ELEMENTARY MATRICES 1

ELEMENTARY MATRICES AND ELEMENTARY

ROW OPERATIONS
Section 1.2 introduced the three elementary row operations for matrices listed below.
1. Interchange two rows.
2. Multiply a row by a nonzero constant.
3. Add a multiple of a row to another row.
In this section, you will see how to use matrix multiplication to perform these operations.

Definition of an Elementary Matrix

An n × n matrix is an elementary matrix when it can be obtained from the
identity matrix In by a single elementary row operation.

Elementary Matrices and

Nonelementary Matrices
Which of the matrices below are elementary? For those that are, describe the
corresponding elementary row operation.

[ ]
1 0 0
[10 ]
0 0
a. 0 3 0 b.
1 0
0 0 1

[ ] [ ]
1 0 0 1 0 0
c. 0 1 0 d. 0 0 1
0 0 0 0 1 0

[ ]
1 0 0
[ ]
1 0
e. f. 0 2 0
2 1
0 0 −1

SOLUTION

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 ELEMENTARY MATRICES 2

Elementary matrices are useful because they enable you to use matrix multiplication
to perform elementary row operations, as demonstrated in Example 2.

Elementary Matrices and

Elementary Row Operations
a. In the matrix product below, E is the elementary matrix in which the first two rows
of I3 are interchanged.
E A

[ ][ ] [ ]
0 1 0 0 2 1 1 −3 6
1 0 0 1 −3 6 = 0 2 1
0 0 1 3 2 −1 3 2 −1
Note that the first two rows of A are interchanged when multiplying on the left by E.
b. In the matrix product below, E is the elementary matrix in which the second row of
I3 is multiplied by 12.
E A

[ ][ ] [ ]
1 0 0 1 0 −4 1 1 0 −4 1
1
0 2 0 0 2 6 −4 = 0 1 3 −2
0 0 1 0 1 3 1 0 1 3 1
Note that the second row of A is multiplied by 12 when multiplying on the left by E.
c. In the matrix product below, E is the elementary matrix in which 2 times the first
row of I3 is added to the second row.
E A

[ ][ ] [ ]
1 0 0 1 0 −1 1 0 −1
2 1 0 −2 −2 3 = 0 −2 1
0 0 1 0 4 5 0 4 5
Note that 2 times the first row of A is added to the second row when multiplying on
the left by E.

THEOREM Representing Elementary Row Operations

Let E be the elementary matrix obtained by performing an elementary row operation on Im. If that
same elementary row operation is performed on an m × n matrix A, then the resulting matrix is the product
EA.

Most applications of elementary row operations require a sequence of operations. For instance, Gaussian
elimination usually requires several elementary row operations to row reduce a matrix. This translates
into multiplication on the left by several elementary matrices. The order of multiplication is important;
the elementary matrix immediately to the left of A corresponds to the row operation performed first. Example 3
demonstrates this process.

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 ELEMENTARY MATRICES 3

Using Elementary Matrices

Find a sequence of elementary matrices that can be used to write the matrix A in
row-echelon form.

[ ]
0 1 3 5
A= 1 −3 0 2
2 −6 2 0

SOLUTION
Elementary Row Elementary
Matrix Operation Matrix
R1 ↔ R2

[ ] [ ]
1 −3 0 2 0 1 0
0 1 3 5 E1 = 1 0 0
2 −6 2 0 0 0 1

[ ] [ ]
1 −3 0 2 1 0 0
0 1 3 5 R3 + (−2)R1 → R3 E2 = 0 1 0
0 0 2 −4 −2 0 1

[ ] [ ]
1 −3 0 2 1 0 0
0 1 3 5 E3 = 0 1 0
0 0 1 −2 (12 )R3 → R3 0 0 1
2

The three elementary matrices E1, E2, and E3 can be used to perform the same elimination.

[ ][ ][ ][ ]
1 0 0 1 0 0 0 1 0 0 1 3 5
B = E3E2E1A = 0 1 0 0 1 0 1 0 0 1 −3 0 2
1
0 0 2 −2 0 1 0 0 1 2 −6 2 0

[ ][ ][ ]
1 0 0 1 0 0 1 −3 0 2
= 0 1 0 0 1 0 0 1 3 5
1
0 0 2 −2 0 1 2 −6 2 0

[ ][ ] [ ]
1 0 0 1 −3 0 2 1 −3 0 2
= 0 1 0 0 1 3 5 = 0 1 3 5
1
0 0 2 0 0 2 −4 0 0 1 −2

Definition of Row Equivalence

Let A and B be m × n matrices. Matrix B is row-equivalent to A when there
exists a finite number of elementary matrices E1, E2, . . . , Ek such that
B = EkEk−1 . . . E2E1A.

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 ELEMENTARY MATRICES 4

THEOREM Elementary Matrices Are Invertible

If E is an elementary matrix, then E−1 exists and is an elementary matrix.

The inverse of an elementary matrix E is the elementary matrix that converts E

back to In. For instance, the inverses of the three elementary matrices in Example 3 are
shown below.
Elementary Matrix Inverse Matrix
R1 ↔ R2 R1 ↔ R2

[ ] [ ]
0 1 0 0 1 0
E1 = 1 0 0 E1−1 = 1 0 0
0 0 1 0 0 1

[ ] [ ]
1 0 0 1 0 0
E2 = 0 1 0 E2−1 = 0 1 0
−2 0 1 R3 + (−2)R1 → R3 2 0 1 R3 + (2)R1 → R3

[ ] [ ]
1 0 0 1 0 0
E3 = 0 1 0 E3−1 = 0 1 0
0 0 1
2 (12 )R3 → R3 0 0 2 (2)R3 → R3

Use matrix multiplication to check these results.

The next theorem states that every invertible matrix can be written as the product
of elementary matrices.

THEOREM A Property of Invertible Matrices

A square matrix A is invertible if and only if it can be written as the product of
elementary matrices.

PROOF
The phrase “if and only if” means that there are actually two parts to the theorem.
On the one hand, you have to show that if A is invertible, then it can be written as the
product of elementary matrices. Then you have to show that if A can be written as
the product of elementary matrices, then A is invertible.
To prove the theorem in one direction, assume A is invertible. From Theorem
2.11 you know that the system of linear equations represented by Ax = O has only the
trivial solution. But this implies that the augmented matrix [A O] can be rewritten in
the form [I O] (using elementary row operations corresponding to E1, E2, . . . , and
Ek). So, Ek . . . E2E1A = I and it follows that A = E1−1E2−1 . . . Ek−1. A can be written
as the product of elementary matrices.
To prove the theorem in the other direction, assume A is the product of elementary
matrices. Every elementary matrix is invertible and the product of invertible matrices
is invertible, so it follows that A is invertible. This completes the proof.

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 ELEMENTARY MATRICES 5

Writing a Matrix as the Product

of Elementary Matrices
Find a sequence of elementary matrices whose product is the nonsingular matrix

[−13 −2
A=
8
. ]
SOLUTION
Begin by finding a sequence of elementary row operations that can be used to rewrite A in reduced row-echelon
form.
Matrix Elementary Row Operation Elementary Matrix
(−1)R1 → R1 −1
[3 ] [ ]
1 2 0
E1 =
8 0 1

[10 2
2 ] R2 + (−3)R1 → R2
E2 = [−31 0
1 ]
[10 2
1 ] ( )R2 → R2
1
2
E3 = [10 0
1
2
]
R1 + (−2)R2 → R1 −2
[0 ] [0 ]
1 0 1
E4 =
1 1
Now, from the matrix product E4E3E2E1A = I, solve for A to obtain
A = E1−1E2−1E3−1E4−1. This implies that A is a product of elementary matrices.
E1−1 E2−1 E3−1 E4−1

[−10 −1 −2
][13 ][10 ][10 ] [ ]
0 0 0 2
A= =
1 1 2 1 3 8

THEOREM Equivalent Conditions

If A is an n × n matrix, then the statements below are equivalent.
1. A is invertible.
2. Ax = b has a unique solution for every n × 1 column matrix b.
3. Ax = O has only the trivial solution.
4. A is row-equivalent to In.
5. A can be written as the product of elementary matrices.

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 ELEMENTARY MATRICES 6

THE LU-FACTORIZATION
At the heart of the most efficient and modern algorithms for solving linear systems Ax = b is the LU-
factorization, in which the square matrix A is expressed as a product, A = LU. In this product, the square
matrix L is lower triangular, which means all the entries above the main diagonal are zero. The square
matrix U is upper triangular, which means all the entries below the main diagonal are zero.

[ ] [ ]
a11 0 0 a11 a12 a13
a21 a22 0 0 a22 a23
a31 a32 a33 0 0 a33

3 × 3 lower triangular matrix 3 × 3 upper triangular matrix

Definition of LU-Factorization
If the n × n matrix A can be written as the product of a lower triangular matrix
L and an upper triangular matrix U, then A = LU is an LU-factorization of A.

LU-Factorizations

[11 ] [ ][10 ]
2 1 0 2
a. = = LU
0 1 1 −2
is an LU-factorization of the matrix

[11 ]
2
A=
0
as the product of the lower triangular matrix

[11 ]
0
L=
1
and the upper triangular matrix

[10 ]
2
U= .
−2

[ ] [ ][ ]
1 −3 0 1 0 0 1 −3 0
b. A = 0 1 3 = 0 1 0 0 1 3 = LU
2 −10 2 2 −4 1 0 0 14

is an LU-factorization of the matrix A.

Computational fluid dynamics (CFD) is the computer-based

LINEAR simulation of such real-life phenomena as fluid flow, heat
ALGEBRA transfer, and chemical reactions. Solving the conservation
APPLIED of energy, mass, and momentum equations involved in a
CFD analysis can involve large systems of linear equations.
So, for efficiency in computing, CFD analyses often use
matrix partitioning and LU-factorization in their algorithms.
Aerospace companies such as Boeing and Airbus have
used CFD analysis in aircraft design. For instance,
engineers at Boeing used CFD analysis to simulate
airflow around a virtual model of their 787 aircraft to help
produce a faster and more efficient design than those of
earlier Boeing aircraft.
Goncharuk/Shutterstock.com

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 ELEMENTARY MATRICES 7

Finding an LU-Factorization of a Matrix

[ ]
1 −3 0
Find an LU-factorization of the matrix A = 0 1 3 .
2 −10 2

SOLUTION
Begin by row reducing A to upper triangular form while keeping track of the elementary
matrices used for each row operation.
Matrix Elementary Row Operation Elementary Matrix

[ ] [ ]
1 −3 0 1 0 0
0 1 3 E1 = 0 1 0
0 −4 2 R3 + (−2)R1 → R3 −2 0 1

[ ] [ ]
1 −3 0 1 0 0
0 1 3 E2 = 0 1 0
0 0 14 R3 + (4)R2 → R3 0 4 1

The reduced matrix above is an upper triangular matrix U, and it follows that
E2E1A = U, or A = E1−1E2−1U. The product of the lower triangular matrices

[ ][ ] [ ]
1 0 0 1 0 0 1 0 0
E1−1E2−1 = 0 1 0 0 1 0 = 0 1 0
2 0 1 0 −4 1 2 −4 1
is a lower triangular matrix L, so the factorization A = LU is complete. Notice that this
is the same LU-factorization as in Example 5(b).

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 ELEMENTARY MATRICES 8

Solving a Linear System Using

LU-Factorization
See LarsonLinearAlgebra.com for an interactive version of this type of example.

Solve the linear system.

x1 − 3x2 = −5
x2 + 3x3 = −1
2x1 − 10x2 + 2x3 = −20
SOLUTION
You obtained an LU-factorization of the coefficient matrix A in Example 6.

[ ]
1 −3 0
A= 0 1 3
2 −10 2

[ ][ ]
1 0 0 1 −3 0
= 0 1 0 0 1 3
2 −4 1 0 0 14
First, let y = Ux and solve the system Ly = b for y.

[ ][ ] [ ]
1 0 0 y1 −5
0 1 0 y2 = −1
2 −4 1 y3 −20
Solve this system using forward substitution. Starting with the first equation, you have
y1 = −5. The second equation gives y2 = −1. Finally, from the third equation,
2y1 − 4y2 + y3 = −20
y3 = −20 − 2y1 + 4y2
y3 = −20 − 2(−5) + 4(−1)
y3 = −14.
The solution of Ly = b is

[ ]
−5
y = −1 .
−14
Now solve the system Ux = y for x using back-substitution.

[ ][ ] [ ]
1 −3 0 x1 −5
0 1 3 x2 = −1
0 0 14 x3 −14
From the bottom equation, x3 = −1. Then, the second equation gives
x2 + 3(−1) = −1
or x2 = 2. Finally, the first equation gives
x1 − 3(2) = −5
or x1 = 1. So, the solution of the original system of equations is

[ ]
1
x= 2 .
−1

Copyright 2017 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.