2.

2 Evaluating Determinants by Row Reduction 113

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
cofactor expansion and hence is the method of choice for large matrices.

A BasicTheorem We begin with a fundamental theorem that will lead us to an efficient procedure for
evaluating 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.

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 C1 , C2 , . . . , Cn 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 · C1 + 0 · C2 + · · · + 0 · Cn = 0

The following useful theorem relates the determinant of a matrix and the determinant
of its transpose.
Because transposing a matrix
changes its columns to rows
and its rows to columns, al- THEOREM 2.2.2 Let A be a square matrix. Then det(A) = det(AT ).
most every theorem about the
rows of a determinant has Proof Since transposing a matrix changes its columns to rows and its rows to columns,
a companion version about
the cofactor expansion of A along any row is the same as the cofactor expansion of AT
columns, and vice versa.
along the corresponding column. Thus, both have the same determinant.

Elementary Row The next theorem shows how an elementary row operation on a square matrix affects the
Operations 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).

Table 1

The first panel of Table 1 Relationship Operation

shows that you can bring a    
ka11 ka12 ka13  a11 a12 a13  In the matrix B the first
common factor from any row  
    row of A was multiplied
(column) of a determinant  a21 a22 a23  = k a21 a22 a23 
    by k .
through the determinant sign.  a31 a32 a33  a31 a32 a33 
This is a slightly different way
of thinking about part (a) of det(B) = k det(A)
Theorem 2.2.3.    
a21 a22 a23  a11 a12 a13  In the matrix B the first and
 
    second rows of A were
a11 a12 a13  = − a21 a22 a23 
    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)
114 Chapter 2 Determinants

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 multiplied
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 interchanged,
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).

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, C11 , C12 , C13 , 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 C11 + ka12 C12 + ka13 C13

a a32 a33 
31
= k(a11 C11 + a12 C12 + a13 C13 )
 
a11 a12 a13 
 
= k a21 a22 a23 
a a32 a33 
31

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.

E X A M P L E 1 Determinants of Elementary Matrices

The following determinants of elementary matrices, which are evaluated by inspection,
Observe that the determinant
illustrate Theorem 2.2.4.
of an elementary matrix can-      
not be zero. 1 0 0 0  0 0 0 1  1 0 0 7
   
0 0  0 0  0 1 0 0 
 3 0  1 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 I4 The first and last rows of 7 times the last row of I4
was multiplied by 3. I4 were interchanged. was added to the first row.

Matrices with Proportional If a square matrix A has two proportional rows, then a row of zeros can be introduced
Rows or Columns by adding a suitable multiple of one of the rows to the other. Similarly for columns. But
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.
 2.2 Evaluating Determinants by Row Reduction 115

THEOREM 2.2.5 If A is a square matrix with two proportional rows or two proportional
columns, then det(A) = 0.

E X A M P L E 2 Proportional Rows or Columns

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

Evaluating Determinants We will now give a method for evaluating determinants that involves substantially less
by Row Reduction 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 de-
terminant of the upper triangular matrix (an easy computation), and then relate that
determinant to that of the original matrix. Here is an example.

E X A M P L E 3 Using Row Reduction to Evaluate a Determinant

Evaluate det(A) where ⎡ ⎤
0 1 5
⎢ ⎥
A = ⎣3 −6 9⎦
2 6 1
Solution We will reduce A to row echelon form (which is upper triangular) and then
apply Theorem 2.1.2.
   
0 5  3 −6 9 
 1 
   
det(A) =  3 −6 9 = − 0 1 5 The first and second rows of
    A were interchanged.
2 6 1 2 6 1
 
Even with today’s fastest com-  1 −2 3 
 
puters it would take millions of  
= −3  0 1 5 A common factor of 3 from
years to calculate a 25 × 25 de-   the first row was taken
terminant by cofactor expan- 2 6 1 through the determinant sign.

sion, so methods based on row  

 1 −2 3 
reduction are often used for 
 
large determinants. For deter- = −3  0 1 5 −2 times the first row was
  added to the third row.
minants of small size (such as  0 10 −5 
those in this text), cofactor ex-  
pansion is often a reasonable  1 −2 3 

choice.  
= −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
116 Chapter 2 Determinants

E X A M P L E 4 Using Column Operations to Evaluate a Determinant

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

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

E X A M P L E 5 Row Operations and Cofactor Expansion

Evaluate det(A) where ⎡ ⎤
3 5 −2 6
⎢1 −1 1⎥
⎢ 2 ⎥
A=⎢ ⎥
⎣2 4 1 5⎦
3 7 5 3
Solution By adding suitable multiples of the second row to the remaining rows, we
obtain  
 0 −1 1 3 

1 − 1 
 2 1
det(A) =  
0 0 3 3
 
0 1 8 0
 
 −1 3 
 1
 
= − 0 3 3 Cofactor expansion along
  the first column
 1 8 0
 
 −1 3 
 1
 
= − 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
 2.2 Evaluating Determinants by Row Reduction 117

Exercise Set 2.2

In Exercises 1–4, verify that det(A) = det(AT ).    
d e f  g h i 
 
−2 3 −6 1    
1. A = 2. A = 15. g h i 16. d e f
−2    
1 4 2 a b c a b c
⎡ ⎤ ⎡ ⎤    
2 −1 3 4 2 −1  3a 3b 3c a + d b+e c + f 
⎢ ⎥ ⎢ ⎥  
3. A = ⎣1 2 4⎦ 4. A = ⎣ 0 2 −3 ⎦    
17. −d −e −f  18.  −d −e −f 
5 −3 6 −1 1 5    
 4g 4h 4i   g h i 
In Exercises 5–8, find the determinant of the given elementary    
matrix by inspection. a + g b + h c + i   a b c 
  
⎡ ⎤    
19.  d e f  20.  2d 2e 2f 
1 0 0 0 ⎡ ⎤    
⎢0 ⎥ 1 0 0  g h i  g + 3a h + 3b i + 3c
⎢ 1 0 0⎥ ⎢ ⎥
5. ⎢ ⎥ 6. ⎣ 0 1 0⎦    
⎣0 0 −5 0⎦  −3a −3b −3c  a b c 
−5 0 1  
0 0 0 1    
21.  d e f  22.  d e f
   
g − 4d h − 4e i − 4f  2a 2b 2c
⎡ ⎤ ⎡ ⎤
1 0 0 0 1 0 0 0
⎢0 0⎥ ⎢0 − 13 0⎥
23. Use row reduction to show that
⎢ 0 1 ⎥ ⎢ 0 ⎥  
7. ⎢ ⎥ 8. ⎢ ⎥ 1 1 
⎣0 1 0 0⎦ ⎣0 0 1 0⎦  1
 
0 0 0 1 0 0 0 1 a b c  = (b − a)(c − a)(c − b)
 2 
a b2 c2 
In Exercises 9–14, evaluate the determinant of the matrix
by first reducing the matrix to row echelon form and then using 24. Verify the formulas in parts (a) and (b) and then make a con-
some combination of row operations and cofactor expansion. jecture about a general result of which these results are special
cases.
⎡ ⎤
⎡ ⎤ ⎡ ⎤ 0 0 a13
3 −6 9 3 6 −9 ⎢ ⎥
⎢ ⎥ ⎢ ⎥ (a) det ⎣0 a22 a23 ⎦ = −a13 a22 a31
9. ⎣−2 7 −2 ⎦ 10. ⎣ 0 0 −2 ⎦
a31 a32 a33
0 1 5 −2 1 5
⎡ ⎤
0 0 0 a14
⎡ ⎤ ⎢0
2 1 3 1 ⎡ ⎤ ⎢ 0 a23 a24 ⎥ ⎥
⎢1 −3 (b) det ⎢ ⎥ = a14 a23 a32 a41
1⎥
1 0
⎢ 0 1 ⎥ ⎢ ⎥ ⎣0 a32 a33 a34 ⎦
11. ⎢ ⎥ 12. ⎣−2 4 1⎦
⎣0 2 1 0⎦ a41 a42 a43 a44
5 −2 2
0 1 2 3
In Exercises 25–28, confirm the identities without evaluating
⎡ ⎤
1 3 1 5 3 the determinants directly.
⎢ ⎥    
⎢−2 −7 −4 2⎥ a1 b1 a1 + b1 + c1  a1 b1 c1 
⎢
0
⎥    
   
13. ⎢
⎢ 0 0 1 0 1⎥⎥ 25. a2 b2 a2 + b2 + c2  = a2 b2 c2 
⎢ 0    
⎣ 0 2 1 1⎥⎦ a3 b3 a3 + b3 + c3  a3 b3 c3 
0 0 0 1 1    
a1 + b1 t a2 + b2 t a3 + b3 t  a1 a2 a3 
  
⎡ ⎤   2  
−2 26. a1 t + b1 a2 t + b2 a3 t + b3  = (1 − t ) b1 b2 b3 
1 3 1    
⎢ 5  c1 c3  c1 c2 c3 
⎢ −9 6 3⎥⎥
c2
14. ⎢ ⎥    
⎣−1 2 −6 −2 ⎦ a1 + b1 a1 − b1 c1  a1 b1 c1 
   
2 8 6 1    
27.  2
a + b 2 a2 − b 2 c2 = − 2 a2 b2 c2 
   
In Exercises 15–22, evaluate the determinant, given that a3 + b3 a3 − b3 c3  a3 b3 c3 
     
a a1 b1 + ta1 c1 + rb1 + sa1  a1 a2 a3 
 b c     
d    
 e f  = −6 28. a2 b2 + ta2 c2 + rb2 + sa2  = b1 b2 b3 
g    
h i a3 b3 + ta3 c3 + rb3 + sa3  c1 c2 c3