CHAPTER 2

Determinants
CHAPTER CONTENTS

2.1 Determinants by Cofactor Expansion 118

2.2 Evaluating Determinants by Row Reduction 126
2.3 Properties of Determinants; Cramer’s Rule 133

Introduction
In this chapter we will study “determinants” or, more precisely, “determinant functions.”
Unlike real-valued functions, such as 𝑓(x) = x 2 , that assign a real number to a real vari-
able x, determinant functions assign a real number 𝑓(𝐴) to a matrix variable 𝐴. Although
determinants first arose in the context of solving systems of linear equations, they are
rarely used for that purpose in real-world applications. While they can be useful for solv-
ing very small linear systems (say, two or three unknowns), our main interest in them
stems from the fact that they link together various concepts in linear algebra and provide
a useful formula for the inverse of a matrix.

2.1 Determinants by Cofactor Expansion

In this section we will define the notion of a “determinant.” This will enable us to develop
a specific formula for the inverse of an invertible matrix, whereas up to now we have had
only a computational procedure. This, in turn, will eventually provide us with a formula
for solutions of certain kinds of linear systems.

Recall from Theorem 1.4.5 that the 2 × 2 matrix

a b
𝐴=[ ]
c d
is invertible if and only if ad − bc ≠ 0 and that the expression ad − bc is called the deter-
Warning It is important minant of the matrix 𝐴. Recall also that this determinant is denoted by writing
to keep in mind that det(A)
|a b|
is a number, whereas A is a det(𝐴) = ad − bc or | | = ad − bc (1)
matrix. |c d|
and that the inverse of 𝐴 can be expressed in terms of the determinant as

1 d −b
𝐴−1 = [ ] (2)
det(𝐴) −c a
118
 2.1 Determinants by Cofactor Expansion 119

Minors and Cofactors

One of our main goals in this chapter is to obtain an analog of Formula (2) that is appli-
cable to square matrices of all orders. For this purpose we will find it convenient to use
subscripted entries when writing matrices or determinants. Thus, if we denote a 2 × 2
matrix as
a a12
𝐴 = [ 11 ]
a21 a22
then the two equations in (1) take the form
|a a12 |
det(𝐴) = | 11 | = a11 a22 − a12 a21 (3)
| a21 a22 |
In situations where it is inconvenient to assign a name to the matrix, we can express this
formula as
a a12
det [ 11 ] = a11 a22 − a12 a21 (4)
a21 a22
There are various methods for defining determinants of higher-order square matrices.
In this text, we will use an “inductive definition” by which we mean that the determinant
of a square matrix of a given order will be defined in terms of determinants of square
matrices of the next lower order. To start the process, let us define the determinant of a
1 × 1 matrix [a11 ] as
det [a11 ] = a11 (5)
from which it follows that Formula (4) can be expressed as
a11 a12
det [ ] = det[a11 ] det[a22 ] − det[a12 ] det[a21 ]
a21 a22
Now that we have established a starting point, we can define determinants of 3 × 3
matrices in terms of determinants of 2 × 2 matrices, then determinants of 4 × 4 matri-
ces in terms of determinants of 3 × 3 matrices, and so forth, ad infinitum. The following
terminology and notation will help to make this inductive process more efficient.

Definition 1

If 𝐴 is a square matrix, then the minor of entry aij is denoted by 𝑀ij and is defined
to be the determinant of the submatrix that remains after the ith row and jth col-
umn are deleted from 𝐴. The number (−1)i+j 𝑀ij is denoted by 𝐶 ij and is called the
cofactor of entry aij .

Historical Note
The term determinant was first introduced by the German mathematician Carl Friedrich
Gauss in 1801 (see p. 16), who used them to “determine” properties of certain kinds of func-
tions. Interestingly, the term matrix is derived from a Latin word for “womb” because it was
viewed as a container of determinants.

EXAMPLE 1 | Finding Minors and Cofactors

Let
3 1 −4
𝐴 = [2 5 6]
1 4 8
120 C H APT ER 2 Determinants

The minor of entry a11 is

Warning We have fol-
lowed the standard con- 3 1 4
vention of using capital 5 6
M11 = 2 5 6 = = 16
letters to denote minors and 4 8
cofactors even though they
1 4 8
are numbers, not matrices.
The cofactor of a11 is
𝐶11 = (−1)1+1 𝑀11 = 𝑀11 = 16
Similarly, the minor of entry a32 is

3 1 4
3 4
M32 = 2 5 6 = = 26
2 6
1 4 8

The cofactor of a32 is

𝐶32 = (−1)3+2 𝑀32 = −𝑀32 = −26

Remark Note that a minor 𝑀ij and its corresponding cofactor 𝐶 ij are either the same or
negatives of each other and that the relating sign (−1)i+j is either +1 or −1 in accordance
with the pattern in the “checkerboard” array
+ − + − + ⋅⋅⋅
⎡ ⎤
⎢− + − + − ⋅ ⋅ ⋅⎥
⎢+ − + − + ⋅ ⋅ ⋅⎥
⎢ ⎥
⎢− + − + − ⋅ ⋅ ⋅⎥
⎢. .. .. .. .. ⎥
.
⎣. . . . . ⎦
For example,
𝐶 11 = 𝑀11 , 𝐶 21 = −𝑀21 , 𝐶 22 = 𝑀22
and so forth. Thus, it is never really necessary to calculate (−1)i+j to obtain 𝐶 ij —you can
simply compute the minor 𝑀ij and then adjust the sign in accordance with the checker-
board pattern. Try this in Example 1.

EXAMPLE 2 | Cofactor Expansions of a 2 × 2 Matrix

The checkerboard pattern for a 2 × 2 matrix 𝐴 = [aij ] is

+ −
[ ]
− +
so that
𝐶11 = 𝑀11 = a22 𝐶12 = −𝑀12 = −a21
𝐶21 = −𝑀21 = −a12 𝐶22 = 𝑀22 = a11
We leave it for you to use Formula (3) to verify that det(𝐴) can be expressed in terms of
cofactors in the following four ways:
| a11 a12 |
det(𝐴) = | |
| a21 a22 |
= a11 𝐶11 + a12 𝐶12
(6)
= a21 𝐶21 + a22 𝐶22
= a11 𝐶11 + a21 𝐶21
= a12 𝐶12 + a22 𝐶22
 2.1 Determinants by Cofactor Expansion 121

Each of the last four equations is called a cofactor expansion of det(𝐴). In each cofactor
expansion the entries and cofactors all come from the same row or same column of 𝐴. For
example, in the first equation the entries and cofactors all come from the first row of 𝐴, in
the second they all come from the second row of 𝐴, in the third they all come from the first
column of 𝐴, and in the fourth they all come from the second column of 𝐴.

Historical Note
The term minor is apparently due to the English mathematician James Sylvester (see p. 36),
who wrote the following in a paper published in 1850: “Now conceive any one line and any
one column be struck out, we get . . . a square, one term less in breadth and depth than the
original square; and by varying in every possible selection of the line and column excluded,
we obtain, supposing the original square to consist of n lines and n columns, n2 such minor
squares, each of which will represent what I term a “First Minor Determinant” relative to the
principal or complete determinant.”

Definition of a General Determinant

Formula (6) is a special case of the following general result, which we will state without
proof.

Theorem 2.1.1

If 𝐴 is an n × n matrix, then regardless of which row or column of 𝐴 is chosen, the

number obtained by multiplying the entries in that row or column by the corre-
sponding cofactors and adding the resulting products is always the same.

This result allows us to make the following definition.

Definition 2

If 𝐴 is an n × n matrix, then the number obtained by multiplying the entries in any

row or column of 𝐴 by the corresponding cofactors and adding the resulting prod-
ucts is called the determinant of A, and the sums themselves are called cofactor
expansions of A. That is,
det(𝐴) = a1j 𝐶 1j + a2j 𝐶 2j + ⋅ ⋅ ⋅ + anj 𝐶 nj (7)
[cofactor expansion along the jth column]

and
det(𝐴) = ai1 𝐶 i1 + ai2 𝐶 i2 + ⋅ ⋅ ⋅ + ain 𝐶 in (8)
[cofactor expansion along the ith row]

EXAMPLE 3 | Cofactor Expansion Along the First Row

Find the determinant of the matrix

3 1 0
𝐴 = [−2 −4 3]
5 4 −2
by cofactor expansion along the first row.
122 C H APT ER 2 Determinants

Solution
| 3 1 0|
| | | −4 3| | −2 3| | −2 −4 |
det(𝐴) = | −2 −4 3| = 3 | | − 1| | + 0| |
| | | 4 −2 | | 5 −2 | | 5 4|
| 5 4 −2 |
= 3(−4) − (1)(−11) + 0 = −1

Historical Note

Cofactor expansion is not the only method for expressing the

determinant of a matrix in terms of determinants of lower order.
For example, although it is not well known, the English mathe-
matician Charles Dodgson, who was the author of Alice’s Adven-
tures in Wonderland and Through the Looking Glass under the
pen name of Lewis Carroll, invented such a method, called
condensation. That method has recently been resurrected from
obscurity because of its suitability for parallel processing on
computers.
[Image: Oscar G. Rejlander/Time & Life Pictures/
Getty Images]
Charles Lutwidge
Dodgson
(Lewis Carroll)
(1832–1898)

EXAMPLE 4 | Cofactor Expansion Along the First Column

Let 𝐴 be the matrix in Example 3, and evaluate det(𝐴) by cofactor expansion along the first
column of 𝐴.
Solution
| 3 1 0|
| | | −4 3| |1 0| | 1 0|
det(𝐴) = | −2 −4 3| = 3| | − (−2) | | + 5| |
| | | 4 −2 | |4 −2 | | −4 3|
| 5 4 −2 |
= 3(−4) − (−2)(−2) + 5(3) = −1

Note that in Example 4 we This agrees with the result obtained in Example 3.
had to compute three cofac-
tors, whereas in Example
3 only two were needed
because the third was mul-
tiplied by zero. As a rule,
the best strategy for cofac-
EXAMPLE 5 | Smart Choice of Row or Column
tor expansion is to expand
along a row or column with
the most zeros. If 𝐴 is the 4 × 4 matrix
1 0 0 −1
⎡ ⎤
⎢3 1 2 2⎥
𝐴=⎢
⎢1 0 −2 1⎥⎥
⎣2 0 0 1⎦
 2.1 Determinants by Cofactor Expansion 123

then to find det(𝐴) it will be easiest to use cofactor expansion along the second column,
since it has the most zeros:
|1 0 −1 |
| |
det(𝐴) = 1 ⋅ | 1 −2 1|
| |
|2 0 1|
For the 3 × 3 determinant, it will be easiest to use cofactor expansion along its second column,
since it has the most zeros:
|1 −1 |
det(𝐴) = (1)(−2) | |
|2 1|
= −2(1 + 2)
= −6

EXAMPLE 6 | Determinant of a Lower Triangular Matrix

The following computation shows that the determinant of a 4 × 4 lower triangular matrix is
the product of its diagonal entries. Each part of the computation uses a cofactor expansion
along the first row.
| a11 0 0 0 |
| | | a22 0 0 |
| a21 a22 0 0 | |
= a11 | a32 a33
|
0 |
|a a32 a33 0 || | |
| 31 | a42 a43 a44 |
| a41 a42 a43 a44 |
| a33 0 |
= a11 a22 | |
| a43 a44 |
= a11 a22 a33 |a44 | = a11 a22 a33 a44

The method illustrated in Example 6 can be easily adapted to prove the following
general result.

Theorem 2.1.2

If 𝐴 is an n × n triangular matrix (upper triangular, lower triangular, or diagonal ),

then det(𝐴) is the product of the entries on the main diagonal of the matrix; that is,
det(𝐴) = a11 a22 ⋅ ⋅ ⋅ ann .

A Useful Technique for Evaluating 2 × 2 and

3 × 3 Determinants
Determinants of 2 × 2 and 3 × 3 matrices can be evaluated very efficiently using the pat-
tern suggested in Figure 2.1.1.

a11 a12 a11 a12 a13 a11 a12

a21 a22 a21 a22 a23 a21 a22
a31 a32 a33 a31 a32

FIGURE 2.1.1
124 C H APT ER 2 Determinants

In the 2 × 2 case, the determinant can be computed by forming the product of the entries
on the rightward arrow and subtracting the product of the entries on the leftward arrow.
In the 3 × 3 case we first recopy the first and second columns as shown in the figure, after
which we can compute the determinant by summing the products of the entries on the
Warning The arrow rightward arrows and subtracting the products on the leftward arrows. These procedures
technique works only for execute the computations
determinants of 2 × 2 and
| a11 a12 |
3 × 3 matrices. It does not | | = a11 a22 − a12 a21
| a21 a22 |
work for matrices of size
4 × 4 or higher. | a11 a12 a13 |
| | |a a | |a a | |a a |
| a21 a22 a23 | = a11 | 22 23 | − a12 | 21 23 | + a13 | 21 22 |
| | | a32 a33 | | a31 a33 | | a31 a32 |
| a31 a32 a33 |
= a11 (a22 a33 − a23 a32 ) − a12 (a21 a33 − a23 a31 ) + a13 (a21 a32 − a22 a31 )
= a11 a22 a33 + a12 a23 a31 + a13 a21 a32 − a13 a22 a31 − a12 a21 a33 − a11 a23 a32
which agrees with the cofactor expansions along the first row.

EXAMPLE 7 | A Technique for Evaluating 2 × 2 and

3 × 3 Determinants

3 1 3 1
= = (3)( 2) (1)(4) = 10
4 2 4 2

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

= [45 + 84 + 96] [105 48 72] = 240

Exercise Set 2.1

In Exercises 1–2, find all the minors and cofactors of the matrix 𝐴. 4. Let
1 −2 3 1 1 2 2 3 −1 1
⎡ ⎤
1. 𝐴 = [ 6 7 −1] 2. 𝐴 = [3 3 6] ⎢−3 2 0 3⎥
𝐴=⎢
−3 1 4 0 1 4
⎢ 3 −2 1 0⎥⎥
⎣ 3 −2 1 4⎦
3. Let
Find
4 −1 1 6
⎡ ⎤ a. 𝑀32 and 𝐶32 . b. 𝑀44 and 𝐶44 .
⎢0 0 −3 3⎥
𝐴=⎢
⎢4 1 0 14⎥⎥ c. 𝑀41 and 𝐶41 . d. 𝑀24 and 𝐶24 .
⎣4 1 3 2⎦
In Exercises 5–8, evaluate the determinant of the given matrix. If the
Find matrix is invertible, use Equation (2) to find its inverse.
a. 𝑀13 and 𝐶13 . b. 𝑀23 and 𝐶23 . 3 5 4 1 −5 7 √2 √6
5. [ ] 6. [ ] 7. [ ] 8. [ ]
c. 𝑀22 and 𝐶22 . d. 𝑀21 and 𝐶21 . −2 4 8 2 −7 −2 4 √3
 2.1 Determinants by Cofactor Expansion 125

In Exercises 9–14, use the arrow technique of Figure 2.1.1 to evaluate In Exercises 27–32, evaluate the determinant of the given matrix by
the determinant. inspection.

| −2 7 6| 1 0 0 2 0 0
|a − 3 5 | | | 27. [0 −1 0] 28. [0 2 0]
9. | | 10. | 5 1 −2 |
| −3 a − 2| | | 0 0 1 0 0 2
| 3 8 4|
0 0 0 0 1 1 1 1
| −2 1 4| | −1 1 2| ⎡ ⎤ ⎡ ⎤
| | | | ⎢1 2 0 0⎥ ⎢0 2 2 2⎥
11. | 3 5 −7 | 12. | 3 0 −5 | 29. ⎢ 30. ⎢
| | | | ⎢0 4 3 0⎥⎥ ⎢0 0 3 3⎥⎥
| 1 6 2| | 1 7 2|
⎣1 2 3 8⎦ ⎣0 0 0 4⎦
|3 0 0| |c −4 3| 1 2 7 −3 −3 0 0 0
| | | | ⎡ ⎤ ⎡ ⎤
13. | 2 −1 5| 14. | 2 1 c2 | ⎢0 1 −4 1⎥ ⎢ 1 2 0 0⎥
| | | | 31. ⎢ 32. ⎢
|1 9 −4 | |4 c−1 2|
⎢0 0 2 7⎥⎥ ⎢ 40 10 −1 0⎥⎥
In Exercises 15–18, find all values of 𝜆 for which det(𝐴) = 0. ⎣0 0 0 3⎦ ⎣100 200 −23 3⎦
33. In each part, show that the value of the determinant is inde-
𝜆−4 0 0
𝜆−2 1 pendent of 𝜃.
15. 𝐴 = [ ] 16. 𝐴 = [ 0 𝜆 2 ]
−5 𝜆+4 | sin 𝜃 cos 𝜃 |
0 3 𝜆−1 a. | |
| − cos 𝜃 sin 𝜃 |
𝜆−4 4 0
𝜆−1 0 | sin 𝜃 cos 𝜃 0|
17. 𝐴 = [ ] 18. 𝐴 = [ −1 𝜆 0 ] | |
2 𝜆+1 b. | − cos 𝜃 sin 𝜃 0|
0 0 𝜆−5 | |
| sin 𝜃 − cos 𝜃 sin 𝜃 + cos 𝜃 1|
19. Evaluate the determinant in Exercise 13 by a cofactor expan- 34. Show that the matrices
sion along
a b d e
a. the first row. b. the first column. 𝐴=[ ] and 𝐵=[ ]
0 c 0 𝑓
c. the second row. d. the second column.
commute if and only if
e. the third row. f. the third column.
|b a−c |
| |=0
20. Evaluate the determinant in Exercise 12 by a cofactor expan- |e d−𝑓|
sion along
35. By inspection, what is the relationship between the following
a. the first row. b. the first column. determinants?
c. the second row. d. the second column. |a b c| |a + 𝜆 b c|
| | | |
e. the third row. f. the third column. d1 = | d 1 𝑓| and d2 = | d 1 𝑓|
| | | |
|g 0 1| | g 0 1|
In Exercises 21–26, evaluate det(𝐴) by a cofactor expansion along a
row or column of your choice. 36. Show that
−3 0 7 3 3 1 1 | tr(𝐴) 1 |
det(𝐴) = | |
21. 𝐴 = [ 2 5 1] 22. 𝐴 = [1 0 −4] 2 | tr(𝐴2 ) tr(𝐴) |
−1 0 5 1 −3 5 for every 2 × 2 matrix 𝐴.

1 k k2 k+1 k−1 7 37. What can you say about an nth-order determinant all of whose
entries are 1? Explain.
23. 𝐴 = [1 k k2 ] 24. 𝐴 = [ 2 k−3 4]
1 k k2 5 k+1 k 38. What is the maximum number of zeros that a 3 × 3 matrix can
have without having a zero determinant? Explain.
3 3 0 5 39. Explain why the determinant of a matrix with integer entries
⎡ ⎤
⎢2 2 0 −2⎥ must be an integer.
25. 𝐴 = ⎢
⎢4 1 −3 0⎥⎥
⎣2 10 3 2⎦ Working with Proofs
4 0 0 1 0 40. Prove that (x 1 , y1 ), (x 2 , y2 ), and (x 3 , y3 ) are collinear points if
⎡ ⎤ and only if
⎢3 3 3 −1 0⎥
⎢ ⎥
26. 𝐴 = ⎢1 2 4 2 3⎥ | x1 y1 1|
⎢9 | |
⎢ 4 6 2 3⎥⎥ | x2 y2 1| = 0
| |
⎣2 2 4 2 3⎦ | x3 y3 1|
126 C H APT ER 2 Determinants

41. Prove that the equation of the line through the distinct points c. The minor 𝑀ij is the same as the cofactor 𝐶ij if i + j is
(a1 , b1 ) and (a2 , b2 ) can be written as even.

|x y 1| d. If 𝐴 is a 3 × 3 symmetric matrix, then 𝐶ij = 𝐶ji for all i

| | and j.
| a1 b1 1| = 0
| |
| a2 b2 1| e. The number obtained by a cofactor expansion of a matrix
𝐴 is independent of the row or column chosen for the
42. Prove that if 𝐴 is upper triangular and 𝐵ij is the matrix that
expansion.
results when the ith row and jth column of 𝐴 are deleted, then
𝐵ij is upper triangular if i < j. f. If 𝐴 is a square matrix whose minors are all zero, then
det(𝐴) = 0.
43. A matrix in which the entries in each row (or in each col-
umn) form a geometric progression starting with 1 is called g. The determinant of a lower triangular matrix is the sum
a Vandermonde matrix in honor of the French medical of the entries along the main diagonal.
doctor, mathematician, and musician Alexandre-Théophile
Vandermonde (February 28, 1735–January 1, 1796). Here are h. For every square matrix 𝐴 and every scalar c, it is true that
two examples. det(c𝐴) = c det(𝐴).

1 a a2 a3 i. For all square matrices 𝐴 and 𝐵, it is true that

1 1 1 ⎡ ⎤
⎢1 b b2 b3 ⎥ det(𝐴 + 𝐵) = det(𝐴) + det(𝐵)
𝑉 = [a b c ] and 𝑉=⎢ 3⎥
⎢1 c c2 c ⎥
a2 b2 c2 j. For every 2 × 2 matrix 𝐴 it is true that
⎣1 d d2 d3 ⎦
det(𝐴2 ) = (det(𝐴))2
Vandermonde matrices arise in a variety of applications, such
as polynomial interpolation (see Formula (14) and Example 6
of Section 1.10). Use cofactor expansion to prove that Working with Technology
T1. a. Use the determinant capability of your technology utility
|1 x1 x 21 |
| | to find the determinant of the matrix
|1 x2 x 22 | = (x 2 − x 1 )(x 3 − x 1 )(x 3 − x 2 )
| | 4.2 −1.3 1.1 6.0
|1 x3 x 23 | ⎡ ⎤
⎢0.0 0.0 −3.2 3.4⎥
𝐴=⎢ ⎥
⎢4.5 1.3 0.0 14.8⎥
True-False Exercises ⎢ ⎥
⎣4.7 1.0 3.4 2.3⎦
TF. In parts (a)–( j) determine whether the statement is true or b. Compare the result obtained in part (a) to that obtained by
false, and justify your answer. a cofactor expansion along the second row of 𝐴.
a b
a. The determinant of the 2 × 2 matrix [ ] is ad + bc. T2. Let 𝐴n be the n × n matrix with 2’s along the main diagonal,
c d
1’s along the diagonal lines immediately above and below the
b. Two square matrices that have the same determinant main diagonal, and zeros everywhere else. Make a conjecture
must have the same size. about the relationship between n and det(𝐴n ).

2.2 Evaluating Determinants by

Row Reduction
In this section we will show how to evaluate a determinant by reducing the associated
matrix to row echelon form. In general, this method requires less computation than cofac-
tor expansion and hence is the method of choice for large matrices.

A Basic Theorem
We begin with a fundamental theorem that will lead us to an efficient procedure for eval-
uating the determinant of a square matrix of any size.

Theorem 2.2.1

Let 𝐴 be a square matrix. If 𝐴 has a row of zeros or a column of zeros, then

det(𝐴) = 0.
 2.2 Evaluating Determinants by Row Reduction 127

Proof Since the determinant of 𝐴 can be found by a cofactor expansion along any row or
column, we can use the row or column of zeros. Thus, if we let 𝐶 1 , 𝐶 2 , . . . , 𝐶 n denote the
cofactors of 𝐴 along that row or column, then it follows from Formula (7) or (8) in Section
2.1 that
det(𝐴) = 0 ⋅ 𝐶 1 + 0 ⋅ 𝐶 2 + ⋅ ⋅ ⋅ + 0 ⋅ 𝐶 n = 0

The following useful theorem relates the determinant of a matrix and the determinant
of its transpose.

Theorem 2.2.2

Let 𝐴 be a square matrix. Then det(𝐴) = det(𝐴𝑇 ). Because transposing a

matrix changes its columns
to rows and its rows to
columns, almost every the-
Proof Since transposing a matrix changes its columns to rows and its rows to columns,
orem about the rows of a
the cofactor expansion of 𝐴 along any row is the same as the cofactor expansion of 𝐴𝑇 along
determinant has a compan-
the corresponding column. Thus, both have the same determinant.
ion version about columns,
and vice versa.
Elementary Row Operations
The next theorem shows how an elementary row operation on a square matrix affects the
value of its determinant. In place of a formal proof we have provided a table to illustrate
the ideas in the 3 × 3 case (see Table 1).

Theorem 2.2.3

Let 𝐴 be an n × n matrix.
(a) If 𝐵 is the matrix that results when a single row or single column of 𝐴 is multi-
plied by a scalar k, then det(𝐵) = k det(𝐴).
(b) If 𝐵 is the matrix that results when two rows or two columns of 𝐴 are inter-
changed, then det(𝐵) = − det(𝐴).
(c) If 𝐵 is the matrix that results when a multiple of one row of 𝐴 is added to another
or when a multiple of one column is added to another, then det(𝐵) = det(𝐴).

TA B LE 1

Relationship Operation The first panel of Table 1

shows that you can bring
| ka11 ka12 ka13 | | a11 a12 a13 | In the matrix 𝐵 the first
| | | | row of 𝐴 was multiplied a common factor from
| a21 a22 a23 | = k | a21 a22 a23 |
| | | | by k. any row (column) of a
| a31 a32 a33 | | a31 a32 a33 | determinant through the
det(𝐵) = k det(𝐴) determinant sign. This is
a slightly different way of
| a21 a22 a23 | | a11 a12 a13 | In the matrix 𝐵 the first and thinking about part (a) of
| | | | second rows of 𝐴 were
| a11 a12 a13 | = − | a21 a22 a23 | Theorem 2.2.3.
| | | | interchanged.
| a31 a32 a33 | | a31 a32 a33 |
det(𝐵) = − det(𝐴)

|a11 + ka21 a12 + ka22 a13 + ka23 | | a11 a12 a13 | In the matrix 𝐵 a multiple of
| | | | the second row of 𝐴 was
| a21 a22 a23 | = | a21 a22 a23 |
| | | | added to the first row.
| a31 a32 a33 | | a31 a32 a33 |
det(𝐵) = det(𝐴)
128 C H APT ER 2 Determinants

We will verify the first equation in Table 1 and leave the other two for you. To start,
note that the determinants on the two sides of the equation differ only in the first row,
so these determinants have the same cofactors, 𝐶 11 , 𝐶 12 , 𝐶 13 , along that row (since those
cofactors depend only on the entries in the second two rows). Thus, expanding the left
side by cofactors along the first row yields

| ka11 ka12 ka13 |

| |
| a21 a22 a23 | = ka11 𝐶 11 + ka12 𝐶 12 + ka13 𝐶 13
| |
| a31 a32 a33 |
= k(a11 𝐶 11 + a12 𝐶 12 + a13 𝐶 13 )
| a11 a12 a13 |
| |
= k | a21 a22 a23 |
| |
| a31 a32 a33 |

Elementary Matrices
It will be useful to consider the special case of Theorem 2.2.3 in which 𝐴 = 𝐼n is the n × n
identity matrix and 𝐸 (rather than 𝐵) denotes the elementary matrix that results when the
row operation is performed on 𝐼n . In this special case Theorem 2.2.3 implies the following
result.

Theorem 2.2.4

Let 𝐸 be an n × n elementary matrix.

(a) If 𝐸 results from multiplying a row of 𝐼n by a nonzero number k, then det(𝐸) = k.
(b) If 𝐸 results from interchanging two rows of 𝐼n , then det(𝐸) = −1.
(c) If 𝐸 results from adding a multiple of one row of 𝐼n to another, then det(𝐸) = 1.

EXAMPLE 1 | Determinants of Elementary Matrices

Observe that the deter- The following determinants of elementary matrices, which are evaluated by inspection, illus-
minant of an elementary trate Theorem 2.2.4.
matrix cannot be zero.
|1 0 0 0| |0 0 0 1| |1 0 0 7|
| | | | | |
|0 3 0 0| |0 1 0 0| |0 1 0 0|
| | = 3, | | = −1, | |=1
|0 0 1 0| |0 0 1 0| |0 0 1 0|
| | | | | |
|0 0 0 1| |1 0 0 0| |0 0 0 1|

The second row of I 4 The first and last rows of 7 times the last row of I 4
was multiplied by 3. I 4 were interchanged. was added to the first row.

Matrices with Proportional Rows or Columns

If a square matrix 𝐴 has two proportional rows, then a row of zeros can be introduced by
adding a suitable multiple of one of those rows to the other. Similarly for columns. But
 2.2 Evaluating Determinants by Row Reduction 129

adding a multiple of one row or column to another does not change the determinant, so
from Theorem 2.2.1, we must have det(𝐴) = 0. This proves the following theorem.

Theorem 2.2.5

If 𝐴 is a square matrix with two proportional rows or two proportional columns,

then det(𝐴) = 0.

EXAMPLE 2 | Proportional Rows or Columns

Each of the following matrices has two proportional rows or columns; thus, each has a deter-
minant of zero.
3 −1 4 −5
1 −2 7 ⎡ ⎤
−1 4 ⎢ 6 −2 5 2⎥
[ ], [−4 8 5], ⎢ ⎥
−2 8 ⎢ 5 8 1 4⎥
2 −4 3 ⎢ ⎥
⎣−9 3 −12 15⎦

Evaluating Determinants by Row Reduction

We will now give a method for evaluating determinants that involves substantially less
computation than cofactor expansion. The idea of the method is to reduce the given matrix
to upper triangular form by elementary row operations, then compute the determinant of
the upper triangular matrix (an easy computation), and then relate that determinant to
that of the original matrix. Here is an example.

EXAMPLE 3 | Using Row Reduction to Evaluate a Determinant

Evaluate det(𝐴) where Even with today’s fastest

0 1 5 computers it would take
millions of years to calculate
𝐴 = [3 −6 9]
a 25 × 25 determinant
2 6 1
by cofactor expansion, so
Solution We will reduce 𝐴 to row echelon form (which is upper triangular) and then apply methods based on row
Theorem 2.1.2. reduction are often used
for large determinants. For
|0 1 5| |3 −6 9|
| | | | determinants of small size
det(𝐴) = | 3 −6 9| = −|0 1 5| The first and second rows of
| | | | 𝐴 were interchanged. (such as those in this text),
|2 6 1| |2 6 1| cofactor expansion is often a
|1 −2 3| reasonable choice.
| |
= −3 | 0 1 5| A common factor of 3 from
| | the first row was taken
|2 6 1| through the determinant sign.
130 C H APT ER 2 Determinants

|1 −2 3|
| |
= −3 | 0 1 5| −2 times the first row was
| | added to the third row.
|0 10 −5 |

|1 −2 3|
| |
= −3 | 0 1 5| −10 times the second row
| | was added to the third row.
|0 0 −55 |

|1 −2 3|
| |
= (−3)(−55) | 0 1 5| A common factor of −55
| | from the last row was taken
|0 0 1| through the determinant sign.

= (−3)(−55)(1) = 165

EXAMPLE 4 | Using Column Operations to

Evaluate a Determinant

Compute the determinant of

1 0 0 3
⎡ ⎤
⎢2 7 0 6⎥
𝐴=⎢
⎢0 6 3 0⎥⎥
⎣7 3 1 −5⎦
Solution This determinant could be computed as above by using elementary row opera-
tions to reduce 𝐴 to row echelon form, but we can put 𝐴 in lower triangular form in one step
Example 4 points out that
by adding −3 times the first column to the fourth to obtain
it is always wise to keep
an eye open for column 1 0 0 0
⎡ ⎤
operations that can shorten ⎢2 7 0 0⎥
det(𝐴) = det ⎢ = (1)(7)(3)(−26) = −546
computations. ⎢0 6 3 0⎥⎥
⎣7 3 1 −26⎦

Cofactor expansion and row or column operations can sometimes be used in combi-
nation to provide an effective method for evaluating determinants. The following example
illustrates this idea.

EXAMPLE 5 | Row Operations and Cofactor Expansion

Evaluate det(𝐴) where

3 5 −2 6
⎡ ⎤
⎢1 2 −1 1⎥
𝐴=⎢
⎢2 4 1 5⎥⎥
⎣3 7 5 3⎦
 2.2 Evaluating Determinants by Row Reduction 131

Solution By adding suitable multiples of the second row to the remaining rows, we obtain
|0 −1 1 3|
| |
1 2 −1 1|
det(𝐴) = ||
|0 0 3 3 ||
|0 1 8 0|

| −1 1 3|
| |
= −| 0 3 3| Cofactor expansion along
| | the first column
| 1 8 0|

| −1 1 3|
| |
= −| 0 3 3| We added the first row to
| | the third row.
| 0 9 3|
|3 3|
= −(−1) | | Cofactor expansion along
|9 3| the first column
= −18

Exercise Set 2.2

In Exercises 1–4, verify that det(𝐴) = det(𝐴𝑇 ). 1 3 1 5 3

⎡ ⎤
⎢−2 −7 0 −4 2⎥
−2 3 −6 1 ⎢ ⎥
1. 𝐴 = [ ] 2. 𝐴 = [ ] 13. ⎢ 0 0 1 0 1⎥
1 4 2 −2
⎢ 0 0 2 1 1⎥
⎢ ⎥
2 −1 3 4 2 −1 ⎣ 0 0 0 1 1⎦
3. 𝐴 = [1 2 4] 4. 𝐴 = [ 0 2 −3] 1 −2 3 1
⎡ ⎤
5 −3 6 −1 1 5 ⎢ 5 −9 6 3⎥
14. ⎢
⎢−1 2 −6 −2⎥⎥
In Exercises 5–8, find the determinant of the given elementary matrix
by inspection. ⎣ 2 8 6 1⎦
In Exercises 15–22, evaluate the determinant, given that
1 0 0 0
⎡ ⎤ 1 0 0 |a b c|
⎢0 1 0 0⎥ | |
5. ⎢ 6. [ 0 1 0] |d e 𝑓 | = −6
⎢0 0 −5 0⎥⎥
|
|g h i|
|
−5 0 1
⎣0 0 0 1⎦
|d e 𝑓| |g h i|
| | | |
15. | g h i| 16. | d e 𝑓|
1 0 0 0 1 0 0 0 | | | |
⎡ ⎤ ⎡ ⎤ |a b c| |a b c|
⎢0 0 1 0⎥ ⎢0 − 31 0 0⎥
7. ⎢ 8. ⎢
⎢0 1 0 0⎥⎥ ⎢0 0 1 0⎥
⎥ | 3a
|
3b 3c |
|
|a + d
|
b+e c+𝑓|
|
⎣0 0 0 1⎦ ⎣0 0 0 1⎦ 17. | −d −e −𝑓 | 18. | −d −e −𝑓 |
| | | |
| 4g 4h 4i | | g h i |
In Exercises 9–14, evaluate the determinant of the matrix
by first reducing the matrix to row echelon form and then using |a + g b+h c + i| | a b c |
| | | |
some combination of row operations and cofactor expansion. 19. | d e 𝑓 | 20. | 2d 2e 2𝑓 |
| | | |
| g h i | | g + 3a h + 3b i + 3c |
3 −6 9 3 6 −9
| −3a −3b −3c | | a b c |
9. [−2 7 −2] 10. [ 0 0 −2] | | | |
21. | d e 𝑓 | 22. | d e 𝑓|
0 1 5 −2 1 5 | | | |
| g − 4d h − 4e i − 4𝑓 | | 2a 2b 2c |

2 1 3 1 23. Use row reduction to show that

⎡ ⎤ 1 −3 0
⎢1 0 1 1⎥ |1 1 1 |
11. ⎢ 12. [−2 4 1]
⎢0 2 1 0⎥⎥
|
|a b
|
c | = (b − a)(c − a)(c − b)
5 −2 2 | 2 |
⎣0 1 2 3⎦ |a b2 c2 |
132 C H APT ER 2 Determinants

24. Verify the formulas in parts (a) and (b) and then make a con- 1 2 0 0 0
jecture about a general result of which these results are special ⎡ ⎤
⎢0 1 2 0 0⎥
cases. ⎢
32. 𝑀 = ⎢0 0 1 0 0⎥⎥
0 0 a13 ⎢ ⎥
a. det [0 a22 a23 ] = −a13 a22 a31 ⎢0 0 0 1 2⎥
⎣2 0 0 0 1⎦
a31 a32 a33
0 0 0 a14 33. Let 𝐴 be an n × n matrix, and let 𝐵 be the matrix that results
⎡ ⎤ when the rows of 𝐴 are written in reverse order. State a theo-
⎢0 0 a23 a24 ⎥
b. det ⎢ = a14 a23 a32 a41 rem that describes how det(𝐴) and det(𝐵) are related.
⎢0 a32 a33 a34 ⎥
⎥
⎣a41 a42 a43 a44 ⎦
34. Find the determinant of the following matrix.

In Exercises 25–28, confirm the identities without evaluating any of a b b b

the determinants directly. ⎡ ⎤
⎢b a b b⎥
| a1 b1 a1 + b1 + c1 | | a1 b1 c1 | ⎢b b a b⎥
| | | | ⎢ ⎥
25. | a2 b2 a2 + b2 + c2 | = | a2 b2 c2 |
| | | | ⎣b b b a⎦
| a3 b3 a3 + b3 + c3 | | a3 b3 c3 |

| a1 + b1 t
|
a 2 + b2 t a 3 + b3 t |
|
| a1
|
a2 a3 |
|
True-False Exercises
26. | a1 t + b1 a2 t + b 2 a3 t + b3 | = (1 − t 2 ) | b1 b2 b3 |
| | | | TF. In parts (a)–(f ) determine whether the statement is true or
| c1 c2 c3 | | c1 c2 c3 | false, and justify your answer.
| a1 + b1 a1 − b 1 c1 | | a1 b1 c1 | a. If 𝐴 is a 4 × 4 matrix and 𝐵 is obtained from 𝐴 by inter-
| | | | changing the first two rows and then interchanging the
27. | a2 + b2 a2 − b 2 c2 | = −2 | a2 b2 c2 |
| | | | last two rows, then det(𝐵) = det(𝐴).
| a3 + b3 a3 − b 3 c3 | | a3 b3 c3 |

| a1 b1 + ta1 c1 + rb1 + sa1 | | a1 a2 a3 | b. If 𝐴 is a 3 × 3 matrix and 𝐵 is obtained from 𝐴 by mul-

| | | |
28. | a2 b2 + ta2 c2 + rb2 + sa2 | = | b1 b2 b3 | tiplying the first column by 4 and multiplying the third
| | | |
| a3 b3 + ta3 c3 + rb3 + sa3 | | c1 c2 c3 | column by 34 , then det(𝐵) = 3 det(𝐴).

In Exercises 29–30, show that det(𝐴) = 0 without directly evaluat- c. If 𝐴 is a 3 × 3 matrix and 𝐵 is obtained from 𝐴 by adding
ing the determinant. 5 times the first row to each of the second and third rows,
−2 8 1 4 then det(𝐵) = 25 det(𝐴).
⎡ ⎤
⎢ 3 2 5 1⎥
29. 𝐴 = ⎢
⎢ 1 10 6 5⎥⎥
d. If 𝐴 is an n × n matrix and 𝐵 is obtained from 𝐴 by mul-
tiplying each row of 𝐴 by its row number, then
⎣ 4 −6 4 −3⎦
−4 1 1 1 1 n(n + 1)
⎡ ⎤ det(𝐵) = det(𝐴)
⎢ 1 −4 1 1 1⎥ 2
30. 𝐴 = ⎢ 1 1 −4 1 1⎥
⎢ ⎥ e. If 𝐴 is a square matrix with two identical columns, then
⎢ 1 1 1 −4 1⎥
det(𝐴) = 0.
⎣ 1 1 1 1 −4⎦

It can be proved that if a square matrix 𝑀 is partitioned into block f. If the sum of the second and fourth row vectors of a 6 × 6
triangular form as matrix 𝐴 is equal to the last row vector, then det(𝐴) = 0.
𝐴 0 𝐴 𝐶
𝑀=[ ] or 𝑀 = [ ] Working with Technology
𝐶 𝐵 0 𝐵
T1. Find the determinant of
in which 𝐴 and 𝐵 are square, then det(𝑀) = det(𝐴) det(𝐵). Use
this result to compute the determinants of the matrices in Exer- 4.2 −1.3 1.1 6.0
cises 31 and 32. ⎡ ⎤
⎢0.0 0.0 −3.2 3.4⎥
1 2 0 8 6 −9 𝐴=⎢ ⎥
⎡ ⎤ ⎢4.5 1.3 0.0 14.8⎥
⎢ 2 5 0 4 7 5⎥ ⎢ ⎥
⎢−1 ⎣4.7 1.0 3.4 2.3⎦
⎢ 3 2 6 9 −2⎥ ⎥
31. 𝑀 = ⎢ ⎥
⎢ 0 0 0 3 0 0⎥ by reducing the matrix to reduced row echelon form, and
⎢ 0 0 0 2 1 0⎥ compare the result obtained in this way to that obtained in
⎢ ⎥
⎣ 0 0 0 −3 8 −4⎦ Exercise T1 of Section 2.1.
 2.3 Properties of Determinants; Cramer’s Rule 133

2.3 Properties of Determinants;

Cramer’s Rule
In this section we will develop some fundamental properties of matrices, and we will use
these results to derive a formula for the inverse of an invertible matrix and formulas for
the solutions of certain kinds of linear systems.

Basic Properties of Determinants

Suppose that 𝐴 and 𝐵 are n × n matrices and k is any scalar. We begin by considering
possible relationships among det(𝐴), det(𝐵), and
det(k𝐴), det(𝐴 + 𝐵), and det(𝐴𝐵)
Since a common factor of any row of a matrix can be moved through the determinant sign,
and since each of the n rows in k𝐴 has a common factor of k, it follows that

det(k𝐴) = kn det(𝐴) (1)

For example,
| ka11 ka12 ka13 | | a11 a12 a13 |
| | | |
| ka21 ka22 ka23 | = k3 | a21 a22 a23 |
| | | |
| ka31 ka32 ka33 | | a31 a32 a33 |
Unfortunately, no simple relationship exists among det(𝐴), det(𝐵), and det(𝐴 + 𝐵). In
particular, det(𝐴 + 𝐵) will usually not be equal to det(𝐴) + det(𝐵). The following example
illustrates this fact.

EXAMPLE 1 | det(A + B) ≠ det(A) + det(B)

Consider
1 2 3 1 4 3
𝐴=[ ], 𝐵=[ ], 𝐴+𝐵 =[ ]
2 5 1 3 3 8
We have det(𝐴) = 1, det(𝐵) = 8, and det(𝐴 + 𝐵) = 23; thus
det(𝐴 + 𝐵) ≠ det(𝐴) + det(𝐵)

In spite of the previous example, there is a useful relationship concerning sums of

determinants that is applicable when the matrices involved are the same except for one
row or column. For example, consider the following two matrices that differ only in the
second row:
a11 a12 a11 a12
𝐴=[ ] and 𝐵 = [ ]
a21 a22 b21 b22
Calculating the determinants of 𝐴 and 𝐵, we obtain
det(𝐴) + det(𝐵) = (a11 a22 − a12 a21 ) + (a11 b22 − a12 b21 )
= a11 (a22 + b22 ) − a12 (a21 + b21 )
a11 a12
= det [ ]
a21 + b21 a22 + b22
Thus
a11 a12 a11 a12 a11 a12
det [ ] + det [ ] = det [ ]
a21 a22 b21 b22 a21 + b21 a22 + b22
This is a special case of the following general result.
134 C H APT ER 2 Determinants

Theorem 2.3.1

Let 𝐴, 𝐵, and 𝐶 be n × n matrices that differ only in a single row, say the rth, and
assume that the rth row of 𝐶 can be obtained by adding corresponding entries in
the rth rows of 𝐴 and 𝐵. Then
det(𝐶) = det(𝐴) + det(𝐵)
The same result holds for columns.

EXAMPLE 2 | Sums of Determinants

We leave it to you to confirm the following equality by evaluating the determinants.

1 7 5 1 7 5 1 7 5
det [ 2 0 3 ] = det [2 0 3] + det [2 0 3]
1+0 4+1 7 + (−1) 1 4 7 0 1 −1

Determinant of a Matrix Product

Considering the complexity of the formulas for determinants and matrix multiplication,
it would seem unlikely that a simple relationship should exist between them. This is what
makes the simplicity of our next result so surprising. We will show that if 𝐴 and 𝐵 are
square matrices of the same size, then
det(𝐴𝐵) = det(𝐴) det(𝐵) (2)
The proof of this theorem is fairly intricate, so we will have to develop some preliminary
results first. We begin with the special case of (2) in which 𝐴 is an elementary matrix.
Because this special case is only a prelude to (2), we call it a lemma.

Lemma 2.3.2

If 𝐵 is an n × n matrix and 𝐸 is an n × n elementary matrix, then

det(𝐸𝐵) = det(𝐸) det(𝐵)

Proof We will consider three cases, each in accordance with the row operation that pro-
duces the matrix 𝐸.

Case 1 If 𝐸 results from multiplying a row of 𝐼n by k, then by Theorem 1.5.1, 𝐸𝐵 results

from 𝐵 by multiplying the corresponding row by k; so from Theorem 2.2.3(a) we have
det(𝐸𝐵) = k det(𝐵)
But from Theorem 2.2.4(a) we have det(𝐸) = k, so
det(𝐸𝐵) = det(𝐸) det(𝐵)

Cases 2 and 3 The proofs of the cases where 𝐸 results from interchanging two rows of
𝐼n or from adding a multiple of one row to another follow the same pattern as Case 1 and
are left as exercises.
 2.3 Properties of Determinants; Cramer’s Rule 135

Remark It follows by repeated applications of Lemma 2.3.2 that if 𝐵 is an n × n matrix

and 𝐸1 , 𝐸2 , . . . , 𝐸r are n × n elementary matrices, then
det(𝐸1 𝐸2 ⋅ ⋅ ⋅ 𝐸r 𝐵) = det(𝐸1 ) det(𝐸2 ) ⋅ ⋅ ⋅ det(𝐸r ) det(𝐵) (3)

Determinant Test for Invertibility

Our next theorem provides an important criterion for determining whether a matrix is
invertible. It also takes us a step closer to establishing Formula (2).

Theorem 2.3.3

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

Proof Let 𝑅 be the reduced row echelon form of 𝐴. As a preliminary step, we will show
that det(𝐴) and det(𝑅) are both zero or both nonzero: Let 𝐸1 , 𝐸2 , . . . , 𝐸r be the elementary
matrices that correspond to the elementary row operations that produce 𝑅 from 𝐴. Thus
𝑅 = 𝐸r ⋅ ⋅ ⋅ 𝐸2 𝐸1 𝐴
and from (3),
det(𝑅) = det(𝐸r ) ⋅ ⋅ ⋅ det(𝐸2 ) det(𝐸1 ) det(𝐴) (4)
We pointed out in the margin note that accompanies Theorem 2.2.4 that the determinant
of an elementary matrix is nonzero. Thus, it follows from Formula (4) that det(𝐴) and
det(𝑅) are either both zero or both nonzero, which sets the stage for the main part of the
proof. If we assume first that 𝐴 is invertible, then it follows from Theorem 1.6.4 that 𝑅 = 𝐼 It follows from Theorems
and hence that det(𝑅) = 1 (≠ 0). This, in turn, implies that det(𝐴) ≠ 0, which is what we 2.3.3 and 2.2.5 that a square
wanted to show. matrix with two propor-
Conversely, assume that det(𝐴) ≠ 0. It follows from this that det(𝑅) ≠ 0, which tells tional rows or two pro-
us that 𝑅 cannot have a row of zeros. Thus, it follows from Theorem 1.4.3 that 𝑅 = 𝐼 and portional columns is not
hence that 𝐴 is invertible by Theorem 1.6.4. invertible.

EXAMPLE 3 | Determinant Test for Invertibility

Since the first and third rows of

1 2 3
𝐴 = [1 0 1]
2 4 6
are proportional, det(𝐴) = 0. Thus 𝐴 is not invertible.

We are now ready for the main result concerning products of matrices.

Theorem 2.3.4

If 𝐴 and 𝐵 are square matrices of the same size, then

det(𝐴𝐵) = det(𝐴) det(𝐵)

Proof We divide the proof into two cases that depend on whether or not 𝐴 is invert-
ible. If the matrix 𝐴 is not invertible, then by Theorem 1.6.5 neither is the product 𝐴𝐵.
136 C H APT ER 2 Determinants

Thus, from Theorem 2.3.3, we have det(𝐴𝐵) = 0 and det(𝐴) = 0, so it follows that det(𝐴𝐵) =
det(𝐴) det(𝐵).
Now assume that 𝐴 is invertible. By Theorem 1.6.4, the matrix 𝐴 is expressible as a
product of elementary matrices, say
𝐴 = 𝐸1 𝐸2 ⋅ ⋅ ⋅ 𝐸r (5)
so
𝐴𝐵 = 𝐸1 𝐸2 ⋅ ⋅ ⋅ 𝐸r 𝐵
Applying (3) to this equation yields
det(𝐴𝐵) = det(𝐸1 ) det(𝐸2 ) ⋅ ⋅ ⋅ det(𝐸r ) det(𝐵)
and applying (3) again yields
det(𝐴𝐵) = det(𝐸1 𝐸2 ⋅ ⋅ ⋅ 𝐸r ) det(𝐵)
which, from (5), can be written as det(𝐴𝐵) = det(𝐴) det(𝐵).

EXAMPLE 4 | Verifying that det(AB) = det(A) det(B)

Consider the matrices

3 1 −1 3 2 17
𝐴=[ ], 𝐵=[ ], 𝐴𝐵 = [ ]
2 1 5 8 3 14
We leave it for you to verify that
det(𝐴) = 1, det(𝐵) = −23, and det(𝐴𝐵) = −23
Thus det(𝐴𝐵) = det(𝐴) det(𝐵), as guaranteed by Theorem 2.3.4.

The following theorem gives a useful relationship between the determinant of an

invertible matrix and the determinant of its inverse.

Theorem 2.3.5

If 𝐴 is invertible, then
1
det(𝐴−1 ) =
det(𝐴)

Historical Note
In 1815 the great French mathematician Augustin Cauchy pub-
lished a landmark paper in which he gave the first systematic
and modern treatment of determinants. It was in that paper
that Theorem 2.3.4 was stated and proved in full generality for
the first time. Special cases of the theorem had been stated and
proved earlier, but it was Cauchy who made the final jump.
[Image: © Bettmann/CORBIS]

Augustin Louis Cauchy

(1789–1857)
 2.3 Properties of Determinants; Cramer’s Rule 137

Proof Since 𝐴−1𝐴 = 𝐼, it follows that det(𝐴−1𝐴) = det(𝐼). Therefore, we must have
det(𝐴−1 ) det(𝐴) = 1. Since det(𝐴) ≠ 0, the proof can be completed by dividing through
by det(𝐴).

Adjoint of a Matrix
In a cofactor expansion we compute det(𝐴) by multiplying the entries in a row or column
by their cofactors and adding the resulting products. It turns out that if one multiplies
the entries in any row by the corresponding cofactors from a different row, the sum of
these products is always zero. (This result also holds for columns.) Although we omit the
general proof, the next example illustrates this fact.

EXAMPLE 5 | Entries and Cofactors from Different Rows

Let
3 2 −1
⎡ ⎤
𝐴 = ⎢1 6 3⎥
⎢ ⎥
⎣2 −4 0⎦
We leave it for you to verify that the cofactors of 𝐴 are
𝐶11 = 12 𝐶12 = 6 𝐶13 = −16
𝐶21 = 4 𝐶22 = 2 𝐶23 = 16
𝐶31 = 12 𝐶32 = −10 𝐶33 = 16
so, for example, the cofactor expansion of det(𝐴) along the first row is
det(𝐴) = 3𝐶11 + 2𝐶12 + (−1)𝐶13 = 36 + 12 + 16 = 64
and along the first column is
det(𝐴) = 3𝐶11 + 𝐶21 + 2𝐶31 = 36 + 4 + 24 = 64
Suppose, however, we multiply the entries in the first row by the corresponding cofactors
from the second row and add the resulting products. The result is
3𝐶21 + 2𝐶22 + (−1)𝐶23 = 12 + 4 − 16 = 0
Or suppose we multiply the entries in the first column by the corresponding cofactors from
the second column and add the resulting products. The result is again zero since
3𝐶12 + 1𝐶22 + 2𝐶32 = 18 + 2 − 20 = 0

Definition 1

If 𝐴 is any n × n matrix and 𝐶 ij is the cofactor of aij , then the matrix

𝐶 𝐶 12 ⋅⋅⋅ 𝐶 1n
⎡ 11 ⎤
⎢𝐶 21 𝐶 22 ⋅⋅⋅ 𝐶 2n ⎥
⎢ .. .. .. ⎥
⎢ . . . ⎥
⎣𝐶 n1 𝐶 n2 ⋅⋅⋅ 𝐶 nn ⎦
is called the matrix of cofactors from A. The transpose of this matrix is called the
adjoint of A and is denoted by adj(𝐴).
138 C H APT ER 2 Determinants

Historical Note
The use of the term adjoint for the transpose of the matrix of
cofactors appears to have been introduced by the American
mathematician L. E. Dickson in a research paper that he pub-
lished in 1902.
[Image: Courtesy of the American Mathematical Society
(www.ams.org)]

Leonard Eugene Dickson

(1874–1954)

EXAMPLE 6 | Adjoint of a 3 × 3 Matrix

Let
3 2 −1
𝐴 = [1 6 3]
2 −4 0
As noted in Example 5, the cofactors of 𝐴 are
𝐶11 = 12 𝐶12 = 6 𝐶13 = −16
𝐶21 = 4 𝐶22 = 2 𝐶23 = 16
𝐶31 = 12 𝐶32 = −10 𝐶33 = 16
so the matrix of cofactors is
12 6 −16
[ 4 2 16]
12 −10 16
and the adjoint of 𝐴 is
12 4 12
adj(𝐴) = [ 6 2 −10]
−16 16 16

In Theorem 1.4.5 we gave a formula for the inverse of a 2 × 2 invertible matrix. Our
next theorem extends that result to n × n invertible matrices.

Theorem 2.3.6

Inverse of a Matrix Using Its Adjoint

If 𝐴 is an invertible matrix, then
1
𝐴−1 = adj(𝐴) (6)
det(𝐴)
 2.3 Properties of Determinants; Cramer’s Rule 139

Proof We show first that

𝐴 adj(𝐴) = det(𝐴)𝐼
Consider the product
a a12 ... a1n
⎡ 11 ⎤
⎢a21 a22 ... a2n ⎥ 𝐶 11 𝐶 21 ... 𝐶 j1 ... 𝐶 n1
⎢ ... .. .. ⎥ ⎡ ⎤
. . ⎥ ⎢𝐶 12 𝐶 22 ... 𝐶 j2 ... 𝐶 n2 ⎥
𝐴 adj(𝐴) = ⎢
⎢ ai1 ai2 ... ain ⎥ ⎢ ... ..
.
..
.
.. ⎥
. ⎥
⎢ .. .. .. ⎥ ⎢
⎢ . . . ⎥ ⎣𝐶 1n 𝐶 2n ... 𝐶 jn ... 𝐶 nn ⎦
⎣an1 an2 ... ann ⎦
The entry in the ith row and jth column of the product 𝐴 adj(𝐴) is
ai1 𝐶 j1 + ai2 𝐶 j2 + ⋅ ⋅ ⋅ + ain 𝐶 jn (7)

(see the shaded lines above).

If i = j, then (7) is the cofactor expansion of det(𝐴) along the ith row of 𝐴 (Theo-
rem 2.1.1), and if i ≠ j, then the a’s and the cofactors come from different rows of 𝐴, so
the value of (7) is zero (as illustrated in Example 5). Therefore,

⎡det(𝐴) 0 ⋅⋅⋅ 0⎤
⎢ 0 det(𝐴) ⋅ ⋅ ⋅ 0⎥
𝐴 adj(𝐴) = ⎢ . .. ⎥ = det(𝐴)𝐼
.. (8)
⎢ .. . ⎥
.
⎢ ⎥
⎣ 0 0 ⋅⋅⋅ det(𝐴)⎦

Since 𝐴 is invertible, det(𝐴) ≠ 0. Therefore, Equation (8) can be rewritten as

1 1
[𝐴 adj(𝐴)] = 𝐼 or 𝐴[ adj(𝐴)] = 𝐼
det(𝐴) det(𝐴)

Multiplying both sides on the left by 𝐴−1 yields

1
𝐴−1 = adj(𝐴)
det(𝐴)

EXAMPLE 7 | Using the Adjoint to Find an Inverse Matrix

Use Formula (6) to find the inverse of the matrix 𝐴 in Example 6.

Solution We showed in Example 5 that det(𝐴) = 64. Thus,
12 4 12
12 4 12 ⎡ 64 64 64 ⎤
1 1 ⎢ ⎥
6 2
𝐴−1 = adj(𝐴) = [ 6 2 −10] = ⎢ 64 64 − 10
64
⎥
det(𝐴) 64 ⎢ ⎥
−16 16 16 ⎢− 16 16 16 ⎥
⎣ 64 64 64
⎦

Cramer’s Rule
Our next theorem uses the formula for the inverse of an invertible matrix to produce a
formula, called Cramer’s rule, for the solution of a linear system 𝐴x = b of n equations
in n unknowns in the case where the coefficient matrix 𝐴 is invertible (or, equivalently,
when det(𝐴) ≠ 0).
140 C H APT ER 2 Determinants

Theorem 2.3.7

Cramer’s Rule
If 𝐴x = b is a system of n linear equations in n unknowns such that det(𝐴) ≠ 0,
then the system has a unique solution. This solution is
det(𝐴1 ) det(𝐴2 ) det(𝐴n )
x1 = , x2 = ,..., xn =
det(𝐴) det(𝐴) det(𝐴)
where 𝐴j is the matrix obtained by replacing the entries in the jth column of 𝐴 by
the entries in the matrix
b
⎡ 1⎤
⎢b2 ⎥
b=⎢.⎥
.
⎢.⎥
⎣bn ⎦

Proof If det(𝐴) ≠ 0, then 𝐴 is invertible, and by Theorem 1.6.2, x = 𝐴−1 b is the unique
solution of 𝐴x = b. Therefore, by Theorem 2.3.6 we have
𝐶 𝐶 21 ⋅⋅⋅ 𝐶 n1 b1
⎡ 11 ⎤⎡ ⎤
1 1 ⎢𝐶 12 𝐶 22 ⋅⋅⋅ 𝐶 n2 ⎥ ⎢b2 ⎥
x = 𝐴−1 b = adj(𝐴)b = .. .. .. ⎥ ⎢ .. ⎥
det(𝐴) det(𝐴) ⎢
⎢ . . . ⎥⎢ . ⎥
⎣𝐶 1n 𝐶 2n ⋅⋅⋅ 𝐶 nn ⎦ ⎣bn ⎦
Multiplying the matrices out gives
b 𝐶 + b2 𝐶 21 + ⋅ ⋅ ⋅ + bn 𝐶 n1
⎡ 1 11 ⎤
1 ⎢b1 𝐶 12 + b2 𝐶 22 + ⋅ ⋅ ⋅ + bn 𝐶 n2 ⎥
x= .. .. ..
det(𝐴) ⎢
⎢ . . .
⎥
⎥
⎣b1 𝐶 1n + b2 𝐶 2n + ⋅ ⋅ ⋅ + bn 𝐶 nn ⎦
The entry in the jth row of x is therefore
b1 𝐶 1j + b2 𝐶 2j + ⋅ ⋅ ⋅ + bn 𝐶 nj
xj = (9)
det(𝐴)
Now let
a a12 ⋅⋅⋅ a1j−1 b1 a1j+1 ⋅⋅⋅ a1n
⎡ 11 ⎤
⎢a21 a22 ⋅⋅⋅ a2j−1 b2 a2j+1 ⋅⋅⋅ a2n ⎥
𝐴j = ⎢ . .. .. .. .. .. ⎥
.
⎢ . . . . . . ⎥
⎣an1 an2 ⋅⋅⋅ anj−1 bn anj+1 ⋅⋅⋅ ann ⎦

Historical Note
Variations of Cramer’s rule were fairly well known before the
Swiss mathematician discussed it in work he published in 1750.
It was Cramer’s superior notation that popularized the method
and led mathematicians to attach his name to it.
[Image: Science Source/Photo Researchers]

Gabriel Cramer
(1704–1752)
 2.3 Properties of Determinants; Cramer’s Rule 141

Since 𝐴 j differs from 𝐴 only in the jth column, it follows that the cofactors of entries
b1 , b2 , . . . , bn in 𝐴j are the same as the cofactors of the corresponding entries in the jth
column of 𝐴. The cofactor expansion of det(𝐴 j ) along the jth column is therefore
det(𝐴j ) = b1 𝐶 1j + b2 𝐶 2j + ⋅ ⋅ ⋅ + bn 𝐶 nj
Substituting this result in (9) gives
det(𝐴 j )
xj =
det(𝐴)

EXAMPLE 8 | Using Cramer’s Rule to Solve a Linear System

Use Cramer’s rule to solve

x1 + + 2x 3 = 6
−3x 1 + 4x 2 + 6x 3 = 30
−x 1 − 2x 2 + 3x 3 = 8
Solution
1 0 2 6 0 2
𝐴 = [−3 4 6], 𝐴1 = [30 4 6],
−1 −2 3 8 −2 3
For n > 3, it is usually more
1 6 2 1 0 6 efficient to solve a linear
𝐴2 = [−3 30 6], 𝐴3 = [−3 4 30] system with n equations
−1 8 3 −1 −2 8 in n unknowns by Gauss–
Jordan elimination than by
Therefore,
Cramer’s rule. Its main use
det(𝐴1 ) −40 −10 det(𝐴2 ) 72 18 is for obtaining properties of
x1 = = = , x2 = = = ,
det(𝐴) 44 11 det(𝐴) 44 11 solutions of a linear system
det(𝐴3 ) 152 38 without actually solving the
x3 = = =
det(𝐴) 44 11 system.

Equivalence Theorem
In Theorem 1.6.4 we listed five results that are equivalent to the invertibility of a matrix 𝐴.
We conclude this section by merging Theorem 2.3.3 with that list to produce the following
theorem that relates all of the major topics we have studied thus far.

Theorem 2.3.8

Equivalent Statements
If 𝐴 is an n × n matrix, then the following statements are equivalent.
(a) 𝐴 is invertible.
(b) 𝐴x = 0 has only the trivial solution.
(c) The reduced row echelon form of 𝐴 is 𝐼n .
(d) 𝐴 can be expressed as a product of elementary matrices.
(e) 𝐴x = b is consistent for every n × 1 matrix b.
(𝑓) 𝐴x = b has exactly one solution for every n × 1 matrix b.
(g) det(𝐴) ≠ 0.
142 C H APT ER 2 Determinants

OPTIONAL: We now have all of the machinery necessary to prove the following two
results, which we stated without proof in Theorem 1.7.1:

• Theorem 1.7.1(c) A triangular matrix is invertible if and only if its diagonal entries
are all nonzero.
• Theorem 1.7.1(d) The inverse of an invertible lower triangular matrix is lower tri-
angular, and the inverse of an invertible upper triangular matrix is upper triangular.

Proof of Theorem 1.7.1(c) Let 𝐴 = [aij ] be a triangular matrix, so that its diagonal entries
are
a11 , a22 , . . . , ann
From Theorem 2.1.2, the matrix 𝐴 is invertible if and only if
det(𝐴) = a11 a22 ⋅ ⋅ ⋅ ann
is nonzero, which is true if and only if the diagonal entries are all nonzero.

Proof of Theorem 1.7.1(d) We will prove the result for upper triangular matrices and
leave the lower triangular case for you. Assume that 𝐴 is upper triangular and invertible.
Since
1
𝐴−1 = adj(𝐴)
det(𝐴)
we can prove that 𝐴−1 is upper triangular by showing that adj(𝐴) is upper triangular or,
equivalently, that the matrix of cofactors is lower triangular. We can do this by showing
that every cofactor 𝐶 ij with i < j (i.e., above the main diagonal) is zero. Since
𝐶 ij = (−1)i+j 𝑀ij
it suffices to show that each minor 𝑀ij with i < j is zero. For this purpose, let 𝐵ij be the
matrix that results when the ith row and jth column of 𝐴 are deleted, so
𝑀ij = det(𝐵ij ) (10)
From the assumption that i < j, it follows that 𝐵ij is upper triangular (see Figure 1.7.1).
Since 𝐴 is upper triangular, its (i + 1)-st row begins with at least i zeros. But the ith row of
𝐵ij is the (i + 1)-st row of 𝐴 with the entry in the jth column removed. Since i < j, none of
the first i zeros is removed by deleting the jth column; thus the ith row of 𝐵ij starts with at
least i zeros, which implies that this row has a zero on the main diagonal. It now follows
from Theorem 2.1.2 that det(𝐵ij ) = 0 and from (10) that 𝑀ij = 0.

Exercise Set 2.3

In Exercises 1–4, verify that det(k𝐴) = kn det(𝐴). 2 1 0 1 −1 3

−1 2 2 2 5. 𝐴 = [3 4 0] and 𝐵 = [7 1 2]
1. 𝐴 = [ ]; k = 2 2. 𝐴 = [ ]; k = −4
3 4 5 −2 0 0 2 5 0 1

2 −1 3 −1 8 2 2 −1 −4
3. 𝐴 = [3 2 1]; k = −2 6. 𝐴 = [ 1 0 −1] and 𝐵 = [1 1 3]
1 4 5 −2 2 2 0 3 −1

1 1 1 In Exercises 7–14, use determinants to decide whether the given

matrix is invertible.
4. 𝐴 = [0 2 3]; k = 3
0 1 −2 2 5 5 2 0 3
In Exercises 5–6, verify that det(𝐴𝐵) = det(𝐵𝐴) and determine 7. 𝐴 = [−1 −1 0] 8. 𝐴 = [ 0 3 2]
whether the equality det(𝐴 + 𝐵) = det(𝐴) + det(𝐵) holds. 2 4 3 −2 0 −4
 2.3 Properties of Determinants; Cramer’s Rule 143

2 −3 5 −3 0 1 31. Use Cramer’s rule to solve for the unknown y without solving
9. 𝐴 = [0 1 −3] 10. 𝐴 = [ 5 0 6] for the unknowns x, z, and 𝑤.
0 0 2 8 0 3 4x + y + z + 𝑤 = 6
4 2 8 1 0 −1 3x + 7y − z + 𝑤 = 1
11. 𝐴 = [ −2 1 −4] 12. 𝐴 = [9 −1 4] 7x + 3y − 5z + 8𝑤 = −3
3 1 6 8 9 −1 x + y + z + 2𝑤 = 3

2 0 0 32. Let 𝐴x = b be the system in Exercise 31.

⎡ √2 − √7 0⎤
13. 𝐴 = [ 8 1 0] 14. 𝐴 = ⎢
⎢3√2 −3√7
⎥
0⎥
a. Solve by Cramer’s rule.
−5 3 6 ⎢ ⎥
⎣ 5 −9 0⎦ b. Solve by Gauss–Jordan elimination.
c. Which method involves fewer computations?
In Exercises 15–18, find the values of k for which the matrix 𝐴 is
invertible. 33. Let
a b c
k−3 −2 k 2 𝐴 = [d e 𝑓]
15. 𝐴 = [ ] 16. 𝐴 = [ ]
−2 k−2 2 k g h i
1 2 4 1 2 0 Assuming that det(𝐴) = −7, find
17. 𝐴 = [3 1 6] 18. 𝐴 = [k 1 k] a. det(3𝐴) b. det(𝐴−1 ) c. det(2𝐴−1 )
k 3 2 0 2 1
a g d
In Exercises 19–23, decide whether the matrix is invertible, and if so, d. det((2𝐴)−1 ) e. det [b h e]
use the adjoint method to find its inverse. c i 𝑓
2 5 5 2 0 3 34. In each part, find the determinant given that 𝐴 is a 4 × 4
19. 𝐴 = [−1 −1 0] 20. 𝐴 = [ 0 3 2] matrix for which det(𝐴) = −2.
2 4 3 −2 0 −4
a. det(−𝐴) b. det(𝐴−1 ) c. det(2𝐴𝑇 ) d. det(𝐴3 )
2 −3 5 2 0 0
21. 𝐴 = [0 1 −3] 22. 𝐴 = [ 8 1 0] 35. In each part, find the determinant given that 𝐴 is a 3 × 3
0 0 2 −5 3 6 matrix for which det(𝐴) = 7.

1 3 1 1 a. det(3𝐴) b. det(𝐴−1 )
⎡ ⎤
⎢2 5 2 2⎥ c. det(2𝐴−1 ) d. det((2𝐴)−1 )
23. 𝐴 = ⎢
1 3 8 9⎥
⎢ ⎥
⎣1 3 2 2⎦
Working with Proofs
In Exercises 24–29, solve by Cramer’s rule, where it applies. 36. Prove that a square matrix 𝐴 is invertible if and only if 𝐴𝑇 𝐴 is
24. 7x 1 − 2x 2 = 3 25. 4x + 5y =2 invertible.
3x 1 + x 2 = 5 11x + y + 2z = 3
x + 5y + 2z = 1 37. Prove that if 𝐴 is a square matrix, then
det(𝐴𝑇 𝐴) = det(𝐴𝐴𝑇 )
26. x − 4y + z = 6 27. x 1 − 3x 2 + x 3 = 4
4x − y + 2z = −1 2x 1 − x 2 = −2 38. Let 𝐴x = b be a system of n linear equations in n unknowns
2x + 2y − 3z = −20 4x 1 − 3x 3 = 0 with integer coefficients and integer constants. Prove that if
det(𝐴) = 1, the solution x has integer entries.
28. −x 1 − 4x 2 + 2x 3 + x4 = −32
2x 1 − x2 + 7x 3 + 9x 4 = 14 39. Prove that if det(𝐴) = 1 and all the entries in 𝐴 are integers,
−x 1 + x2 + 3x 3 + x4 = 11 then all the entries in 𝐴−1 are integers.
x1 − 2x 2 + x3 − 4x 4 = −4

29. 3x 1 − x 2 + x 3 = 4
True-False Exercises
−x 1 + 7x 2 − 2x 3 = 1 TF. In parts (a)–(l) determine whether the statement is true or
2x 1 + 6x 2 − x 3 = 5 false, and justify your answer.

30. Show that the matrix a. If 𝐴 is a 3 × 3 matrix, then det(2𝐴) = 2 det(𝐴).

cos 𝜃 sin 𝜃 0 b. If 𝐴 and 𝐵 are square matrices of the same size such that
𝐴 = [ − sin 𝜃 cos 𝜃 0] det(𝐴) = det(𝐵), then det(𝐴 + 𝐵) = 2 det(𝐴).
0 0 1 c. If 𝐴 and 𝐵 are square matrices of the same size and 𝐴 is
−1
invertible, then
is invertible for all values of 𝜃; then find 𝐴 using Theo-
rem 2.3.6. det(𝐴−1 𝐵𝐴) = det(𝐵)
144 C H APT ER 2 Determinants

d. A square matrix 𝐴 is invertible if and only if det(𝐴) = 0. in which 𝜖 > 0. Since det(𝐴) = 𝜖 ≠ 0, it follows from The-
𝑇 orem 2.3.8 that 𝐴 is invertible. Compute det(𝐴) for various
e. The matrix of cofactors of 𝐴 is precisely [adj(𝐴)] .
small nonzero values of 𝜖 until you find a value that produces
f. For every n × n matrix 𝐴, we have det(𝐴) = 0, thereby leading you to conclude erroneously that
𝐴 ⋅ adj(𝐴) = (det(𝐴))𝐼n 𝐴 is not invertible. Discuss the cause of this.

g. If 𝐴 is a square matrix and the linear system 𝐴x = 0 has T2. We know from Exercise 39 that if 𝐴 is a square matrix then
multiple solutions for x, then det(𝐴) = 0. det(𝐴𝑇 𝐴) = det(𝐴𝐴𝑇 ). By experimenting, make a conjec-
ture as to whether this is true if 𝐴 is not square.
h. If 𝐴 is an n × n matrix and there exists an n × 1 matrix b
such that the linear system 𝐴x = b has no solutions, then T3. The French mathematician Jacques Hadamard (1865–1963)
the reduced row echelon form of 𝐴 cannot be 𝐼n . proved that if 𝐴 is an n × n matrix each of whose entries sat-
isfies the condition |aij | ≤ 𝑀, then
i. If 𝐸 is an elementary matrix, then 𝐸x = 0 has only the
trivial solution.
|det(𝐴)| ≤ √nn 𝑀 n
j. If 𝐴 is an invertible matrix, then the linear system 𝐴x = 0
has only the trivial solution if and only if the linear system (Hadamard’s inequality). For the following matrix 𝐴, use
𝐴−1 x = 0 has only the trivial solution. this result to find an interval of possible values for det(𝐴),
and then use your technology utility to show that the value of
k. If 𝐴 is invertible, then adj(𝐴) must also be invertible. det(𝐴) falls within this interval.
l. If 𝐴 has a row of zeros, then so does adj(𝐴).
0.3 −2.4 −1.7 2.5
⎡ ⎤
Working with Technology ⎢0.2 −0.3 −1.2 1.4⎥
𝐴=⎢ ⎥
⎢2.5 2.3 0.0 1.8⎥
T1. Consider the matrix ⎢ ⎥
1 1 ⎣1.7 1.0 −2.1 2.3⎦
𝐴=[ ]
1 1+𝜖

Chapter 2 Supplementary Exercises

In Exercises 1–8, evaluate the determinant of the given matrix by (a) 12. Use the determinant to decide whether the matrices in Exer-
cofactor expansion and (b) using elementary row operations to intro- cises 5–8 are invertible.
duce zeros into the matrix.
In Exercises 13–15, find the given determinant by any method.
−4 2 7 −1
1. [ ] 2. [ ] |3 −4 a|
3 3 −2 −6 | 5 b − 3| | |
13. | | 14. | a2 1 2|
|b − 2 −3 | | |
−1 5 2 −1 −2 −3 |2 a−1 4|
3. [ 0 2 −1] 4. [ −4 −5 −6]
−3 1 1 −7 −8 −9 |0 0 0 0 −3 |
| |
|0 0 0 −4 0|
3 0 −1 −5 1 4
15. || 0 0 −1 0 0 ||
5. [1 1 1] 6. [ 3 0 2] |0 2 0 0 0|
0 4 2 1 −2 2 | |
|5 0 0 0 0|
3 6 0 1 −1 −2 −3 −4
⎡ ⎤ ⎡ ⎤
⎢−2 3 1 4⎥ ⎢ 4 3 2 1⎥ 16. Solve for x.
7. ⎢ 8. ⎢ ⎥
⎢ 1 0 −1 1⎥⎥ ⎢ 1 2 3 4⎥ |x −1 | |
|1 0 −3 |
|
⎣−4 −3 −2 −1⎦ | | = |2 x −6 |
⎣−9 2 −2 2⎦ |3 1 − x| | |
|1 3 x − 5|
9. Evaluate the determinants in Exercises 3–6 by using the arrow
technique (see Example 7 in Section 2.1).
In Exercises 17–24, use the adjoint method (Theorem 2.3.6) to find
10. a. Construct a 4 × 4 matrix whose determinant is easy to com- the inverse of the given matrix, if it exists.
pute using cofactor expansion but hard to evaluate using
elementary row operations. 17. The matrix in Exercise 1. 18. The matrix in Exercise 2.

b. Construct a 4 × 4 matrix whose determinant is easy to com-

19. The matrix in Exercise 3. 20. The matrix in Exercise 4.
pute using elementary row operations but hard to evaluate
using cofactor expansion.
21. The matrix in Exercise 5. 22. The matrix in Exercise 6.
11. Use the determinant to decide whether the matrices in Exer-
cises 1–4 are invertible. 23. The matrix in Exercise 7. 24. The matrix in Exercise 8.
 Chapter 2 Supplementary Exercises 145

25. Use Cramer’s rule to solve for x′ and y′ in terms of x and y. 34. a. In the accompanying figure, the area of the triangle 𝐴𝐵𝐶
x= 3 ′ 4 ′ can be expressed as
5x − 5y
4 ′ 3 ′ area 𝐴𝐵𝐶 = area 𝐴𝐷𝐸𝐶 + area 𝐶𝐸𝐹𝐵 − area 𝐴𝐷𝐹𝐵
y= 5x + 5y

26. Use Cramer’s rule to solve for x′ and y′ in terms of x and y. Use this and the fact that the area of a trapezoid equals 12
x = x′ cos 𝜃 − y′ sin 𝜃 the altitude times the sum of the parallel sides to show that
y = x′ sin 𝜃 + y′ cos 𝜃 | x1 y1 1|
1| |
27. By examining the determinant of the coefficient matrix, show area 𝐴𝐵𝐶 = | x2 y2 1|
2| |
that the following system has a nontrivial solution if and only | x3 y3 1|
if 𝛼 = 𝛽.
x + y + 𝛼z = 0 [Note: In the derivation of this formula, the vertices are
labeled such that the triangle is traced counterclockwise
x + y + 𝛽z = 0
proceeding from (x 1 , y1 ) to (x 2 , y2 ) to (x 3 , y3 ). For a clock-
𝛼x + 𝛽y + z = 0 wise orientation, the determinant above yields the negative
28. Let 𝐴 be a 3 × 3 matrix, each of whose entries is 1 or 0. What of the area.]
is the largest possible value for det(𝐴)? b. Use the result in (a) to find the area of the triangle with ver-
tices (3, 3), (4, 0), (−2, −1).
29. a. For the triangle in the accompanying figure, use trigonom-
etry to show that
b cos 𝛾 + c cos 𝛽 = a
C(x3, y3)
c cos 𝛼 + a cos 𝛾 = b
a cos 𝛽 + b cos 𝛼 = c B(x2, y2)
and then apply Cramer’s rule to show that A(x1, y1)
b2 + c2 − a2
cos 𝛼 =
2bc
b. Use Cramer’s rule to obtain similar formulas for cos 𝛽 and D E F
cos 𝛾.
FIGURE Ex-34

γ
b a
α β
c
35. Use the fact that

FIGURE Ex-29
21375, 38798, 34162, 40223, 79154

are all divisible by 19 to show that

|2 1 3 7 5|
30. Use determinants to show that for all real values of 𝜆, the only | |
solution of |3 8 7 9 8|
x − 2y = 𝜆x |3 4 1 6 2 ||
|
x − y = 𝜆y |4 0 2 2 3|
| |
is x = 0, y = 0. |7 9 1 5 4|
31. Prove: If 𝐴 is invertible, then adj(𝐴) is invertible and is divisible by 19 without directly evaluating the determinant.
1
[adj(𝐴)]−1 = 𝐴 = adj(𝐴−1 ) 36. Without directly evaluating the determinant, show that
det(𝐴)
32. Prove: If 𝐴 is an n × n matrix, then | sin 𝛼 cos 𝛼 sin(𝛼 + 𝛿) |
| |
det[adj(𝐴)] = [det(𝐴)] n−1 | sin 𝛽 cos 𝛽 sin(𝛽 + 𝛿) | = 0
| |
| sin 𝛾 cos 𝛾 sin(𝛾 + 𝛿) |
33. Prove: If the entries in each row of an n × n matrix 𝐴 add up
to zero, then the determinant of 𝐴 is zero. [Hint: Consider the 𝑇 a b
37. Let 𝑇 ∶ 𝑅2 → 𝑅 be the mapping (a, b, c, d) ⟶ det[ ]. Is
product 𝐴x, where x is the n × 1 matrix, each of whose entries c d
is one.] this a linear transformation? Justify your answer.