𝐵)𝒙 = 𝐴𝒙 + 𝐵𝒙.
• The system
𝐴x = 𝟎
is called the associated homogeneous system, obtained from the original
system 𝐴𝐱 = 𝐛 by replacing all the constants by zeros.
• Suppose 𝑥1 is a solution to 𝐴𝐱 = 𝐛 and 𝐱 0 is a solution to 𝐴𝐱 = 𝟎. Then
𝐱1 + 𝐱 0 is another solution to 𝐴x = 𝐛.
17
SECTION 2 .2 Equations, Matrices

Theorem 3
Suppose 𝒙1 is any particular solution to the system 𝐴𝒙 = 𝒃 of linear
equations. Then every solution 𝒙2 to 𝐴𝒙 = 𝒃 has the form
𝒙2 = 𝒙0 + 𝒙1
for some solution 𝒙0 of the associated homogeneous system 𝐴𝒙 = 𝟎.

EXAMPLE 7
Express every solution to the following system as the sum of a specific
solution plus a solution to the associated homogeneous system.
𝑥1 − 𝑥2 − 𝑥3 + 3𝑥4 = 2
2𝑥1 − 𝑥2 − 3𝑥3 + 4𝑥4 = 6
𝑥1 −2𝑥3 + 𝑥4 = 4
18
SECTION 2 .2 Equations, Matrices
The Dot Product
Definition 2.6
If 𝑎1 , 𝑎2 , … , 𝑎𝑛 and 𝑏1 , 𝑏2 , … , 𝑏𝑛 are two ordered 𝑛-tuples, their dot
product is defined to be the number
𝑎1 𝑏1 + 𝑎2 𝑏2 + ⋯ + 𝑎𝑛 𝑏𝑛
obtained by multiplying corresponding entries and adding the results.

Theorem 4 Dot Product Rule


Let A be an m × n matrix and let x be an n-vector.
Then each entry of the vector Ax is the dot product of
the corresponding row of A with x.
19
SECTION 2 .2 Equations, Matrices

EXAMPLE 8
2
2 −1 3 5
1
If 𝐴 = 0 2 −3 1 and 𝐱 = , compute 𝐴𝐱.
0
−3 4 1 2
−2

EXAMPLE 9
Write the following system of linear equations in the form 𝐴x = 𝐛.
5𝑥1 − 𝑥2 + 2𝑥3 + 𝑥4 − 3𝑥5 = 8
𝑥1 + 𝑥2 + 3𝑥3 − 5𝑥4 + 2𝑥5 = −2
−𝑥1 + 𝑥2 − 2𝑥3 + (−3)𝑥5 = 0
20
SECTION 2 .2 Equations, Matrices

EXAMPLE 10
If A is the zero m × n matrix, then Ax = 0 for each n-vector x.
Definition 2.7
For each 𝑛 ≥ 2, the identity matrix 𝐼𝑛 is the n × n matrix with 1s on the
main diagonal (upper left to lower right), and zeros elsewhere.
The first few identity matrices are
1 0 0 0
1 0 0
1 0 0 1 0 0
𝐼2 = , 𝐼3 = 0 1 ,
0 4 𝐼 = ⋯
0 1 0 0 1 0
0 0 1
0 0 0 1
EXAMPLE 11
For each 𝑛 ≥ 2 we have 𝐼𝑛 𝒙 = 𝒙 for each n-vector x in ℝ𝑛 .
21
SECTION 2 .2 Equations, Matrices
EXAMPLE 12
Let 𝐴 = [𝒂1 𝒂2 … 𝒂𝑛 ] be any 𝑚 × 𝑛 matrix with columns 𝐚1 , 𝐚2 , … , 𝐚𝑛 . If 𝒆𝑗
denotes column j of the 𝑛 × 𝑛 identity matrix 𝐼𝑛 , then 𝐴𝒆𝑗 = 𝐚𝑗 for each
𝑗 = 1, 2, … , 𝑛.

Theorem 5
Let 𝐴 and 𝐵 be 𝑚 × 𝑛 matrices. If 𝐴𝒙 = 𝐵𝒙 for all 𝒙 in ℝ𝑛 , then 𝐴 = 𝐵.
Transformations
We identify:
• A point (𝑎1 , 𝑎2 ) as the vector
𝑎1 , 𝑎2 𝑇 ∈ ℝ2 .
• A point (𝑎1 , 𝑎2 , 𝑎3 ) as the vector
𝑎1 , 𝑎2 , 𝑎3 𝑇 ∈ ℝ3 . 22
SECTION 2 .2 Equations, Matrices

EXAMPLE 13
Consider the transformation of ℝ2 given by reflection in
𝑎1
the 𝑥 axis. This operation carries the vector 𝑎 to its
2
𝑎1
reflection −𝑎 as in Figure 3. Now observe that
2
𝑎1 1 0 𝑎1
−𝑎2 = 0 −1 𝑎2
𝑎1
so reflecting 𝑎 in the 𝑥 axis can be achieved by
2
1 0
multiplying by the matrix .
0 −1

23
SECTION 2 .2 Equations, Matrices

Functions 𝑇: ℝ𝑛 → ℝ𝑚 are called transformations


from ℝ𝑛 to ℝ𝑚 . Such a transformation 𝑇 is a rule
that assigns to every vector 𝐱 in ℝ𝑛 a uniquely
determined vector 𝑇(𝐱) in ℝ𝑚 called the image of 𝐱
under 𝑇. We denote this state of affairs by writing
𝑇
𝑇: ℝ → ℝ or ℝ → ℝ𝑚
𝑛 𝑚 𝑛

EXAMPLE 14

𝑥1 𝑥1 + 𝑥2
𝑥2 𝑥2 + 𝑥3
The formula 𝑇 𝑥 = defines a transformation ℝ4 → ℝ3 .
3 𝑥3 + 𝑥4
𝑥4
24
SECTION 2 .2 Equations, Matrices
Definition 2.8
If A is any 𝑚 × 𝑛 matrix, multiplication by A gives a transformation
𝑇𝐴 : ℝ𝑛 → ℝ𝑚 defined by 𝑇𝐴 (𝑥) = 𝐴𝑥 for every 𝑥 ∈ ℝ𝑛 .
𝑇𝐴 is called the matrix transformation induced by A

EXAMPLE 15
Let 𝑅𝜋 : ℝ2 → ℝ2 denote counterclockwise rotation
2
𝜋
about the origin through radians (that is, 90∘ ). Show
2
0 −1
that 𝑅 is induced by the matrix
𝜋 .
2 1 0

25
SECTION 2 .3 Matrix Multiplication

Definition 2.9 Matrix Multiplication


Let 𝐴 be an 𝑚 × 𝑛 matrix, let 𝐵 be an 𝑛 × 𝑘 matrix, and write 𝐵 =
𝒃1 𝒃2 ⋯ 𝒃𝑘 where 𝒃𝑗 is column j of 𝐵 for each 𝑗. The product matrix
𝐴𝐵 is the 𝑚 × 𝑘 matrix defined as follows:
𝐴𝐵 = 𝐴 𝒃1 , 𝒃2 , ⋯ , 𝒃𝑘 = 𝐴𝒃1 , 𝐴𝒃2 , ⋯ , 𝐴𝒃𝑘

Theorem 1
Let 𝐴 be an 𝑚 × 𝑛 matrix and let 𝐵 be an 𝑛 × 𝑘 matrix. Then the product
matrix 𝐴𝐵 is 𝑚 × 𝑘 and satisfies

𝐴(𝐵𝐱) = (𝐴𝐵)𝐱 for all 𝐱 in ℝ𝑘

26
SECTION 2 .3 Matrix Multiplication

EXAMPLE 1
2 3 5 8 9
Compute 𝐴𝐵 if 𝐴 = 1 4 7 and 𝐵 = 7 2 .
0 1 8 6 1

Theorem 1 Dot Product Rule


Let A and B be matrices of sizes m × n and n × k,
respectively. Then the (i, j)-entry of AB is the dot
product of row i of A with column j of B.

27
SECTION 2 .3 Matrix Multiplication
Compatibility Rule
• Let A and B denote matrices. If A is m × n and B is n’ × k, the product AB
can be formed if and only if n = n’.
• In this case the size of the product matrix AB is m × k, and we say that AB
is defined, or that A and B are compatible for multiplication.

Convention
Whenever a product of matrices is written, it is tacitly assumed that the
sizes of the factors are such that the product is defined.

28
SECTION 2 .3 Matrix Multiplication

EXAMPLE 2
2 3 5 8 9
Compute 𝐴𝐵 if 𝐴 = 1 4 7 and 𝐵 = 7 2 .
0 1 8 6 1

EXAMPLE 3
Compute the (1,3) - and (3,4)-entries of 𝐴𝐵 where
2 1 6 0
3 −1 2
𝐴= and 𝐵 = 0 2 3 4 .
0 1 4
−1 0 5 8

29
SECTION 2 .3 Matrix Multiplication
EXAMPLE 4
5
If 𝐴 = 1 3 2 and 𝐵 = 6 , compute 𝐴2 , 𝐴𝐵, 𝐵𝐴, and 𝐵2 when they are
4
defined.

EXAMPLE 5
6 9 1 2
Let 𝐴 = and 𝐵 = . Compute 𝐴2 , 𝐴𝐵, 𝐵𝐴.
−4 −6 −1 0

EXAMPLE 6
If A is any matrix, then IA = A and AI = A, and where I denotes an identity
matrix of a size so that the multiplications are defined.
30
SECTION 2 .3 Matrix Multiplication
Theorem 3
Assume that 𝑎 is any scalar, and that 𝐴, 𝐵, and 𝐶 are matrices of sizes such
that the indicated matrix products are defined. Then:
1. 𝐼𝐴 = 𝐴 and 𝐴𝐼 = 𝐴 where I denotes an identity matrix.
2. 𝐴(𝐵𝐶) = (𝐴𝐵)𝐶
3. 𝐴(𝐵 + 𝐶) = 𝐴𝐵 + 𝐴𝐶.
4. (𝐵 + 𝐶)𝐴 = 𝐵𝐴 + 𝐶𝐴.
5. 𝑎(𝐴𝐵) = (𝑎𝐴)𝐵 = 𝐴(𝑎𝐵).
6. (𝐴𝐵)𝑇 = 𝐵𝑇 𝐴𝑇

EXAMPLE 7
Simplify the expression A(BC - CD) + A(C - B)D - AB(C - D).
31
SECTION 2 .3 Matrix Multiplication
Block Multiplication
Definition 2.10
It is often useful to consider matrices whose entries are themselves matrices
(called blocks). A matrix viewed in this way is said to be partitioned into blocks.

32
SECTION 2 .3 Matrix Multiplication
• In other words, we can compute the product AB by ordinary matrix
multiplication, using blocks as entries. The only requirement is that the
blocks be compatible.
• That is, the sizes of the blocks must be such that all (matrix) products of
blocks that occur make sense.
• This means that the number of columns in each block of A must equal the
number of rows in the corresponding block of B.

Theorem 4 Block Multiplication


If matrices A and B are partitioned compatibly into blocks, the product AB can
be computed by matrix multiplication using blocks as entries.

33
SECTION 2 .3 Matrix Multiplication
Theorem 5
𝐵 𝑋 𝐵1 𝑋1
Suppose matrices 𝐴 = and 𝐴1 = are partitioned as shown
0 𝐶 0 𝐶1
where 𝐵 and 𝐵1 are square matrices of the same size, and 𝐶 and 𝐶1 are also
square of the same size. These are compatible partitionings and block
multiplication gives
𝐵 𝑋 𝐵1 𝑋1 𝐵𝐵1 𝐵𝑋1 + 𝑋𝐶1
𝐴𝐴1 = =
0 𝐶 0 𝐶1 0 𝐶𝐶1

EXAMPLE 2
𝐼 𝑋
Obtain a formula for 𝐴𝑘 where 𝐴 = is square and 𝐼 is an identity
0 0
matrix. 34
SECTION 2 .4 Matrix Inverses
Definition 2.11
This suggests the following definition. If 𝐴 is a square matrix, a matrix 𝐵 is
called an inverse of 𝐴 if and only if
𝐴𝐵 = 𝐼 and 𝐵𝐴 = 𝐼
A matrix 𝐴 that has an inverse is called an invertible matrix.

EXAMPLE 1
−1 1 0 1
Show that 𝐵 = is an inverse of 𝐴 = .
1 0 1 1

EXAMPLE 2
0 0
Show that 𝐴 = has no inverse.
1 3 35
SECTION 2 .4 Matrix Inverses
Theorem 1
If B and C are both inverses of A, then B = C.

If A is an invertible matrix, the (unique) inverse of A is denoted 𝐴−1 .

EXAMPLE 3
0 −1
If 𝐴 = , show that 𝐴3 = 𝐼 and so find 𝐴−1 .
1 −1
𝑎 𝑏
Let 𝐴 = . We define the determinant det 𝐴 and the adjugate adj 𝐴 of
𝑐 𝑑
the matrix 𝐴 as follows:
𝑎 𝑏 𝑎 𝑏 𝑑 −𝑏
det = 𝑎𝑑 − 𝑏𝑐, and adj =
𝑐 𝑑 𝑐 𝑑 −𝑐 𝑎 36
SECTION 2 .4 Matrix Inverses

EXAMPLE 4
𝑎 𝑏
If 𝐴 = , show that 𝐴 has an inverse if and only if det 𝐴 ≠ 0, and in this
𝑐 𝑑
−1 1
case 𝐴 = adj 𝐴
det 𝐴

Inverses and Linear Systems


Theorem 2
Suppose a system of n equations in 𝑛 variables is written in matrix form as
𝐴𝑥 = 𝑏
If the 𝑛 × 𝑛 coefficient matrix 𝐴 is invertible, the system has the unique
solution
𝒙 = 𝐴−1 𝒃 37
SECTION 2 .4 Matrix Inverses

EXAMPLE 5
5𝑥1 − 3𝑥2 = −4
Use Example 4 to solve the system ቊ
7𝑥1 + 4𝑥2 = 8

An Inversion Method
Matrix Inversion Algorithm
If 𝐴 is an invertible (square) matrix, there exists a sequence of elementary row
operations that carry 𝐴 to the identity matrix I of the same size, written 𝐴 → 𝐼.
This same series of row operations carries 𝐼 to 𝐴−1 ; that is, 𝐼 → 𝐴−1 . The
algorithm can be summarized as follows:
[𝐴 𝐼] → 𝐼 𝐴−1
where the row operations on 𝐴 and 𝐼 are carried out simultaneously.
38
SECTION 2 .4 Matrix Inverses
EXAMPLE 6
Use the inversion algorithm to find the inverse of the matrix
2 7 1
𝐴 = 1 4 −1
1 3 0

Theorem 3
If 𝐴 is an 𝑛 × 𝑛 matrix, either 𝐴 can be reduced to I by elementary row
operations or it cannot. In the first case, the algorithm produces 𝐴−1 ; in the
second case, 𝐴−1 does not exist.

39
SECTION 2 .4 Matrix Inverses

EXAMPLE 7
Cancellation Laws Let 𝐴 be an invertible matrix. Show that:
(1) If 𝐴𝐵 = 𝐴𝐶, then 𝐵 = 𝐶.
(2) If 𝐵𝐴 = 𝐶𝐴, then 𝐵 = 𝐶.

EXAMPLE 8
If 𝐴 is an invertible matrix, show that the transpose 𝐴𝑇 is also invertible. Show
further that the inverse of 𝐴𝑇 is just the transpose of 𝐴−1 ; in symbols,
𝐴𝑇 −1 = 𝐴−1 𝑇

EXAMPLE 9
If 𝐴 and 𝐵 are invertible 𝑛 × 𝑛 matrices, show that their product 𝐴𝐵 is also
invertible and (𝐴𝐵)−1 = 𝐵−1 𝐴−1 . 40
SECTION 2 .4 Matrix Inverses
Theorem 4
All the following matrices are square matrices of the same size.
1. 𝐼 is invertible and 𝐼 −1 = 𝐼.
2. If 𝐴 is invertible, so is 𝐴−1 , and 𝐴−1 −1 = 𝐴.
3. If 𝐴 and 𝐵 are invertible, so is 𝐴𝐵, and (𝐴𝐵)−1 = 𝐵−1 𝐴−1 .
4. If 𝐴1 , 𝐴2 , … , 𝐴𝑘 are all invertible, so is their product 𝐴1 𝐴2 ⋯ 𝐴𝑘 , and
𝐴1 𝐴2 ⋯ 𝐴𝑘 −1 = 𝐴−1 𝑘 ⋯ 𝐴−1 −1
2 𝐴1 .
−1
5. If 𝐴 is invertible, so is 𝐴𝑘 for any 𝑘 ≥ 1, and 𝐴𝑘 = 𝐴−1 𝑘 .
6. If 𝐴 is invertible and 𝑎 ≠ 0 is a number, then 𝑎𝐴 is invertible and (𝑎𝐴)−1 =
1 −1
𝐴 .
𝑎
7. If 𝐴 is invertible, so is its transpose 𝐴𝑇 , and 𝐴𝑇 −1 = 𝐴−1 𝑇 .
41
SECTION 2 .4 Matrix Inverses
Corollary 1
A square matrix 𝐴 is invertible if and only if 𝐴𝑇 is invertible.
EXAMPLE 10
𝑇 −1 2 1
Find 𝐴 if 𝐴 − 2𝐼 =
−1 0

42
SECTION 2 .4 Matrix Inverses
Theorem 5 Inverse Theorem
The following conditions are equivalent for an 𝑛 × 𝑛 matrix 𝐴 :
1. 𝐴 is invertible.
2. The homogeneous system 𝐴𝒙 = 0 has only the trivial solution 𝒙 = 𝟎.
3. 𝐴 can be carried to the identity matrix 𝐼𝑛 by elementary row operations.
4. The system 𝐴𝒙 = 𝒃 has at least one solution 𝒙 for every choice of column 𝒃.
5. There exists an 𝑛 × 𝑛 matrix 𝐶 such that 𝐴𝐶 = 𝐼𝑛 .

43
SECTION 2 .4 Matrix Inverses
Corollary 1
If 𝐴 and 𝐶 are square matrices such that 𝐴𝐶 = 𝐼, then also 𝐶𝐴 = 𝐼. In
particular, both 𝐴 and 𝐶 are invertible, 𝐶 = 𝐴−1 , and 𝐴 = 𝐶 −1

Corollary 2
An 𝑛 × 𝑛 matrix 𝐴 is invertible if and only if rank 𝐴 = 𝑛.

44
SECTION 2 .5 Elementary Matrices
Definition 2.12
An 𝑛 × 𝑛 matrix E is called an elementary matrix if it can be obtained from
the identity matrix 𝐼𝑛 by a single elementary row operation (called the
operation corresponding to E). We say that E is of type I, II, or III if the
operation is of that type.

Lemma 1
If an elementary row operation is performed on an 𝑚 × 𝑛 matrix A, the result
is EA where E is the elementary matrix obtained by performing the same
operation on the 𝑚 × 𝑚 identity matrix.

45
SECTION 2 .5 Elementary Matrices
Lemma 2
Every elementary matrix E is invertible, and 𝐸 −1 is also a elementary matrix
(of the same type). Moreover, 𝐸 −1 corresponds to the inverse of the row
operation that produces E

EXAMPLE 1
Find the inverse of each of the elementary matrices
0 1 0 1 0 0 1 0 5
𝐸1 = 1 0 0 , 𝐸2 = 0 1 0 , and 𝐸3 = 0 1 0 .
0 0 1 0 0 9 0 0 1 46
SECTION 2 .5 Elementary Matrices

Theorem 1
Suppose 𝐴 is 𝑚 × 𝑛 and 𝐴 → 𝐵 by elementary row operations.
1. 𝐵 = 𝑈𝐴 where 𝑈 is an 𝑚 × 𝑚 invertible matrix.
2. 𝑈 can be computed by 𝐴𝐼𝑚 → [𝐵 𝑈] using the operations carrying 𝐴 → 𝐵.
3. 𝑈 = 𝐸𝑘 𝐸𝑘−1 ⋯ 𝐸2 𝐸1 where 𝐸1 , 𝐸1 , … , 𝐸𝑘 are the elementary matrices
corresponding (in order) to the elementary row operations carrying 𝐴 to 𝐵.
𝐴 → 𝐸1 𝐴 → 𝐸2 𝐸1 𝐴 → 𝐸3 𝐸2 𝐸1 𝐴 → … → 𝐸𝑘 𝐸𝑘−1 … 𝐸2 𝐸1 𝐴 = 𝐵
𝑈
EXAMPLE 2
2 3 1
If 𝐴 = , express the reduced row-echelon form 𝑅 of 𝐴 as 𝑅 = 𝑈𝐴
1 2 1
where 𝑈 is invertible.
47
SECTION 2 .5 Elementary Matrices
Theorem 2
A square matrix is invertible if and only if it is a product of elementary
matrices.

EXAMPLE 3
−2 3
Express 𝐴 = as a product of elementary matrices.
1 0
Uniqueness of the Reduced Row-echelon Form
Theorem 4
If a matrix A is carried to reduced row-echelon matrices R and S by row
operations, then R = S.
48
SECTION 2 .7 LU - Factorization
SECTION 2 .7 LU - Factorization
SECTION 2 .7 LU - Factorization
SECTION 2 .7 LU - Factorization
Theorem 1
If A can be lower reduced to a row-echelon matrix U, then
A = LU
where L is lower triangular and invertible and U is upper triangular and row-
echelon.

Definition 2.14
A factorization A = LU as in Theorem 1 is called an LU-factorization of A.
SECTION 2 .7 LU - Factorization
For convenience, the first nonzero column from the left in a
matrix A is called the leading column of A.
LU - Algorithm
Let A be an 𝑚 × 𝑛 matrix of rank r, and suppose that A can be lower reduced to a
row-echelon matrix U. Then A = LU where the lower triangular, invertible matrix L
is constructed as follows:
1. If A = 0, take L = 𝐼𝑚 and U = 0.
2. If A ≠ 0, write 𝐴1 = 𝐴 and let 𝒄1 be the leading column of 𝐴1 . Use 𝒄1 to create
the first leading 1 and create zeros below it (using lower reduction). When this is
completed, let 𝐴2 denote the matrix consisting of rows 2 to m of the matrix just
created.
3. If 𝐴2 ≠ 0, let 𝒄2 be the leading column of 𝐴2 and repeat Step 2 on 𝐴2 to create
𝐴3 .
4. Continue in this way until U is reached, where all rows below the last leading 1
consist of zeros. This will happen after r steps.
5. Create L by placing 𝒄1 , 𝒄2 , … , 𝒄𝑟 at the bottom of the first r columns of 𝐼𝑚 .
SECTION 2 .7 LU - Factorization
SECTION 2 .7 LU - Factorization
SECTION 2 .7 LU - Factorization
SECTION 2 .7 LU - Factorization
SECTION 2 .8 An Application to Input-Output Economic Models

• In 1973 Wassily Leontief was awarded the Nobel prize in economics for his
work on mathematical models.
• Roughly speaking, an economic system in this model consists of several
industries, each of which produces a product and each of which uses some
of the production of the other industries.
The following example is typical.

OUTPUT
Farming Housing Garment
CONSUMPTION Farming 0.4 0.2 0.3
Housing 0.2 0.6 0.4
Garment 0.4 0.2 0.3
58
SECTION 2 .8 An Application to Input-Output Economic Models

EXAMPLE 1
A primitive society has three basic needs: food, shelter, and clothing. There are
thus three industries in the society-the farming, housing, and garment
industries-that produce these commodities. Each of these industries consumes
a certain proportion of the total output of each commodity according to the
following table.
OUTPUT
Farming Housing Garment
CONSUMPTION Farming 0.4 0.2 0.3
Housing 0.2 0.6 0.4
Garment 0.4 0.2 0.3 59
SECTION 2 .8 An Application to Input-Output Economic Models

• In general, suppose an economy has n industries. We assume first that the


economy is closed (that is, no product is exported or imported).
• Given two industries i and j, let 𝑒𝑖𝑗 denote the proportion of the total
annual output of industry j that is consumed by industry i. Then E = [𝑒𝑖𝑗 ]
is called the input-output matrix for the economy. Clearly
1) 0 ≤ 𝑒𝑖𝑗 ≤ 1 for all 𝑖 and 𝑗.
2) 𝑒1𝑗 + 𝑒2𝑗 + ⋯ + 𝑒𝑛𝑗 = 1 for each 𝑗.
• Matrices satisfying conditions 1) and 2) are called stochastic matrices.
• The closed economic system is said to be in equilibrium if the annual
expenditure equals the annual revenue for each industry—that is, if
𝑒1𝑗 𝑝1 + 𝑒2𝑗 𝑝2 + ⋯ + 𝑒𝑖𝑗 𝑝𝑛 = 𝑝𝑖 for each 𝑖 = 1,2, … , 𝑛 60
SECTION 2 .8 An Application to Input-Output Economic Models

• If we write 𝐩 = 𝑝1 … 𝑝𝑛 𝑇 , these equations can be written as the matrix


equation
𝐸𝐩 = 𝐩
This is called the equilibrium condition, and the solutions 𝐩 are called
equilibrium price structures.

Theorem 1
Let 𝐸 be any 𝑛 × 𝑛 stochastic matrix. Then there is a nonzero 𝑛 × 1 matrix 𝒑
with nonnegative entries such that 𝐸𝒑 = 𝒑. If all the entries of 𝐸 are positive,
the matrix 𝒑 can be chosen with all entries positive.

61
SECTION 2 .8 An Application to Input-Output Economic Models
EXAMPLE 2
Find the equilibrium price structures for four industries if the input-output
matrix is
.6 .2 .1 .1
.3 .4 .2 0
𝐸=
.1 .3 .5 .2
0 .1 .2 .7
Find the prices if the total value of business is $1000.

The Open Model


We now assume that there is a demand for products in the open sector of the
economy, which is the part of the economy other than the producing industries
(for example, consumers). Let 𝒅𝒊 denote the total value of the demand for
product i in the open sector. 62
SECTION 2 .8 An Application to Input-Output Economic Models
• So the total annual revenue 𝑝𝑖 of industry i breaks down as follows:
𝑝𝑖 = (𝑒𝑖1 𝑝1 + 𝑒𝑖2 𝑝2 + ⋯ + 𝑒𝑖𝑛 𝑝𝑛 ) + 𝑑𝑖 for each 𝑖 = 1, 2, … , 𝑛.
• The column 𝒅 = 𝑑1 … 𝑑𝑛 𝑇 is called the demand matrix, and this gives a
matrix equation
𝒑 = 𝐸𝒑 + 𝒅 or 𝐼 − 𝐸 𝒑 = 𝒅.
• If 𝐴 = [𝑎𝑖𝑗 ] and 𝐵 = [𝑏𝑖𝑗 ] are matrices of the same size,
• We write 𝐴 > 𝐵 if 𝑎𝑖𝑗 > 𝑏𝑖𝑗 for all 𝑖 and 𝑗
• We write 𝐴 ≥ 𝐵 if 𝑎𝑖𝑗 ≥ 𝑏𝑖𝑗 for all 𝑖 and 𝑗.
Theorem 2
Let 𝐸 ≥ 0 be a square matrix. Then 𝐼 − 𝐸 is invertible and 𝐼 − 𝐸 −1 ≥ 0 if
and only if there exists a column 𝒑 > 0 such that 𝒑 > 𝐸𝒑.
63
SECTION 2 .8 An Application to Input-Output Economic Models

EXAMPLE 3
0.6 0.2 0.3
If 𝐸 = 0.1 0.4 0.2 , show that 𝐼 − 𝐸 is invertible and (𝐼 − 𝐸)−1 ≥ 0
0.2 0.5 0.1

64

You might also like