Lineng PDF
Lineng PDF
Lineng PDF
Marcel B. Finan
All
c Rights Reserved
Last Updated
Marcel B. Finan
Russellville, Arkansas
January 2015
i
ii PREFACE
Contents
Preface i
Matrices 61
7 Matrices and Matrix Operations . . . . . . . . . . . . . . . . . . . 62
8 Matrix Multiplication . . . . . . . . . . . . . . . . . . . . . . . . 74
9 The Inverse of a Square Matrix . . . . . . . . . . . . . . . . . . . 84
10 Elementary Matrices . . . . . . . . . . . . . . . . . . . . . . . . 92
11 Finding A−1 Using Elementary Matrices . . . . . . . . . . . . . 101
Determinants 109
12 Determinants by Cofactor Expansion . . . . . . . . . . . . . . . 110
13 Evaluating Determinants by Row Reduction . . . . . . . . . . . 118
14 Additional Properties of the Determinant . . . . . . . . . . . . . 127
15 Finding A−1 Using Cofactors . . . . . . . . . . . . . . . . . . . . 133
16 Application of Determinants to Systems: Cramer’s Rule . . . . . 140
1
2 CONTENTS
Index 305
Linear Systems of Equations
In this chapter we shall develop the theory of general systems of linear equa-
tions. The tool we will use to find the solutions is the row-echelon form of
a matrix. In fact, the solutions can be read off from the row-echelon form
of the augmented matrix of the system. The solution technique, known as
elimination method, is developed in Section 4.
3
4 LINEAR SYSTEMS OF EQUATIONS
or equivalently
3x1 + 4x2 + 2x3 = 13.35
5x1 + 2x2 + 3x3 = 19.50
−x1 + 4x2 + x3 = 0.30.
Thus, the problem is to find the values of x1 , x2 , and x3 . A system like the
one above is called a linear system.
Many practical problems can be reduced to solving systems of linear equa-
tions. The main purpose of linear algebra is to find systematic methods for
solving these systems. So it is natural to start our discussion of linear algebra
by studying linear equations.
a1 x1 + a2 x2 + ... + an xn = b (1.1)
Example 1.1
Determine whether the given equations are linear or not (i.e., non-linear):
(a) 3x1 − 4x2 + 5x3 = 6.
(b) 4x1 − 5x2 = x1 x2 .
√
(c) x2 = 2 x1 − 6.
(d) x1 + sin x2 + x3 = 1.
(e) x1 − x2 + x3 = sin 3.
Solution
(a) The given equation is in the form given by (1.1) and therefore is linear.
(b) The equation is non-linear because the term on the right side of the equa-
tion involves a product of the variables x1 and x2 .
√
(c) A non-linear equation because the term 2 x1 involves a square root of
the variable x1 .
(d) Since x2 is an argument of a trigonometric function, the given equation
is non-linear.
(e) The equation is linear according to (1.1)
In the case of n = 2, sometimes we will drop the subscripts and use in-
stead x1 = x and x2 = y. For example, ax + by = c. Geometrically, this is a
straight line in the xy−coordinate system. Likewise, for n = 3, we will use
x1 = x, x2 = y, and x3 = z and write ax + by + cz = d which is a plane in
the xyz−coordinate system.
A solution of a linear equation (1.1) in n unknowns is a finite ordered
collection of numbers s1 , s2 , ..., sn which make (1.1) a true equality when
x1 = s1 , x2 = s2 , · · · , xn = sn are substituted in (1.1). The collection of
all solutions of a linear equation is called the solution set or the general
solution.
Example 1.2
Show that (5 + 4s − 7t, s, t), where s, t ∈ R, is a solution to the equation
x1 − 4x2 + 7x3 = 5.
Solution
x1 = 5 + 4s − 7t, x2 = s, and x3 = t is a solution to the given equation
because
x1 − 4x2 + 7x3 = (5 + 4s − 7t) − 4s + 7t = 5
6 LINEAR SYSTEMS OF EQUATIONS
Many problems in the sciences lead to solving more than one linear equation.
The general situation can be described by a linear system.
A system of linear equations or simply a linear system is any finite
collection of linear equations. A linear system of m equations in n variables
has the form
a11 x1 + a12 x2 + · · · + a1n xn = b1
a21 x1 + a22 x2 + · · · + a2n xn = b2
.. ..
.=.
am1 x1 + am2 x2 + · · · + amn xn = bm .
Note that the coefficients aij consist of two subscripts. The subscript i in-
dicates the equation in which the coefficient occurs, and the subscript j
indicates which unknown it multiplies.
When a linear system has more equations than unknowns, we call the sys-
tem overdetermined. When the system has more unknowns than equations
then we call the system underdetermined.
A solution of a linear system in n unknowns is a finite ordered collection of
numbers s1 , s2 , ..., sn for which the substitution
x1 = s1 , x2 = s2 , · · · , xn = sn
(s1 , s2 , · · · , sn ).
The collection of all solutions of a linear system is called the solution set
or the general solution. To solve a linear system is to find its general
solution.
A linear system can have infinitely many solutions (dependent system),
exactly one solution (independent system) or no solutions at all. When a
linear system has a solution we say that the system is consistent. Otherwise,
the system is said to be inconsistent. Thus, for the case n = 2, a linear
system is consistent if the two lines either intersect at one point (independent)
or they coincide (dependent). In the case the two lines are parallel, the system
is inconsistent. For the case, n = 3, replace a line by a plane.
1 SYSTEMS OF LINEAR EQUATIONS 7
Example 1.3
Find the general solution of the linear system
(
x+ y =7
2x + 4y = 18.
Solution.
Multiply the first equation of the system by −2 and then add the resulting
equation to the second equation to find 2y = 4. Solving for y we find y = 2.
Plugging this value in one of the equations of the given system and then
solving for x one finds x = 5
Example 1.4
Solve the system (
7x + 2y = 16
−21x − 6y = 24.
Solution.
Graphing the two lines we find
Example 1.5
Solve the system (
9x + y = 36
3x + 31 y = 12.
8 LINEAR SYSTEMS OF EQUATIONS
Solution.
Graphing the two lines we find
Thus, the system is consistent and dependent. Note that the two equa-
tions are basically the same: 9x + y = 36. Letting y = t, where t is called a
parameter, we can solve for x and find x = 36−t
9
. Thus, the general solution
is defined by the parametric equations
36 − t
x= , y=t
9
Example 1.6
By letting x3 = t, find the general solution of the linear system
(
x1 + x2 + x3 = 7
2x1 + 4x2 + x3 = 18.
Solution.
By letting x3 = t the given system can be rewritten in the form
(
x1 + x2 = 7 − t
2x1 + 4x2 = 18 − t.
Practice Problems
Problem 1.1
Define the following terms:
(a) Solution to a system of m linear equations in n unknowns.
(b) Consistent system of linear equations.
(c) Inconsistent system of linear equations.
(d) Dependent system of linear equations.
(e) Independent system of linear equations.
Problem 1.2
Which of the following equations are not linear and why?
(a) x21 + 3x2 − 2x3 = 5.
(b) x1 + x1 x2 + 2x3 = 1.
(c) x1 + x22 + x3 = 5.
Problem 1.3
Show that (2s + 12t + 13, s, −s − 3t − 3, t) is a solution to the system
(
2x1 + 5x2 + 9x3 + 3x4 = −1
x1 + 2x2 + 4x3 = 1.
Problem 1.4
Solve each of the following systems graphically:
(a) (
4x1 − 3x2 = 0
2x1 + 3x2 = 18.
(b) (
4x1 − 6x2 = 10
6x1 − 9x2 = 15.
(c) (
2x1 + x2 = 3
2x1 + x2 = 1.
Which of the above systems is consistent and which is inconsistent?
10 LINEAR SYSTEMS OF EQUATIONS
Problem 1.5
Determine whether the system of equations is linear or non-linear.
(a)
ln x1 + x2 + x3 = 3
2x1 + x2 − 5x3 = 1
−x1 + 5x2 + 3x3 = −1.
(b)
3x1 + 4x2 + 2x3 = 13.35
5x1 + 2x2 + 3x3 = 19.50
−x1 + 4x2 + x3 = 0.30.
Problem 1.6
Find the parametric equations of the solution set to the equation −x1 +5x2 +
3x3 − 2x4 = −1.
Problem 1.7
Write a system of linear equations consisting of three equations in three
unknowns with
(a) no solutions;
(b) exactly one solution;
(c) infinitely many solutions.
Problem 1.8
For what values of h and k the system below has (a) no solution, (b) a unique
solution, and (c) many solutions.
(
x1 + 3x2 = 2
3x1 + hx2 = k.
Problem 1.9
True/False:
(a) A general solution of a linear system is an explicit description of all the
solutions of the system.
(b) A linear system with either one solution or infinitely many solutions is
said to be inconsistent.
(c) Finding a parametric description of the solution set of a linear system is
the same as solving the system.
(d) A linear system with a unique solution is consistent and dependent.
1 SYSTEMS OF LINEAR EQUATIONS 11
Problem 1.10
Find a linear equation in the variables x and y that has the general solution
x = 5 + 2t and y = t.
Problem 1.11
Find a relationship between a, b, c so that the following system is consistent.
x1 + x2 + 2x3 = a
x1 + x3 = b
2x1 + x2 + 3x3 = c.
Problem 1.12
Determine which points are solutions to the system
(
x1 − x2 = 6
4x1 + 3x2 = −4.
Problem 1.13
The graph of a system of linear equations is given below.
Problem 1.14
You invest a total of $5,800 in two investments earning 3.5% and 5.5% simple
interest. Your goal is to have a total annual interest income of $283. Write a
system of linear equations that represents this situation where x1 represents
the amount invested in the 3.5% fund and x2 represents the amount invested
in the 5.5% fund.
Problem 1.15
Consider the system (
mx1 + x2 = m2
x1 + mx2 = 1.
(a) For which value(s) of m does the system have no solutions?
(b) What about infinitely many solutions?
Problem 1.16
A little boy opened his piggy bank which has coins consisting of pennies and
nickels. The total amount in the jar is $8.80. If there are twice as many
nickels as pennies, how many pennies does the boy have? How many nickels?
Problem 1.17
Two angles are complementary. One angle is 81◦ less than twice the other.
Find the two angles.
Problem 1.18
The sum of three numbers is 14. The largest is 4 times the smallest, while
the sum of the smallest and twice the largest is 18. Find the numbers.
Problem 1.19
Show that the system
x1 − 2x2 = −1
4x1 + 3x2 = 7
3x1 + 5x2 = 8.
Problem 1.20
A linear equation in three variables can be visualized as a plane in the three
1 SYSTEMS OF LINEAR EQUATIONS 13
Example 2.1
Show that the system (
x1 − 3x2 = −7
2x1 + x2 = 7
is equivalent to the system
8x1 − 3x2 = 7
3x1 − 2x2 = 0
10x1 − 2x2 = 14.
Solution.
Solving the first system one finds the solution x1 = 2, x2 = 3. Similarly,
solving the second system one finds the solution x1 = 2 and x2 = 3. Hence,
the two systems are equivalent
Example 2.2
Show that if x1 + kx2 = c and x1 + `x2 = d are equivalent then k = l and
c = d.
Solution.
The general solution to the first equation is the ordered pair (c − kt, t) where
t is an arbitrary number. For this solution to be a solution to the second
equation as well, we must have c − kt + `t = d for all t ∈ R. In particular,
if t = 0 we find c = d. Thus, kt = `t for all t ∈ R. Letting t = 1 we find k = `
Our basic algebraic method for solving a linear system is known as the
method of elimination. The method consists of reducing the original
system to an equivalent system that is easier to solve. The reduced sys-
tem has the shape of an upper (resp. lower) triangle. This new system can
2 EQUIVALENT SYSTEMS AND ELEMENTARY ROW OPERATIONS: THE ELIMINATION METH
To indicate which operation is being used in the process one can use the
following shorthand notation. For example, r3 ← 21 r3 represents the row
operation of type (I) where each entry of row 3 is being replaced by 12 that
entry. Similar interpretations for types (II) and (III) operations.
The following theorem asserts that the system obtained from the original
system by means of elementary row operations has the same set of solutions
as the original one.
Theorem 2.1
Suppose that an elementary row operation is performed on a linear system.
Then the resulting system is equivalent to the original system.
Example 2.3
Use the elimination method described above to solve the system
x1 + x2 − x3 = 3
x1 − 3x2 + 2x3 = 1
2x1 − 2x2 + x3 = 4.
Solution.
Step 1: We eliminate x1 from the second and third equations by performing
two operations r2 ← r2 − r1 and r3 ← r3 − 2r1 obtaining
x1 + x2 − x3 = 3
− 4x2 + 3x3 = −2
− 4x2 + 3x3 = −2.
Solution.
Step 1: To eliminate the variable x1 from the second and third equations
we perform the operations r2 ← 3r2 − 2r1 and r3 ← 3r3 − 4r1 obtaining the
system
3x1 + 4x2 + x3 = 1
x2 − 2x3 = −2
− 7x2 − 7x3 = −10.
Step 2: Now, to eliminate the variable x3 from the third equation we apply
the operation r3 ← r3 + 7r2 to obtain
3x1 + 4x2 + x3 = 1
x2 − 2x3 = −2
− 21x3 = −24.
Practice Problems
Problem 2.1
Solve each of the following systems using the method of elimination:
(a) (
4x1 − 3x2 = 0
2x1 + 3x2 = 18.
(b) (
4x1 − 6x2 = 10
6x1 − 9x2 = 15.
(c) (
2x1 + x2 = 3
2x1 + x2 = 1.
Which of the above systems is consistent and which is inconsistent?
Problem 2.2
Find the values of A, B, C in the following partial fraction decomposition
x2 − x + 3 Ax + B C
2
= 2 + .
(x + 2)(2x − 1) x +2 2x − 1
Problem 2.3
Find a quadratic equation of the form y = ax2 + bx + c that goes through
the points (−2, 20), (1, 5), and (3, 25).
Problem 2.4
Solve the following system using the method of elimination.
5x1 − 5x2 − 15x3 = 40
4x1 − 2x2 − 6x3 = 19
3x1 − 6x2 − 17x3 = 41.
Problem 2.5
Solve the following system using elimination.
2x1 + x2 + x3 = −1
x1 + 2x2 + x3 = 0
3x1 − 2x3 = 5.
18 LINEAR SYSTEMS OF EQUATIONS
Problem 2.6
Find the general solution of the linear system
(
x1 − 2x2 + 3x3 + x4 = −3
2x1 − x2 + 3x3 − x4 = 0.
Problem 2.7
Find a, b, and c so that the system
x1 + ax2 + cx3 = 0
bx1 + cx2 − 3x3 = 1
ax + 2x + bx = 5
1 2 3
and
21x1 + 6x2 + 6x3 = 63
− 4x2 + 6x3 = 2
x3 = 3.
Problem 2.9
Solve the following system by elimination.
3x1 + x2 + 2x3 = 13
2x1 + 3x2 + 4x3 = 19
x1 + 4x2 + 3x3 = 15.
Problem 2.10
Solve the following system by elimination.
x1 − 2x2 + 3x3 = 7
2x1 + x2 + x3 = 4
−3x1 + 2x2 − 2x3 = −10.
2 EQUIVALENT SYSTEMS AND ELEMENTARY ROW OPERATIONS: THE ELIMINATION METH
Problem 2.11
John, Philip, and Grace play a round of golf together. Their combined score
is 224. John’s score was 8 more than Philip’s, and Grace’s score was 7 more
than John’s. What was each person’s score?
Problem 2.12
Set up the system of equations to solve the following problem. Do not solve
it.
Billy’s Restaurant ordered 200 flowers for Mother’s Day. They ordered car-
nations at $1.50 each, roses at $5.75 each, and daisies at $2.60 each. They
ordered mostly carnations, and 20 fewer roses than daisies. The total order
came to $589.50. How many of each type of flower was ordered?
Problem 2.13
In each case, tell whether the operation is a valid row operation. If it is, say
what it does (in words).
(a) r1 ↔ r3 .
(b) r2 ← r2 + 5.
(c) r3 ← r3 + 4r2 .
(d) r4 ← 5r4 .
Problem 2.14
Find the general solution to the system
x1 + x2 + x3 + 2x4 = 1
2x2 + 2x3 + x4 = 0
2x1 + 2x3 + x4 = 1.
Problem 2.15
Determine the value(s) of a so that the following system is inconsistent.
x1 + 2x2 +
x3 = a
x1 + x2 + ax3 = 1
3x + 4x + (a2 − 2)x = 1.
1 2 3
Problem 2.16
True or false:
(a) Elementary row operations on an augmented matrix never change the
solution set of the associated linear system.
(b) Two matrices are row equivalent if they have the same number of rows.
(c) Two linear systems are equivalent if they have the same solution set.
20 LINEAR SYSTEMS OF EQUATIONS
Problem 2.17
Determine the value(s) of h so that the system
(
x1 + hx2 = −3,
−2x1 + 4x2 = 6.
Problem 2.20
(a) The elementary row operation ri ↔ rj swaps rows i and j. What elemen-
tary row operation will undo this operation?
(b) The elementary row operation ri ← αri multiplies row i by a non-zero
constant α. What elementary row operation will undo this operation?
(c) The elementary row operation ri ↔ ri + αrj adds α row j to row i. What
elementary row operation will undo this operation?
Problem 2.21
Show that (s1 , s2 , · · · , sn ) is a solution to a1 x1 + a2 x2 + · · · + an xn = b if and
only if (s1 , s2 , · · · , sn ) is a solution to ka1 x1 + ka2 x2 + · · · + an xn = b for any
non-zero scalar k.
Problem 2.22
Show that (s1 , s2 , · · · , sn ) is a solution to a1 x1 + a2 x2 + · · · + an xn = a and
b1 x1 + b2 x2 + · · · + bn xn = b if and only if (s1 , s2 , · · · , sn ) is a solution to
(ka1 + b1 )x1 + (ka2 + b2 )x2 + · · · + (kan + bn )xn = ka + b and a1 x1 + a2 x2 +
· · · + an xn = a.
2 EQUIVALENT SYSTEMS AND ELEMENTARY ROW OPERATIONS: THE ELIMINATION METH
Problem 2.23
Using Problems 2.21 and 2.22, prove Theorem 2.1.
22 LINEAR SYSTEMS OF EQUATIONS
Solution.
The augmented matrix for the system is
1 −2 1 0
0 2 −8 8
−4 5 9 −9
3 SOLVING LINEAR SYSTEMS USING AUGMENTED MATRICES 23
Example 3.2
Solve the following linear system using the method described above.
x2 + 5x3 = −4
x1 + 4x2 + 3x3 = −2
2x1 + 7x2 + x3 = −1.
Solution.
The augmented matrix for the system is
0 1 5 −4
1 4 3 −2
2 7 1 −1
Example 3.3
Determine if the following system is consistent.
x2 − 4x3 = 8
2x1 − 3x2 + 2x3 = 1
5x1 − 8x2 + 7x3 = 1.
Solution.
The augmented matrix of the given system is
0 1 −4 8
2 −3 2 1
5 −8 7 1
This last system has no solution ( the last equation requires x1 , x2 , and x3
to satisfy the equation 0x1 + 0x2 + 0x3 = 5 and no such x1 , x2 , and x3 exist).
Hence the original system is inconsistent
Pay close attention to the last row of the triangular matrix of the previ-
ous exercise. This situation is typical of an inconsistent system.
26 LINEAR SYSTEMS OF EQUATIONS
Practice Problems
Problem 3.1
Solve the following linear system using the elimination method of this section.
x1 + 2x2
=0
−x1 + 3x2 + 3x3 = −2
x2 + x3 = 0.
Problem 3.2
Find an equation involving g, h, and k that makes the following augmented
matrix corresponds to a consistent system.
2 5 −3 g
4 7 −4 h
−6 −3 1 k
Problem 3.3
Solve the following system using elementary row operations on the augmented
matrix:
5x1 − 5x2 − 15x3 = 40
4x1 − 2x2 − 6x3 = 19
3x1 − 6x2 − 17x3 = 41.
Problem 3.4
Solve the following system using elementary row operations on the augmented
matrix:
2x1 + x2 + x3 = −1
x1 + 2x2 + x3 = 0
3x1 − 2x3 = 5.
Problem 3.5
Solve the following system using elementary row operations on the augmented
matrix:
x1 − x2 + 2x3 + x4 = 0
2x1 + 2x2 − x4 = 0
3x1 + x2 + 2x3 + x4 = 0.
3 SOLVING LINEAR SYSTEMS USING AUGMENTED MATRICES 27
Problem 3.6
Find the value(s) of a for which the following system has a nontrivial solution.
Find the general solution.
x1 + 2x2 + x3 = 0
x1 + 3x2 + 6x3 = 0
2x + 3x + ax = 0.
1 2 3
Problem 3.7
Solve the linear system whose augmented matrix is given by
1 1 2 8
−1 −2 3 1
3 −7 4 10
Problem 3.8
Solve the linear system whose augmented matrix is reduced to the following
triangular form
1 −3 7 1
0 1 4 0
0 0 0 1
Problem 3.9
Solve the linear system whose augmented matrix is reduced to the following
triangular form
1 0 0 −7 8
0 1 0 3 2
0 0 1 1 −5
Problem 3.10
Reduce the matrix to triangular matrix.
−1 −1 0 0
0 0 2 3
4 0 −2 1
3 −1 0 4
28 LINEAR SYSTEMS OF EQUATIONS
Problem 3.11
Solve the following system using elementary row operations on the augmented
matrix:
3x1 + x2 + 7x3 + 2x4 = 13
2x − 4x + 14x − x = −10
1 2 3 4
5x1 + 11x2 − 7x3 + 8x4 = 59
2x1 + 5x2 − 4x3 − 3x4 = 39.
Problem 3.12
Determine the value(s) of a so that the line whose parametric equations are
given by
x1 = −3 + t
x2 = 2−t
x3 = 1 + at
Problem 3.13
Find the general solution to the system
x1 − 2x2 + 2x3 − x4 = 3
3x1 + x2 + 6x3 + 11x4 = 16
2x1 − x2 + 4x3 + x4 = 9.
Problem 3.14
Solve the system
x1 − 2x2 − 6x3 = 12
2x1 + 4x2 + 12x3 = −17
x1 − 4x2 − 12x3 = 22.
Problem 3.15
Solve the system
x1 − 2x2 − 3x3 = 0
−x1 + x2 + 2x3 = 3
2x2 + x3 = −8.
3 SOLVING LINEAR SYSTEMS USING AUGMENTED MATRICES 29
Problem 3.16
True or false. Justify your answer.
(a) A general solution of a system is an explicit description of all solutions
of the system.
(b) If one row in an reduced form of an augmented matrix is [0 0 0 5 0],
then the associated linear system is inconsistent.
(c) If one row in an reduced form of an augmented matrix is [0 0 0 0 0],
then the associated linear system is inconsistent.
(d) The row reduction algorithm applies only to augmented matrices for a
linear system.
Problem 3.17
The augmented matrix
1 2 0 −3
0 1 −3 0
0 0 k −4
represents a system of equations in the variables x1 , x2 , and x3 .
(a) For what values of k is there no solution?
(b) For what value of k is there exactly one solution? What is that solution?
Problem 3.18
The cubic function p(x) = a0 +a1 t+a2 t2 +a3 t3 crosses the points (20, 106), (30, 123), (40, 132)
and (50, 151). Find the value of p(60).
Problem 3.19
Solve the system
2x3 + 2x4 − 6x5 = −2
x1 + 2x2 + x3 + 4x4 − 12x5 = −3
x1 + 2x2 + x3 + 2x4 − 6x5 = −1.
Problem 3.20
Solve the system
x1 + x2 + x3 = 11
x1 − x2 + 3x3 = 5
2x1 + 2x2 + 2x3 = 15.
30 LINEAR SYSTEMS OF EQUATIONS
Solution.
(a)The given matrix is in row-echelon form but not in reduced row-echelon
form since the (1, 2)−entry is not zero.
(b) The given matrix satisfies the characterization of a reduced row-echelon
form
Theorem 4.1
Every nonzero matrix can be brought to (reduced) row-echelon form by a
finite number of elementary row operations.
Proof.
The proof consists of the following steps:
Step 1. Find the first column from the left containing a non-zero entry (call
it a), and move the row containing that entry to the top position.
Step 2. Multiply the row from Step 1 by a1 to create a leading 1.
Step 3. By subtracting multiples of that row from the rows below it, make
each entry below the leading 1 zero.
Step 4. This completes the first row. Now repeat steps 1-3 on the matrix
consisting of the remaining rows.
The process stops when either no rows remain in step 4 or the remaining
rows consist of zeros. The entire matrix is now in row-echelon form.
To find the reduced row-echelon form we need the following additional step.
Step 5. Beginning with the last nonzero row and working upward, add suit-
able multiples of each row to the rows above to introduce zeros above the
leading 1
Example 4.2
Use Gauss-Jordan elimination to transform the following matrix first into
32 LINEAR SYSTEMS OF EQUATIONS
Step 1: r1 ↔ r4
1 4 5 −9 −7
−1 −2 −1 3 1
−2 −3 0 3 −1
0 −3 −6 4 9
Step 2: r2 ← r2 + r1 and r3 ← r3 + 2r1
1 4 5 −9 −7
0
2 4 −6 −6
0 5 10 −15 −15
0 −3 −6 4 9
Step 3: r2 ← 12 r2 and r3 ← 15 r3
1 4 5 −9 −7
0
1 2 −3 −3
0 1 2 −3 −3
0 −3 −6 4 9
Step 4: r3 ← r3 − r2 and r4 ← r4 + 3r2
1 4 5 −9 −7
0 1 2 −3 −3
0 0 0 0 0
0 0 0 −5 0
Step 5: r3 ↔ r4
1 4 5 −9 −7
0
1 2 −3 −3
0 0 0 −5 0
0 0 0 0 0
4 ECHELON FORM AND REDUCED ECHELON FORM: GAUSSIAN ELIMINATION33
Step 6: r5 ← − 51 r5
1 4 5 −9 −7
0
1 2 −3 −3
0 0 0 1 0
0 0 0 0 0
Step 7: r1 ← r1 − 4r2
1 0 −3 3 5
0
1 2 −3 −3
0 0 0 1 0
0 0 0 0 0
Step 8: r1 ← r1 − 3r3 and r2 ← r2 + 3r3
1 0 −3 0 5
0 1
2 0 −3
0 0 0 1 0
0 0 0 0 0
Example 4.3
Use Gauss-Jordan elimination to transform the following matrix first into
row-echelon form and then into reduced row-echelon form
0 3 −6 6 4 −5
3 −7 8 −5 8 9
3 −9 12 −9 6 15
Solution.
By following the steps in the Gauss-Jordan algorithm we find
Step 1: r3 ← 13 r3
0 3 −6 6 4 −5
3 −7 8 −5 8 9
1 −3 4 −3 2 5
Step 2: r1 ↔ r3
1 −3 4 −3 2 5
3 −7 8 −5 8 9
0 3 −6 6 4 −5
34 LINEAR SYSTEMS OF EQUATIONS
Step 3: r2 ← r2 − 3r1
1 −3 4 −3 2 5
0 2 −4 4 2 −6
0 3 −6 6 4 −5
Step 4: r2 ← 21 r2
1 −3 4 −3 2 5
0 1 −2 2 1 −3
0 3 −6 6 4 −5
Step 5: r3 ← r3 − 3r2
1 −3 4 −3 2 5
0 1 −2 2 1 −3
0 0 0 0 1 4
Step 6: r1 ← r1 + 3r2
1 0 −2 3 5 −4
0 1 −2 2 1 −3
0 0 0 0 1 4
Step 7: r1 ← r1 − 5r3 and r2 ← r2 − r3
1 0 −2 3 0 −24
0 1 −2 2 0 −7
0 0 0 0 1 4
Remark 4.2
It can be shown that no matter how the elementary row operations are varied,
one will always arrive at the same reduced row-echelon form; that is, the
reduced row echelon form is unique. On the contrary row-echelon form is
not unique. However, the number of leading 1’s of two different row-echelon
forms is the same. That is, two row-echelon matrices have the same number
of nonzero rows. This number is known as the rank of the matrix and is
denoted by rank(A).
Example 4.4
Consider the system (
ax + by = k
cx + dy = l.
4 ECHELON FORM AND REDUCED ECHELON FORM: GAUSSIAN ELIMINATION35
Solution.
The coefficient matrix is the matrix
a b
.
c d
Step 1: r1 ↔ r2
c d
0 b
Step 2: r1 ← 1c r1 and r2 ← 1b r2
1 dc
0 1
36 LINEAR SYSTEMS OF EQUATIONS
Step 3: r1 ← r1 − dc r2
1 0
0 1
Example 4.5
Find the rank of each of the following matrices
(a)
2 1 4
A = 3 2 5
0 −1 1
(b)
3 1 0 1 −9
B = 0 −2 12 −8 −6
2 −3 22 −14 −17
Solution.
(a) We use Gaussian elimination to reduce the given matrix into row-echelon
form as follows:
Step 1: r2 ← r2 − r1
2 1 4
1 1 1
0 −1 1
Step 2: r1 ↔ r2
1 1 1
2 1 4
0 −1 1
Step 3: r2 ← −r2 + 2r1
1 1 1
0 1 −2
0 −1 1
Step 4: r3 ← −r3 − r2
1 1 1
0 1 −2
0 0 1
Thus, rank(A) = 3.
(b) As in (a), we reduce the matrix into row-echelon form as follows:
4 ECHELON FORM AND REDUCED ECHELON FORM: GAUSSIAN ELIMINATION37
Step 1: r1 ← r1 − r3
1 4 −22 15 8
0 −2 12 −8 −6
2 −3 22 −14 −17
Step 2: r3 ← r3 − 2r1
1 4 −22 15 25
0 −2 12 −8 −6
0 −11 −22 −44 −33
Step 3: r2 ← − 21 r2
1 4 −22 15 8
0 1 −6 4 3
0 −11 −22 −44 −33
Step 4: r3 ← r3 + 11r2
1 4 −22 15 8
0 1 −6 4 3
0 0 −88 0 0
Step 5: r3 ← 18 r3
1 4 −22 15 8
0 1 −6 4 3
0 0 1 0 0
Hence, rank(B) = 3
38 LINEAR SYSTEMS OF EQUATIONS
Practice Problems
Problem 4.1
Use Gaussian elimination to reduce the given matrix to row echelon form.
1 −2 3 1 −3
.
2 −1 3 −1 0
Problem 4.2
Use Gaussian elimination to reduce the given matrix to row echelon form.
−1 0 2 −3
0 3 −1 7 .
3 2 0 7
Problem 4.3
Use Gaussian elimination to reduce the given matrix to row echelon form.
5 −5 −15 40
4 −2 −6 19 .
3 −6 −17 41
Problem 4.4
Use Gaussian elimination to reduce the given matrix to row echelon form.
2 1 1 −1
1 2 1 0 .
3 0 −2 5
Problem 4.5
Which of the following matrices are not in reduced row-ehelon form and
why?
(a)
1 −2 0 0
0 0 0 0
.
0 0 1 0
0 0 0 1
(b)
1 0 0 3
0 2 0 −2 .
0 0 3 0
4 ECHELON FORM AND REDUCED ECHELON FORM: GAUSSIAN ELIMINATION39
(c)
1 0 4
0 1 −2 .
0 0 0
Problem 4.6
Use Gaussian elimination to convert the following matrix into a row-echelon
matrix.
1 −3 1 −1 0 −1
−1 3 0 3 1 3
.
2 −6 3 0 −1 2
−1 3 1 5 1 6
Problem 4.7
Use Gauss-Jordan elimination to convert the following matrix into reduced
row-echelon form.
−2 1 1 15
6 −1 −2 −36
.
1 −1 −1 −11
−5 −5 −5 −14
Problem 4.8
Use Gauss-Jordan elimination to convert the following matrix into reduced
row-echelon form.
3 1 7 2 13
2 −4 14 −1 −10
.
5 11 −7 8 59
2 5 −4 −3 39
Problem 4.9
Use Gaussian elimination to convert the following matrix into row-echelon
form.
−1 −1 0 0
0 0 2 3
.
4 0 −2 1
3 −1 0 4
Problem 4.10
Use Gaussian elimination to convert the following matrix into row-echelon
40 LINEAR SYSTEMS OF EQUATIONS
form.
1 1 2 8
−1 −2 3 1 .
3 −7 4 10
Problem 4.11
Find the rank of each of the following matrices.
(a)
−1 −1 0 0
0 0 2 3
.
4 0 −2 1
3 −1 0 4
(b)
1 −1 3
2 0 4 .
−1 −3 1
Problem 4.12
Use Gauss-Jordan elimination to reduce the matrix into reduced row echelon
form
1 1 1 3
2 3 4 11 .
4 9 16 41
Problem 4.13
Use Gauss-Jordan elimination to reduce the matrix into reduced row echelon
form
2 −2 1 3
3 1 −1 7 .
1 −3 2 0
Problem 4.14
Find the rank of the matrix
2 −1 3
1 0 1
.
0 2 −1
1 1 4
4 ECHELON FORM AND REDUCED ECHELON FORM: GAUSSIAN ELIMINATION41
Problem 4.15
Find the rank of the following matrix.
1 2 3
1 0 4 .
0 2 0
Problem 4.16
Use Gaussian elimination to reduce the matrix to reduced row echelon form.
1 1 1 5
2 3 5 8
4 0 5 2
Problem 4.17
Use Gaussian elimination to reduce the matrix to reduced row echelon form.
1 2 −3 2
6 3 −9 6
7 14 −21 13
Problem 4.18
Use Gaussian elimination to reduce the matrix to reduced row echelon form.
0 4 1 2
2 6 −2 3
4 8 −5 4
Problem 4.19
Find the rank of the matrix
−1 0 −1 2
2 0 2 0
1 0 1 −1
Problem 4.20
Find the rank of the matrix
1 2 −1 3 0 1
−1 −2 2 −2 −1 1
1 2 0 4 0 6
0 0 2 2 −1 7
42 LINEAR SYSTEMS OF EQUATIONS
Step 4: r3 ↔ r4
1 3 −2 0 2 0 0
0 0 1 2 0 3 1
0 0 0 0 0 6 2
0 0 0 0 0 0 0
Step 5: r3 ← 61 r3
1 3 −2 0 2 0 0
0 0 1 2 0 3 1
0 0 0 0 0 1 13
0 0 0 0 0 0 0
The leading variables are x1 , x3 , and x6 . The free variables are x2 , x4 , and x5
One way to solve a linear system is to apply the elementary row opera-
tions to reduce the augmented matrix to a (reduced) row-echelon form. If
the augmented matrix is in reduced row-echelon form then to obtain the gen-
eral solution one just has to move all independent variables to the right side
of the equations and consider them as parameters. The dependent variables
are given in terms of these parameters.
Example 5.2
Solve the following linear system.
x1 + 2x2
+ x4 =6
x3 + 6x4 =7
x5 = 1.
Solution.
The augmented matrix is already in row-echelon form. The free variables are
x2 and x4 . So let x2 = s and x4 = t. Solving the system starting from the
bottom we find x1 = −2s − t + 6, x3 = 7 − 6t, and x5 = 1
44 LINEAR SYSTEMS OF EQUATIONS
If the augmented matrix does not have the reduced row-echelon form but
the row-echelon form then the general solution also can be easily found by
using the method of backward substitution.
Example 5.3
Solve the following linear system
x1 − 3x2 + x3 − x4 = 2
x2 + 2x3 − x4 = 3
x3 + x4 = 1.
Solution.
The augmented matrix is in row-echelon form. The free variable is x4 = t.
Solving for the leading variables we find, x1 = 11t + 4, x2 = 3t + 1, and
x3 = 1 − t
Theorem 5.1
A system of m linear equations in n unknowns can have exactly one solution,
infinitely many solutions, or no solutions at all. If the augmented matrix in
the (reduced) row echelon form
(1) has a row of the form [0, 0, · · · , 0, b] where b is a nonzero constant, then
the system has no solutions;
(2) has independent variables and no rows of the form [0, 0, · · · , 0, b] with
b 6= 0 then the system has infinitely many solutions;
(3) no independent variables and no rows of the form [0, 0, · · · , 0, b] with
b 6= 0, then the system has exactly one solution.
Example 5.4
Find the general solution of the system whose augmented matrix is given by
1 2 −7
−1 −1 1 .
2 1 5
5 ECHELON FORMS AND SOLUTIONS TO LINEAR SYSTEMS 45
Solution.
We first reduce the system to row-echelon form as follows.
Step 2: r3 ← r3 + 3r2
1 2 −7
0 1 −6
0 0 1
The corresponding system is given by
x1 + 2x2 = −7
x2 = −6
0x1 + 0x2 = 1
Example 5.5
Find the general solution of the system whose augmented matrix is given by
1 −2 0 0 7 −3
0
1 0 0 −3 1
.
0 0 0 1 5 −4
0 0 0 0 0 0
Solution.
By adding two times the second row to the first row we find the reduced
row-echelon form of the augmented matrix.
1 0 0 0 1 −1
0 1 0 0 −3 1
0 0 0 1 5 −4
0 0 0 0 0 0
It follows that the free variables are x3 = s and x5 = t. Solving for the leading
variables we find x1 = −1 − t, x2 = 1 + 3t, and x4 = −4 − 5t
46 LINEAR SYSTEMS OF EQUATIONS
Example 5.6
Determine the value(s) of h such that the following matrix is the augmented
matrix of a consistent linear system
1 4 2
.
−3 h −1
Solution.
By adding three times the first row to the second row we find
1 4 2
0 12 + h 5
The system is consistent if and only if 12 + h 6= 0; that is, h 6= −12
Example 5.7
Find (if possible) conditions on the numbers a, b, and c such that the following
system is consistent
x1 + 3x2 + x3 = a
−x − 2x2 + x3 = b
1
3x1 + 7x2 − x3 = c.
Solution.
The augmented matrix of the system is
1 3 1 a
−1 −2 1 b
3 7 −1 c
Now apply Gaussian elimination as follows.
Practice Problems
Problem 5.1
Using Gaussian elimination, solve the linear system whose augmented matrix
is given by
1 1 2 8
−1 −2 3 1 .
3 −7 4 10
Problem 5.2
Solve the linear system whose augmented matrix is reduced to the following
reduced row-echelon form
1 0 0 −7 8
0 1 0 3 2 .
0 0 1 1 −5
Problem 5.3
Solve the linear system whose augmented matrix is reduced to the following
row-echelon form
1 −3 7 1
0 1 4 0 .
0 0 0 1
Problem 5.4
Solve the following system using Gauss-Jordan elimination.
3x1 + x2 + 7x3 + 2x4 = 13
2x − 4x + 14x − x = −10
1 2 3 4
5x1 + 11x2 − 7x3 + 8x4 = 59
2x1 + 5x2 − 4x3 − 3x4 = 39.
Problem 5.5
Solve the following system.
2x1 + x2 + x3 = −1
x1 + 2x2 + x3 = 0
3x1 − 2x3 = 5.
48 LINEAR SYSTEMS OF EQUATIONS
Problem 5.6
Solve the following system using elementary row operations on the augmented
matrix:
5x1 − 5x2 − 15x3 = 40
4x1 − 2x2 − 6x3 = 19
3x1 − 6x2 − 17x3 = 41.
Problem 5.7
Reduce the following system to row echelon form and then find the solution.
2x1 + x2 − x3 + 2x4 = 5
4x + 5x − 3x + 6x = 9
1 2 3 4
−2x1 + 5x2 − 2x3 + 6x4 = 4
4x1 + 11x2 − 4x3 + 8x4 = 2.
Problem 5.8
Reduce the following system to row echelon form and then find the solution.
2x1 − 5x2 + 3x3 = −4
x1 − 2x2 − 3x3 = 3
−3x1 + 4x2 + 2x3 = −4.
Problem 5.9
Reduce the following system to reduced row echelon form and then find the
solution.
2x1 + 4x2 + 2x3 + 4x4 + 2x5 = 4
2x + 4x + 3x + 3x + 3x = 4
1 2 3 4 5
3x1 + 6x2 + 6x3 + 3x4 + 6x5 = 6
x3 − x4 − x5 = 4.
Problem 5.10
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem.
x1 + 2x2 + 3x3 + 4x4 + 3x5 = 1
2x1 + 4x2 + 6x3 + 2x4 + 6x5 = 2
3x1 + 6x2 + 18x3 + 9x4 + 9x5 = −6
4x1 + 8x2 + 12x3 + 10x4 + 12x5 = 4
5x1 + 10x2 + 24x3 + 11x4 + 15x5 = −4.
5 ECHELON FORMS AND SOLUTIONS TO LINEAR SYSTEMS 49
Problem 5.11
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem.
x1 + x2 + x3 = 5
2x1 + 3x2 + 5x3 = 8
4x1 + 5x3 = 2.
Problem 5.12
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem.
x1 + 2x2 − 3x3 = 2
6x1 + 3x2 − 9x3 = 6
7x1 + 14x2 − 21x3 = 13.
Problem 5.13
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem.
4x1 + 8x2 − 5x3 = 4
2x1 + 6x2 − 2x3 = 3
4x2 + x3 = 2.
Problem 5.14
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem. (
x1 − 2x2 + x3 − 3x4 = 0
3x1 − 6x2 + 2x3 − 7x4 = 0.
Problem 5.15
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem.
−x1 + 3x2 + x3 = 16
x1 − 5x2 + 3x3 = 2
−2x1 + 5x2 − 3x3 = 6.
Problem 5.16
Using the Gauss-Jordan elimination method, solve the following linear sys-
50 LINEAR SYSTEMS OF EQUATIONS
tem.
x1 + x2 + x3 = 5
2x1 + 3x2 + 5x3 = 8
4x1 + 5x3 = 2.
Problem 5.17
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem.
x1 + 2x2 − 3x3 = 2
6x1 + 3x2 − 9x3 = 6
7x1 + 14x2 − 21x3 = 13.
Problem 5.18
Using the Gauss-Jordan elimination method, solve the following linear sys-
tem.
4x2 + x3 = 2
2x1 + 6x2 − 2x3 = 3
4x1 + 8x2 − 5x3 = 4.
Problem 5.19
Using the Gaussian elimination method, solve the following linear system.
−x1 − x3 = 2
2x1 − 2x3 = 0
x1 − x3 = −1.
Problem 5.20
Using the Gaussian elimination method, solve the following linear system.
2x1 + 4x2 + 4x3 + 2x4 = 16
4x + 8x + 6x + 8x = 32
1 2 3 4
14x1 + 29x2 + 32x3 + 16x4 = 112
10x1 + 17x2 + 10x3 + 2x4 = 28.
Problem 5.21
Consider a system of m linear equations in n unknowns such that the aug-
mented matrix in the (reduced) row echelon form has a row of the form
[0, 0, · · · , 0, b] where b is a nonzero constant. Show that the system has no
solutions.
5 ECHELON FORMS AND SOLUTIONS TO LINEAR SYSTEMS 51
Problem 5.22
Consider a system of m linear equations in n unknowns such that the aug-
mented matrix in the (reduced) row echelon form has independent variables
and no rows of the form [0, 0, · · · , 0, b] with b 6= 0. Show that the system has
infinitely many solutions.
Problem 5.23
Consider a system of m linear equations in n unknowns such that the aug-
mented matrix in the (reduced) row echelon form has no independent vari-
ables and no rows of the form [0, 0, · · · , 0, b] with b 6= 0. Show that the system
has exactly one solution.
52 LINEAR SYSTEMS OF EQUATIONS
Theorem 6.1
A homogeneous system in n unknowns and m equations has infinitely many
solutions if either
(1) the rank of the row echelon form of the augmented matrix is less than n;
or
(2) the number of unknowns exceeds the number of equations, i.e. m < n.
That is, the system is underdetermined.
Proof.
Applying the Gauss-Jordan elimination to the augmented matrix [A|0] we ob-
tain the matrix [B|0]. The number of nonzero rows of B is equals to rank(A).
(1) Suppose first that rank(A) = r < n. In this case, the system Bx = 0
has r equations in n unknowns. Thus, the system has n − r independent
variables and consequently the system Bx = 0 has a nontrivial solution. By
Theorem 2.1, the system Ax = 0 has a nontrivial solution.
(2) Suppose m < n. If the system has only the trivial solution then by (1)
we must have rank(A) ≥ n. This implies that m < n ≤ rank(A) ≤ m, a
contradiction
6 HOMOGENEOUS SYSTEMS OF LINEAR EQUATIONS 53
Example 6.1
Solve the following homogeneous system using Gauss-Jordan elimination.
2x1 + 2x2 − x3
+ x5 = 0
−x − x + 2x − 3x + x = 0
1 2 3 4 5
x1 + x2 − 2x3 − x5 = 0
x3 + x4 + x5 = 0.
Solution.
The reduction of the augmented matrix to reduced row-echelon form is out-
lined below.
2 2 −1 0 1 0
−1 −1 2 −3 1 0
1 1 −2 0 −1 0
0 0 1 1 1 0
Step 1: r3 ← r3 + r2
2 2 −1 0 1 0
−1 −1 2 −3 1 0
0 0 0 −3 0 0
0 0 1 1 1 0
Step 2: r3 ↔ r4 and r1 ↔ r2
−1 −1 2 −3 1 0
2
2 −1 0 1 0
0 0 1 1 1 0
0 0 0 −3 0 0
Step 3: r2 ← r2 + 2r1 and r4 ← − 13 r4
−1 −1 2 −3 1 0
0
0 3 −6 3 0
0 0 1 1 1 0
0 0 0 1 0 0
Step 4: r1 ← −r1 and r2 ← 13 r2
1 1 −2 3 −1 0
0 0
1 −2 1 0
0 0 1 1 1 0
0 0 0 1 0 0
54 LINEAR SYSTEMS OF EQUATIONS
Step 5: r3 ← r3 − r2
1 1 −2 3 −1 0
0
0 1 −2 1 0
0 0 0 3 0 0
0 0 0 1 0 0
Step 6: r4 ← r4 − 31 r3 and r3 ← 13 r3
1 1 −2 3 −1 0
0
0 1 −2 1 0
0 0 0 1 0 0
0 0 0 0 0 0
Step 7: r1 ← r1 − 3r3 and r2 ← r2 + 2r3
1 1 −2 0 −1 0
0 0 1 0 1 0
0 0 0 1 0 0
0 0 0 0 0 0
Step 8: r1 ← r1 + 2r2
1 1 0 0 1 0
0 0 1 0 1 0
.
0 0 0 1 0 0
0 0 0 0 0 0
The corresponding system is
x 1 + x 2
+ x5 = 0
x3 + x5 = 0
x4 = 0.
The free variables are x2 = s, x5 = t and the general solution is given by the
formula: x1 = −s − t, x2 = s, x3 = −t, x4 = 0, x5 = t
Example 6.2
Solve the following homogeneous system using Gaussian elimination.
x1 + 3x2 + 5x3 + x4 = 0
4x1 − 7x2 − 3x3 − x4 = 0
3x1 + 2x2 + 7x3 + 8x4 = 0.
6 HOMOGENEOUS SYSTEMS OF LINEAR EQUATIONS 55
Solution.
The augmented matrix for the system is
1 3 5 1 0
4 −7 −3 −1 0
3 2 7 8 0
Remark 6.1
Part (2) of Theorem 6.1 applies only to homogeneous linear systems. A
non-homogeneous system (right-hand side has non-zero entries) with more
unknowns than equations need not be consistent as shown in the next exam-
ple.
56 LINEAR SYSTEMS OF EQUATIONS
Example 6.3
Show that the following system is inconsistent.
(
x1 + x2 + x3 = 0
2x1 + 2x2 + 2x3 = 4.
Solution.
Multiplying the first equation by −2 and adding the resulting equation to the
second we obtain 0 = 4 which is impossible. So the system is inconsistent
6 HOMOGENEOUS SYSTEMS OF LINEAR EQUATIONS 57
Practice Problems
Problem 6.1
Justify the following facts:
(a) All homogeneous linear systems are consistent.
(b) A homogeneous linear system with fewer equations than unknowns has
infinitely many solutions.
Problem 6.2
Find the value(s) of a for which the following system has a non-trivial solu-
tion. Find the general solution.
x1 + 2x2 + x3 = 0
x1 + 3x2 + 6x3 = 0
2x + 3x + ax = 0.
1 2 3
Problem 6.3
Solve the following homogeneous system.
x1 − x2 + 2x3 + x4 = 0
2x1 + 2x2 − x4 = 0
3x1 + x2 + 2x3 + x4 = 0.
Problem 6.4
Solve the homogeneous linear system.
x1 + x2 − 2x3 = 0
3x1 + 2x2 + 4x3 = 0
4x1 + 3x2 + 3x3 = 0.
Problem 6.5
Solve the homogeneous linear system.
x1 + x2 − 2x3 = 0
3x1 + 2x2 + 4x3 = 0
4x1 + 3x2 + 2x3 = 0.
58 LINEAR SYSTEMS OF EQUATIONS
Problem 6.6
Solve the homogeneous linear system.
(
2x1 + 4x2 − 6x3 = 0
4x1 + 8x2 − 12x3 = 0.
Problem 6.7
Solve the homogeneous linear system.
x1 + x2
+ 3x4 = 0
2x1 + x2 − x3 + x4 = 0
3x1 − x2 − x3 + 2x4 = 0.
Problem 6.8
Solve the homogeneous linear system.
x1 + x2
− x4 = 0
−2x1 − 3x2 + 4x3 + 5x4 = 0
2x1 + 4x2 − 2x4 = 0.
Problem 6.9
Solve the homogeneous system.
3x1 + 2x2 + x3 = 0
6x1 + 4x2 =0
x2 + x3 = 0.
Problem 6.10
Use Gaussian elimination to solve the following homogeneous system of equa-
tions.
x1 − x2 − x3 + 3x4 = 0
x1 + x2 − 2x3 + x4 = 0
4x1 − 2x2 + 4x3 + x4 = 0.
Problem 6.11
Use Gaussian elimination to solve the following homogeneous system of equa-
tions.
−x1 + x2 − x3 = 0
3x1 − x2 − x3 = 0
2x1 + x2 − 3x3 = 0.
6 HOMOGENEOUS SYSTEMS OF LINEAR EQUATIONS 59
Problem 6.12
Use Gaussian elimination to solve the following homogeneous system of equa-
tions.
x1 − x2 − x3 + x4 = 0
2x1 − 2x2 + x3 + x4 = 0
5x1 − 5x2 − 2x3 + 4x4 = 0.
Problem 6.13
Without solving the system, justify the reason why the system has infinitely
many solutions. (
x1 − x2 + 3x3 = 0
2x1 + x2 + 3x3 = 0.
Problem 6.14
By just using the rank of the augmented matrix, show that the following
system has infinitely many solutions.
−x1 − x2
=0
− 2x3 + x4 = 0
4x
1
− 4x2 − 2x3 + x4 = 0
3x1 − x2 + 4x4 = 0.
Problem 6.15
Show that the following two systems are equivalent:
−x1 + x2 + 4x3 = 0
x1 + 3x2 + 8x3 = 0
x1 + 2x2 + 5x3 = 0
and (
x1 − x3 = 0
x2 + 3x3 = 0.
Problem 6.16
True or false. Justify your answer.
(a) All homogeneous linear systems are consistent.
(b) Linear systems with more unknowns than equations are always consistent.
(c) A homogeneous system with fewer equations than unknowns has infinitely
many solutions.
60 LINEAR SYSTEMS OF EQUATIONS
Problem 6.17
Solve the system
x1 + 2x2 − x3 + 3x4
=0
−x − 2x + 2x − 2x − x = 0
1 2 3 4 5
x1 + 2x2 + 4x4 =0
2x3 + 2x4 − x5 = 0.
Problem 6.18
Show that if (x1 , x2 , · · · , xn ) and (y1 , y2 , · · · , yn ) are two solutions to a m ×
n homogeneous linear system then α(x1 , x2 , · · · , xn ) + β(y1 , y2 , · · · , yn ) =
(αx1 + βy1 , αx2 + βy2 , · · · , αxn + βyn ) is also a solution.
Problem 6.19
Show with an example that the result of Problem 6.19 fails for linear non-
homogeneous systems.
Problem 6.20
If the Gauss-Jordan form of the augmented matrix of a homogeneous system
has a row of zeros, are there necessarily any free variables? If there are free
variables, is there necessarily a row of zeros?
Matrices
Matrices are essential in the study of linear algebra. The concept of matrices
has become a tool in all branches of mathematics, the sciences, and engi-
neering. They arise in many contexts other than as augmented matrices for
systems of linear equations. In this chapter we shall consider this concept as
objects in their own right and develop their properties for use in our later
discussions.
61
62 MATRICES
Two matrices are said to be equal if they have the same size and their cor-
responding entries are all equal. If the matrix A is not equal to the matrix
B we write A 6= B.
Example 7.1
Find x1 , x2 and x3 such that
x1 + x2 + 2x3 0 1 9 0 1
2 3 2x1 + 4x2 − 3x3 = 2 3 1 .
4 3x1 + 6x2 − 5x3 5 4 0 5
7 MATRICES AND MATRIX OPERATIONS 63
Solution.
Because corresponding entries must be equal, this gives the following linear
system
x1 + x2 + 2x3 = 9
2x1 + 4x2 − 3x3 = 1
3x1 + 6x2 − 5x3 = 0
Example 7.2
Solve the following matrix equation for a, b, c, and d
a−b b+c 8 1
=
3d + c 2a − 4d 7 6
Solution.
Equating corresponding entries we get the system
a−b
=8
b+c =1
c + 3d = 7
2a − 4d = 6.
The augmented matrix is
1 −1 0 0 8
0 1 1 0 1
.
0 0 1 3 7
2 0 0 −4 6
We next apply Gaussian elimination as follows.
Step 1: r4 ← r4 − 2r1
1 −1 0 0 8
0 1 1 0 1
0 0 1 3 7
0 2 0 −4 −10
Step 2: r4 ← r4 − 2r2
1 −1 0 0 8
0 1 1 0 1
0 0 1 3 7
0 0 −2 −4 −12
Step 3: r4 ← r4 + 2r3
1 −1 0 0 8
0 1 1 0 1
.
0 0 1 3 7
0 0 0 2 2
7 MATRICES AND MATRIX OPERATIONS 65
Step 4: r4 ← 21 r4
1 −1 0 0 8
0 1 1 0 1
.
0 0 1 3 7
0 0 0 1 1
The equivalent system to the given one is
a−b
=8
b+c =1
c + 3d = 7
d = 1.
Example 7.3
Consider the matrices
2 1 2 1 2 1 0
A= ,B = ,C = .
3 4 3 5 3 4 0
Solution.
We have
4 2
A+B = .
6 9
A + C and B + C are undefined since A and C are of different sizes as well
as B and C
From now on, a constant number will be called a scalar. The scalar prod-
uct of a real number c and a matrix A = [aij ] is the matrix cA = [caij ]. That
is, the product cA is the matrix obtained by multiplying each entry of A by
c. Hence, −A = (−1)A. We define, A − B = A + (−B).
66 MATRICES
Example 7.4
Consider the matrices
2 3 4 0 2 7
A= ,B = .
1 2 1 1 −3 5
Compute A − 3B.
Solution.
Using the above definitions we have
2 −3 −17
A − 3B =
−2 11 −14
The following theorem list the properties of matrix addition and multiplica-
tion of a matrix by a scalar.
Theorem 7.1
Let A, B, and C be m × n and let c, d be scalars. Then
(i) A + B = B + A,
(ii) (A + B) + C = A + (B + C) = A + B + C,
(iii) A + 0 = 0 + A = A,
(iv) A + (−A) = 0,
(v) c(A + B) = cA + cB,
(vi) (c + d)A = cA + dA,
(vii) (cd)A = c(dA).
Example 7.5
Solve the following matrix equation.
3 2 a b 1 0
+ = .
−1 1 c d −1 2
Solution.
Adding and then equating corresponding entries we obtain a = −2, b =
−2, c = 0, and d = 1
If A is a square matrix then the sum of the entries on the main diagonal
is called the trace of A and is denoted by tr(A).
7 MATRICES AND MATRIX OPERATIONS 67
Example 7.6
Find the trace of the coefficient matrix of the system
− x2 + 3x3 = 1
x1 + 2x3 = 2
−3x1 − 2x2 = 4.
Solution.
If A is the coefficient matrix of the system then
0 −1 3
A= 1 0 2 .
−3 −2 0
The trace of A is the number tr(A) = 0 + 0 + 0 = 0
Two useful properties of the trace of a matrix are given in the following
theorem.
Theorem 7.2
Let A = [aij ] and B = [bij ] be two n × n matrices and c be a scalar. Then
(i) tr(A + B) = tr(A) + tr(B),
(ii) tr(cA) = c tr(A).
Proof.
= ni=1 (aii +Pbii ) = ni=1 aii + ni=1 bii = tr(A) + tr(B).
P P P
(i) tr(A + B)P
(ii) tr(cA) = ni=1 caii = c ni=1 aii = c tr(A)
Example 7.7
Find the transpose of the matrix
2 3 4
A= .
1 2 1
68 MATRICES
Solution.
The transpose of A is the matrix
2 1
A T = 3 2
4 1
The following result lists some of the properties of the transpose of a matrix.
Theorem 7.3
Let A = [aij ], and B = [bij ] be two m × n matrices, C = [cij ] be an n × n
matrix, and c a scalar. Then
(i) (AT )T = A,
(ii) (A + B)T = AT + B T ,
(iii) (cA)T = cAT ,
(iv) tr(C T ) = tr(C).
Proof.
(i) (AT )T = [aji ]T = [aij ] = A.
(ii) (A + B)T = [aij + bij ]T = [aji + bji ] = [aji ] + [bji ] = AT + B T .
(iii) (cA)T = [ca T T
Pijn] = [caji ] = c[aji ] = cA .
T
(iv) tr(C ) = i=1 cii = tr(C)
Example 7.8
A square matrix A is called symmetric if AT = A. A square matrix A is
called skew-symmetric if AT = −A.
(a) Show that the matrix
1 2 3
A = 2 3 1
3 1 2
is a symmetric matrix.
(b) Show that the matrix
0 2 3
A = −2 0 −4
−3 4 0
is a skew-symmetric matrix.
(c) Show that a matrix that is both symmetric and skew-symmetric is the
7 MATRICES AND MATRIX OPERATIONS 69
zero matrix.
(d) Show that for any square matrix A, the matrix S = 21 (A + AT ) is sym-
metric and the matrix K = 12 (A − AT ) is skew-symmetric.
(e) Show that if A is a square matrix, then A = S + K, where S is symmetric
and K is skew-symmetric.
(f) Show that the representation in (d) is unique.
Solution.
(a) A is symmetric since
1 2 3
AT = 2 3 1 = A.
3 1 2
Example 7.9
Let A be an n × n matrix.
(a) Show that if A is symmetric then A and AT have the same main diagonal.
(b) Show that if A is skew-symmetric then the entries on the main diagonal
are 0.
(c) If A and B are symmetric then so is A + B.
70 MATRICES
Solution.
(a) Let A = [aij ] be symmetric. Let AT = [bij ]. Then bij = aji for all
1 ≤ i, j ≤ n. In particular, when i = j we have bii = aii . That is, A and AT
have the same main diagonal.
(b) Since A is skew-symmetric, we have aij = −aji . In particular, aii = −aii
and this implies that aii = 0.
(c) Suppose A and B are symmetric. Then (A + B)T = AT + B T = A + B.
That is, A + B is symmetric
Example 7.10
Let A be an m × n matrix and α a real number. Show that if αA = 0 then
either α = 0 or A = 0.
Solution.
Let A = [aij ]. Then αA = [αaij ]. Suppose αA = 0. Then αaij = 0 for all
0 ≤ i ≤ m and 0 ≤ j ≤ n. If α 6= 0 then aij = 0 for all indices i and j. In
this case, A = 0
7 MATRICES AND MATRIX OPERATIONS 71
Practice Problems
Problem 7.1
Compute the matrix
T
2 1 1 −1
3 −2 .
−1 0 2 3
Problem 7.2
Find w, x, y, and z.
1 2 w 1 2 −1
2 x 4 = 2 −3 4 .
y −4 z 0 −4 5
Problem 7.3
Determine two numbers s and t such that the following matrix is symmetric.
2 s t
A = 2s 0 s + t .
3 3 t
Problem 7.4
Let A be the matrix
a b
A= .
c d
Show that
1 0 0 1 0 0 0 0
A=a +b +c +d .
0 0 0 0 1 0 0 1
Problem 7.5
Let A = 1 1 −1 , B = 0 1 2 , C = 3 0 1 . Suppose rA+sB+tC =
0. Show that s = r = t = 0.
Problem 7.6
Compute
1 9 −2 8 −4 3
+ .
3 6 0 −7 1 6
72 MATRICES
Problem 7.7
Determine whether the matrix is symmetric or skew-symmetric.
11 6 1
A= 6 3 −1 .
1 −1 −6
Problem 7.8
Determine whether the matrix is symmetric or skew-symmetric.
0 3 −1 −5
−3 0 7 −2
A= 1 −7
.
0 0
5 2 0 0
Problem 7.9
Consider the matrix
0 3 −1 −5
−3 0 7 −2
A=
1 −7
.
0 0
5 2 0 0
Find 4tr(7A).
Problem 7.10
Consider the matrices
11 6 1 0 3 −1
A= 6 3 −1 , B = −3 0 7 .
1 −1 −6 1 −7 0
Problem 7.11
Write the matrix A =
12 7 1
−2 −4 0
0 −8 2
as a sum of a symmetric matrix S and a skew-symmetric matrix K.
7 MATRICES AND MATRIX OPERATIONS 73
Problem 7.12
Show that there is no square matrix A of dimension n such that A−AT = In .
Problem 7.13
What value(s) of x satisfies the matrix equation
T
0 −x 0 7x + 14
= .
3x − 2 5 x+8 5
Problem 7.14
2 −3
1 −2 0
Let A = 0 2 and B = . Calculate 3(AT + 2B).
5 1 0
7 −2
Problem 7.15
Give an example of a 5 × 5 symmetric matrix with non-zero entries.
Problem 7.16
Solve the system
2 −5 3 −4
x1 1 + x2 −2 + x3 −3 =
3
−3 4 2 −4
Problem 7.17
Let A and B be two square matrices with trace 0. Show that tr(αA+βB) = 0
where α and β are scalars.
Problem 7.18
Let f be a function defined on the set of all n × n matrices given by f (A) =
tr(A). Show that f (αA + βB) = αf (A) + βf (B), where α and β are scalars.
Problem 7.19
Let f be a function defined on the set of all m × n matrices given by f (A) =
AT . Show that f (αA + βB) = αf (A) + βf (B), where α and β are scalars.
Problem 7.20
Suppose that A and B are skew-symmetric. Show that αA + βB is also
skew-symmetric, where α and β are scalars.
Problem 7.21
Prove Theorem 7.1.
74 MATRICES
8 Matrix Multiplication
In the previous section we discussed some basic properties associated with
matrix addition and scalar multiplication. Here we introduce another impor-
tant operation involving matrices−the product.
Let A = [aij ] be a matrix of size m × n and B = (bij ) be a matrix of size
n × p. Then the product matrix is a matrix of size m × p and entries
(AB)ij = ai1 b1j + ai2 b2j + ... + ain bnj ,
that is, (AB)ij is obtained by multiplying componentwise the entries of the
ith row of A by the entries of the j th column of B. It is very important to keep
in mind that the number of columns of the first matrix must be equal to the
number of rows of the second matrix; otherwise the product is undefined.
An interesting question associated with matrix multiplication is the following:
If A and B are square matrices then is it always true that AB = BA?
The answer to this question is negative. In general, matrix multiplication is
not commutative, as the following example shows.
Example 8.1
Let
1 2 2 −1
A= ,B = .
3 2 −3 4
Show that AB 6= BA. Hence, matrix multiplication is not commutative.
Solution.
Using the definition of matrix multiplication we find
−4 7 −1 2
AB = , BA = .
0 5 9 2
Hence, AB 6= BA
Example 8.2
Consider the matrices
2 1 2 1 −1 −2
A= ,B = ,C = .
3 4 3 5 11 4
(a) Compare A(BC) and (AB)C.
(b) Compare A(B + C) and AB + AC.
(c) Compute I2 A and AI2 , where I2 is the 2 × 2 identity matrix.
8 MATRIX MULTIPLICATION 75
Solution.
(a)
70 14
A(BC) = (AB)C = .
235 56
(b)
16 7
A(B + C) = AB + AC = .
59 33
(c) AI2 = I2 A = A
Example 8.3
Let A be a m × n and B be a n × p matrices. Show that if
(a) B has a column of zeros then the same is true for AB.
(b) A has a row of zeros then the same is true for AB.
Solution.
(a) Suppose the the j th column of B is a column of zeros. That is, b1j =
b2j = · · · = bnj = 0. Then for 1 ≤ i ≤ m, we have
n
X
(AB)ij = aik bkj = 0.
k=1
Example 8.4
Let A and B be two n × n matrices.
(a) Show that tr(AB) = tr(BA).
(b) Show that AB − BA = In is impossible.
Solution.
(a) Let A =
P[a ij ]P
and B = [bij ]. Then
tr(AB) = i=1 ( nk=1 aik bki ) = nk=1 ( ni=1 bki aik ) = tr(BA).
n P P
76 MATRICES
Then the matrix of the coefficients of the xi ’s is called the coefficient ma-
trix:
a11 a12 ... a1n
a21 a22 ... a2n
A= ...
.
... ... ...
am1 am2 ... amn
The matrix of the coefficients of the xi ’s and the right hand side coefficients
is called the augmented matrix:
a11 a12 ... a1n b1
a21 a22 ... a2n b2
.
... ... ... ... ...
am1 am2 ... amn bm
Now, if we let
x1
x2
x = ..
.
xn
and
b1
b2
b = ..
.
bm
then the above system can be represented in matrix notation as
Ax = b.
8 MATRIX MULTIPLICATION 77
Example 8.5
Consider the linear system
x1 − 2x2 + x3 = 0
2x2 − 8x3 = 8
−4x1 + 5x2 + 9x3 = −9.
(a) Find the coefficient and augmented matrices of the linear system.
(b) Find the matrix notation.
Solution.
(a) The coefficient matrix of this system is
1 −2 1
0 2 −8
−4 5 9
and the augmented matrix is
1 −2 1 0
0 2 −8 8 .
−4 5 9 −9
(b) We can write the given system in matrix form as
1 −2 1 x1 0
0 2 −8 x2 =
8
−4 5 9 x3 −9
As the reader has noticed so far, most of the basic rules of arithmetic of real
numbers also hold for matrices but a few do not. In Example 8.1 we have
seen that matrix multiplication is not commutative. The following exercise
shows that the cancellation law of numbers does not hold for matrix product.
Example 8.6
(a) Consider the matrices
1 0 0 0 0 0
A= ,B = ,C = .
0 0 1 0 0 1
Compare AB and AC. Is it true that B = C?
(b) Find two square matrices A and B such that AB = 0 but A 6= 0 and
B 6= 0.
78 MATRICES
Solution.
(a) Note that B 6= C even though AB = AC = 0.
(b) The given matrices satisfy AB = 0 with A 6= 0 and B 6= 0
Theorem 8.1
Let A be a matrix of size m × n. Then
(a) A(BC) = (AB)C, where B is of size n × p, C of size p × q.
(b) A(B + C) = AB + AC, where B and C are of size n × p.
(c) (B + C)A = BA + CA, where B and C are of size l × m.
(d) c(AB) = (cA)B = A(cB), where c denotes a scalar and B is of size n × p.
(e) Im A = AIn = A.
Theorem 8.2
Let A = [aij ], B = [bij ] be matrices of sizes m × n and n × m respectively.
Then (AB)T = B T AT .
Example 8.7
Let A be any matrix. Show that AAT and AT A are symmetric matrices.
Solution.
First note that for any matrix A the matrices AAT and AT A are well-
defined. Since (AAT )T = (AT )T AT = AAT , AAT is symmetric. Similarly,
(AT A)T = AT (AT )T = AT A
Example 8.8
Suppose that
1 2
A= .
3 4
Compute A3 .
8 MATRIX MULTIPLICATION 79
Solution.
Multiplying the matrix A by itself three times we obtain
3 37 54
A =
81 118
We end this section with the following result about matrix powers properties.
Theorem 8.3
For any non-negative integers s, t we have
(a) As+t = As At
(b) (As )t = Ast .
Proof.
(a) Fix s. Let t = 1. Then by definition above we have As+1 = As A. Now,
we prove by induction on t that As+t = As At . The equality holds for t = 1.
As the induction hypothesis, suppose that As+t = As At . Then As+(t+1) =
A(s+t)+1 = As+t A = (As At )A = As (At A) = As At+1 .
(b) Fix s. We prove by induction on t that (As )t = Ast . The equality holds
for t = 1. As the induction hypothesis, suppose that (As )t = Ast . Then
(As )t+1 = (As )t (As ) = (Ast )As = Ast+s = As(t+1)
80 MATRICES
Practice Problems
Problem 8.1
Write the linear system whose augmented matrix is given by
2 −1 0 −1
−3 2 1 0 .
0 1 1 3
Problem 8.2
Consider the linear system
2x1 + 3x2 − 4x3 + x4 = 5
−2x1 + x3 =7
3x1 + 2x2 − 4x3 = 3.
(a) Find the coefficient and augmented matrices of the linear system.
(b) Find the matrix notation.
Problem 8.3
Let A be an arbitrary matrix. Under what conditions is the product AAT
defined?
Problem 8.4
An n × n matrix A is said to be idempotent if A2 = A.
(a) Show that the matrix
1 1 1
A=
2 1 1
is idempotent.
(b) Show that if A is idempotent then the matrix (In −A) is also idempotent.
Problem 8.5
The purpose of this exercise is to show that the rule (ab)n = an bn does not
hold with matrix multiplication. Consider the matrices
2 −4 3 2
A= ,B = .
1 3 −1 5
Problem 8.6
Show that AB = BA if and only if AT B T = B T AT .
Problem 8.7
Let A and B be symmetric matrices. Show that AB is symmetric if and only
if AB = BA.
Problem 8.8
A matrix B is said to be the square root of a matrix A if BB = A. Find
two square roots of the matrix
2 2
A= .
2 2
Problem 8.9
Find k such that
1 1 0 k
k 1 1 1 0
2 1 = 0.
0 2 −3 1
Problem 8.10
Express the matrix notation as a system of linear equations.
3 −1 2 x1 2
4 3 7 x2 = −1 .
−2 1 5 x3 4
Problem 8.11
1 2 0 −3
The augmented matrix 0 1 −3 0 represents a system of equations in
0 0 a −4
the variables x1 , x2 , and x3 .
(a) For what values of a is there no solution?
(b) For what values of a is there exactly one solution?
(c) Find the solution to this system of equations if a = −2.
Problem 8.12
Consider the following system of equations
x1 + x2 + 2x3 = 1
2x1 + 3x2 + 3x3 = −2
3x1 + 3x2 + 7x3 = 3.
82 MATRICES
Problem 8.13
12 −1 −3 2 0 −3 2
Consider the matrices A = −5 1 1 and B = −1 3 3 0 .
−3 0 1 −2 4 −1 2
Determine the dimension of AB and find the value of (2, 3) entry of AB.
Problem 8.14 2
1 −2 0
Compute 0 3 1 .
−4 2 1
Problem 8.15
1 −2 3
−1 2 −3
Let A = . Solve the system AT x = 0.
2 −3 1
−1 1 2
Problem 8.16
A square matrix A of size n is said to be orthogonal if AAT = AT A = In .
Show that the matrix
cos θ − sin θ
A=
sin θ cos θ
is orthogonal.
Problem 8.17
If
1 −2
A=
−2 5
and
−1 2 −1
AB = ,
6 −9 3
deteremine the first and the second columns of B.
8 MATRIX MULTIPLICATION 83
Problem 8.18
Let L and D be two square matrices each of size n with D being a diagonal
matrix. Show that the matrix A = LDLT is symmetric.
Problem 8.19
Let L and D be two square matrices each of size n, where L is an orthogonal
matrix, i.e., LLT = LT L = In . Show that if A = LDLT then An = LDn LT ,
where n is a positive integer.
Problem 8.20
Let
1 −2
A=
2 −4
and
2 3
B= .
1 32
Show that AB = 0. Thus, the zero product property of real numbers1 does
not hold for matrices.
Problem 8.21
Let A be a matrix of size m × n, B is of size n × p, and C of size p × q. Show
that A(BC) = (AB)C.
Problem 8.22
Let A be a matrix of size m×n, B and C of size n×p. Show that A(B +C) =
AB + AC.
Problem 8.23
Let A be a matrix of size m×n, B and C of size `×m. Show that (B +C)A =
BA + CA.
Problem 8.24
Let A be a matrix of size m × n and B size n × p. Show that c(AB) =
(cA)B = A(cB).
Problem 8.25
Prove Theorem 8.2.
1
If a, b, ∈ R and ab = 0 then a = 0 or b = 0.
84 MATRICES
Example 9.1
Show that the matrix
−2 1
B= 3
2
− 12
is the inverse of the matrix
1 2
A= .
3 4
Solution.
Using matrix multiplication one checks that AB = BA = I2
Example 9.2
Show that the matrix
1 0
A=
0 0
is singular.
Solution.
Let B = (bij ) be a 2 × 2 matrix. Then
b11 b12 1 0 b11 0
BA = = 6 I2 .
=
b21 b22 0 0 b21 0
Thus, A is singular
9 THE INVERSE OF A SQUARE MATRIX 85
Example 9.3
Let
1 0
1 0 0
A= , B = 0 1 .
0 1 0
0 0
Show that AB = I2 .
Solution.
Simple matrix multiplication shows that AB = I2 . However, this does not
imply that B is the inverse of A since
1 0 0
BA = 0 1 0
0 0 0
Example 9.4
Show that the identity matrix is invertible but the zero matrix is not.
Solution.
Since In In = In , In is nonsingular and its inverse is In . Now, for any n × n
matrix B we have B0 = 0 6= In so that the zero matrix is not invertible
Theorem 9.1
The inverse of a matrix is unique.
Proof.
The proof is by contradiction. Suppose A has more than one inverse. Let
B and C be two distinct inverses of A. We have, B = BIn = B(AC) =
(BA)C = In C = C. A contradiction. Hence, A has a unique inverse
86 MATRICES
Theorem 9.2
If A and B are two square matrices of size n × n such that AB = In then
BA = In and B = A−1 .
For an invertible matrix A one can now define the negative power of a square
matrix as follows: For any positive integer n ≥ 1, we define A−n = (A−1 )n .
The next theorem lists some of the useful facts about inverse matrices.
Theorem 9.3
Let A and B be two square matrices of the same size n × n.
(a) If A and B are invertible matrices then AB is invertible and (AB)−1 =
B −1 A−1 .
(b) If A is invertible then A−1 is invertible and (A−1 )−1 = A.
(c) If A is invertible then AT is invertible and (AT )−1 = (A−1 )T
Proof.
(a) If A and B are invertible then AA−1 = A−1 A = In and BB −1 = B −1 B =
In . In This case, (AB)(B −1 A−1 ) = A[B(B −1 A−1 )] = A[(BB −1 )A−1 ] =
A(In A−1 ) = AA−1 = In . Similarly, (B −1 A−1 )(AB) = In . It follows that
B −1 A−1 is the inverse of AB.
(b) Since A−1 A = AA−1 = In , A is the inverse of A−1 , i.e. (A−1 )−1 = A.
(c) Since AA−1 = A−1 A = In , by taking the transpose of both sides we get
(A−1 )T AT = AT (A−1 )T = In . This shows that AT is invertible with inverse
(A−1 )T
Example 9.5
If A is invertible and k 6= 0 show that (kA)−1 = k1 A−1 .
9 THE INVERSE OF A SQUARE MATRIX 87
Solution.
Suppose that A is invertible and k 6= 0. Then (kA)A−1 = k(AA−1 ) = kIn .
This implies (kA)( k1 A−1 ) = In . Thus, kA is invertible with inverse equals to
1 −1
k
A
Example 9.6
(a) Under what conditions a diagonal matrix is invertible?
(b) Is the sum of two invertible matrices necessarily invertible?
Solution.
(a) Let D = [dii ] be a diagonal n × n matrix. Let B = [bij ] be an n × n
matrix
Pn such that DB = In . Using matrix multiplication, we find (DB)ij =
k=1 ik bkj . It follows that (DB)ij = dii bij = 0 for i 6= j and (DB)ii =
d
dii bii = 1. If dii 6= 0 for all 1 ≤ i ≤ n then bij = 0 for i 6= j and bii = d1ii .
Thus, if d11 d22 h· · · dinn 6= 0 then D is invertible and its inverse is the diagonal
matrix D−1 = d1ii .
(b) The following two matrices are invertible but their sum , which is the
zero matrix, is not.
1 0 −1 0
,
0 −1 0 1
Example 9.7
Consider the 2 × 2 matrix
a b
A= .
c d
Show that if ad − bc 6= 0 then A−1 exists and is given by
−1 1 d −b
A = .
ad − bc −c a
Solution.
Let
x y
B=
z w
be a matrix such that BA = I2 . Then using matrix multiplication we find
ax + cy bx + dy 1 0
= .
az + cw bz + dw 0 1
88 MATRICES
and (
az + cw = 0
bz + dw = 0.
In the first system, using elimination we find (ad − bc)y = −b and (ad −
bc)x = d. Similarly, using the second system we find (ad − bc)z = −c and
(ad − bc)w = a. If ad − bc 6= 0 then one can solve for x, y, z, and w and in
this case B = A−1 as given in the statement of the problem
Finally, we mention here that matrix inverses can be used to solve systems
of linear equations as suggested by the following theorem.
Theorem 9.4
If A is an n × n invertible matrix and b is a column matrix then the equation
Ax = b has a unique solution x = A−1 b.
Proof.
Since A(A−1 b) = (AA−1 )b = In b = b, we find that A−1 b is a solution to the
equation Ax = b. Now, if y is another solution then y = In y = (A−1 A)y =
A−1 (Ay) = A−1 b
9 THE INVERSE OF A SQUARE MATRIX 89
Practice Problems
Problem 9.1
(a) Find two 2 × 2 singular matrices whose sum in nonsingular.
(b) Find two 2 × 2 nonsingular matrices whose sum is singular.
Problem 9.2
Show that the matrix
1 4 0
A = 2 5 0
3 6 0
is singular.
Problem 9.3
Let
1 2
A= .
1 3
Find A−3 .
Problem 9.4
Let
−1 2 −1
A = .
3 5
Find A.
Problem 9.5
Let A and B be square matrices such that AB = 0. Show that if A is
invertible then B is the zero matrix.
Problem 9.6
Find the inverse of the matrix
sin θ cos θ
A= .
− cos θ sin θ
Problem 9.7
Find the matrix A given that
−1 −1 2
(I2 + 2A) = .
4 5
90 MATRICES
Problem 9.8
Find the matrix A given that
T −1 −3 −1
(5A ) = .
5 2
Problem 9.9
Show that if a square matrix A satisfies the equation A2 − 3A − In = 0 then
A−1 = A − 3In .
Problem 9.10
Simplify: (AB)−1 (AC −1 )(D−1 C −1 )−1 D−1 .
Problem 9.11
Let A and B be two square matrices. Show that (ABA−1 )2 = AB 2 A−1 .
Problem 9.12
Two square matrices A and B are said to be similar, denoted A ≈ B, if
there exists an invertible matrix P such that B = P −1 AP.
(a) Show that A ≈ A.
(b) Show that if A ≈ B then B ≈ A.
(c) Show that if A ≈ B and B ≈ C then A ≈ C.
Problem 9.13
0 −1
Let A = . Compare A, AT , and A−1 .
−1 0
Problem 9.14
Suppose A, B, and X are n × n matrices such that A, X, and A − AX are
invertible. Moreover, suppose that
(A − AX)−1 = X −1 B.
Problem 9.15
Suppose that A is similar to a diagonal matrix D. Show that AT is also
similar to D.
9 THE INVERSE OF A SQUARE MATRIX 91
Problem 9.16
Suppose that A is a n × n orthogonal square matrix. What is its inverse?
Problem 9.17
Find the inverse of the matrix
1 0 0 0
0 2 0 0
0 0 3 0
0 0 0 4
Problem 9.18
Solve the system below by finding A−1 .
(
8x1 − x2 = 17
−4x1 + x2 = −5.
Problem 9.19
Solve the system below by finding A−1 .
(
−2x1 + x2 = 1
3x1 − 3x2 = 2.
Problem 9.20
Let A be an n × n invertible square matrix. Show that the homogeneous
system Ax = 0 has only the trivial solution.
92 MATRICES
10 Elementary Matrices
In this section we introduce a special type of invertible matrices, the so-called
elementary matrices, and we discuss some of their properties. As we shall see,
elementary matrices will be used in the next section to develop an algorithm
for finding the inverse of a square matrix.
An n × n elementary matrix is a matrix obtained from the identity matrix
by performing one single elementary row operation.
Example 10.1
Show that the following matrices are elementary matrices
(a)
1 0 0
0 1 0 ,
0 0 1
(b)
1 0 0
0 0 1 ,
0 1 0
(c)
1 0 3
0 1 0 .
0 0 1
Solution.
We list the operations that produce the given elementary matrices.
(a) r1 ← r1 .
(b) r2 ↔ r3 .
(c) r1 ← r1 + 3r3
Example 10.2
Consider the matrix
1 0 2 3
A = 2 −1 3 6 .
1 4 4 0
(a) Find the row equivalent matrix to A obtained by adding 3 times the first
row of A to the third row. Call the equivalent matrix B.
10 ELEMENTARY MATRICES 93
The conclusion of the above example holds for any matrix of size m × n.
Theorem 10.1
If the elementary matrix E results from performing a certain row operation
on Im and if A is an m × n matrix, then the product of EA is the matrix
that results when this same row operation is performed on A.
It follows from the above theorem that a matrix A is row equivalent to
a matrix B, denoted by A ∼ B, if and only if B = Ek Ek−1 · · · E1 A, where
E1 , E2 , · · · , Ek are elementary matrices. See Problem 10.16.
The above theorem is primarily of theoretical interest and will be used for
developping some results about matrices and systems of linear equations.
From a computational point of view, it is preferred to perform row operations
directly rather than multiply on the left by an elementary matrix. Also, this
theorem says that an elementary row operation on A can be achieved by
premultiplying A by the corresponding elementary matrix E.
Given any elementary row operation, there is another row operation ( called
its inverse) that reverse the effect of the first operation. The inverses are
described in the following chart.
Theorem 10.2
Every elementary matrix is invertible, and the inverse is an elementary ma-
trix.
Proof.
Let A be any n × n matrix. Let E be an elementary matrix obtained by
applying a row elementary operation ρ on In . By Theorem 10.1, applying ρ
on A produces a matrix EA. Applying the inverse operation ρ−1 to EA gives
F (EA) where F is the elementary matrix obtained from In by applying the
operation ρ−1 . Since inverse row operations cancel the effect of each other, it
follows that F EA = A. Since A was arbitrary, we can choose A = In . Hence,
F E = In . A similar argument shows that EF = In . Hence E is invertible
and E −1 = F
Example 10.3
Write down the inverses of the following elementary matrices:
0 1 0 1 0 0 1 0 5
(a)E1 = 1 0 0 , (b)E2 = 0
1 0 , (c)E3 = 0 1 0 .
0 0 1 0 0 9 0 0 1
Solution.
Using the inverses of elementary row operations introduced above, we find
(a) E1−1 = E1 since E1 E1 = I3 . Note that E1 = E1T .
(b)
1 0 0
E2−1 = 0 1 0 .
0 0 91
Note that E2T = E2 . (c)
1 0 −5
E3−1 = 0 1 0
0 0 1
Example 10.4
If E is an elementary matrix show that E T is also an elementary matrix of
the same type.
10 ELEMENTARY MATRICES 95
Solution.
(a) Suppose that E is the elementary matrix obtained by swapping rows i
and j of In with i < j. But then E T is obtained by interchanging rows i and
j of In and so is an elementary matrix. Note that E T = E.
(b) Suppose E is obtained by multiplying the ith row of In by a non-zero
constant k. But then E T is obtained by multiplying the ith row of In by k
and so is an elementary matrix. Note that E T = E.
(c) Suppose E is obtained by adding k times the ith row of In to the j th row
(i < j). But then E T is obtained by adding k times the j th row of In to the
ith row. Hence, E T is an elementary matrix
96 MATRICES
Practice Problems
Problem 10.1
Which of the following are elementary matrices?
(a)
1 1 0
0 0 1 .
0 1 0
(b)
1 0
.
−5 1
(c)
1 0 0
0 1 9 .
0 0 1
(d)
2 0 0 2
0 1 0 0
.
0 0 1 0
0 0 0 1
Problem 10.2
Let A be a 4 × 3 matrix. Find the elementary matrix E, which as a premulti-
plier of A, that is, as EA, performs the following elementary row operations
on A :
(a) Multiplies the second row of A by −2.
(b) Adds 3 times the third row of A to the fourth row of A.
(c) Interchanges the first and third rows of A.
Problem 10.3
For each of the following elementary matrices, describe the corresponding
elementary row operation and write the inverse.
(a)
0 0 1
E = 0 1 0 .
1 0 0
10 ELEMENTARY MATRICES 97
(b)
1 0 0
E = −2 1 0 .
0 0 1
(c)
1 0 0
E = 0 1 0 .
0 0 5
Problem 10.4
Consider the matrices
3 4 1 8 1 5 3 4 1
A = 2 −7 −1 , B = 2 −7 −1 , C = 2 −7 −1 .
8 1 5 3 4 1 2 −7 3
Problem 10.5
What should we premultiply a 3 × 3 matrix if we want to interchange rows
1 and 3?
Problem 10.6
Let
1 0 0 1 1 0 1 0 0
E1 = 0 2 0 , E2 = 0 1 0 , E3 = 0 0 1 .
0 0 1 0 0 1 0 1 0
Problem 10.7
List all 3 × 3 elementary matrices corresponding to type I elementary row
operations.
Problem 10.8
List all 3 × 3 elementary matrices corresponding to type II elementary row
operations.
98 MATRICES
Problem 10.9
Write down the inverses of the following elementary matrices:
0 0 1 3 0 0 1 0 0
E1 = 0 1 0 , E2 = 0 1
0 , E3 = −2 1 0 .
1 0 0 0 0 1 0 0 1
Problem 10.10
Consider the following elementary matrices:
0 0 1 3 0 0 1 0 0
E1 = 0 1
0 , E2 = 0 1 0 , E3 = −2 1 0 .
1 0 0 0 0 1 0 0 1
Find
1 0 2
E1 E2 E3 −2 3 4 .
0 5 −3
Problem 10.11
Write the following matrix as a product of elementary matrices:
1 0 −1
0 1 1 .
0 0 −1
Problem 10.12
Explain why the following matrix cannot be written as a product of elemen-
tary matrices.
1 1 1
A= 0 1 1 .
3 3 3
Problem 10.13
Let E1 , E2 , E3 be 3 × 3 elementary matrices that act on an 3 × m matrix A,
such that
(i) E1 A is obtained from A by multiplying the third row by −1;
(ii) E2 A is obtained from A by subtracting the third row from the second;
(iii) E3 A is obtained from A by adding the third row to the first.
(a) Find E1 , E2 , E3 .
(b) Find (E3 E2 E1 )−1 .
10 ELEMENTARY MATRICES 99
Problem 10.14
Let
1 0 −2
A = 0 4 3
0 0 1
(a) Find elementary matrices E1 , E2 , E3 such that E3 E2 E1 A = I3 .
(b) Write A as a product of elementary matrices.
Problem 10.15
Let A be a matrix with 4 rows. For each elementary row operation on A
below, find the corresponding elementary matrices and their inverses.
(a) Adding 2 times row 3 to row 4;
(b) Swapping rows 1 and 4;
(c) Multiplying row 2 by 6.
Problem 10.16
We say that a matrix A is row equivalent to a matrix B if and only if
B = Ek Ek−1 · · · E1 A, where E1 , E2 , · · · , Ek are elementary matrices. We
write A ∼ B.
(a) Show that A ∼ A for any matrix A.
(b) Show that if A ∼ B then B ∼ A.
(c) Show that is A ∼ B and B ∼ C then A ∼ C.
Problem 10.17
Let A be an n × n matrix. Show that if A ∼ In then A is the product of
elementary matrices.
Problem 10.18
Show that if a square matrix A can be written as the product of elementary
matrices then A is invertible.
Problem 10.19
Let A be an n × n matrix such that A = E1 E2 · · · Ek where each Ei swaps
some two rows of In . Show that AAT = In .
Problem 10.20
Let
3 4 5
A= .
2 −1 0
100 MATRICES
Theorem 11.1
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) A is row equivalent to In .
(d) rank(A) = n.
Proof.
(a) ⇒ (b) : Suppose that A is invertible and x0 is a solution to Ax = 0. Then
Ax0 = 0. Multiply both sides of this equation by A−1 to obtain A−1 Ax0 =
A−1 0, that is, x0 = 0. Hence, the trivial solution is the only solution.
(b) ⇒ (c) : Suppose that Ax = 0 has only the trivial solution. Then the
reduced row-echelon form of the augmented matrix has no rows of zeros or
free variables. Hence it must look like
1 0 0 ... 0 0
0 1 0 . . . 0 0
0 0 1 . . . 0 0
.
.. .. .. .. ..
. . . . .
0 0 0 ··· 1 0
If we disgard the last column of the previous matrix we can conclude that A
can be reduced to In by a sequence of elementary row operations, i.e., A is
row equivalent to In .
(c) ⇒ (d) : Suppose that A is row equivalent to In . Then rank(A) =
102 MATRICES
rank(In ) = n.
Step 2: r3 ← r3 + 2r2
1 2 3 1 0 0
0 1 −3 −2 1 0
0 0 −1 −5 2 1
Step 3: r1 ← r1 − 2r2
1 0 9 5 −2 0
0 1 −3 −2 1 0
0 0 −1 −5 2 1
Step 5: r3 ← −r3
1 0 0 −40 16 9
0 1 0 13 −5 −3
0 0 1 5 −2 −1
It follows that
−40 16 9
A−1 = 13 −5 −3
5 −2 −1
Example 11.2
Show that the following homogeneous system has only the trivial solution.
x1 + 2x2 + 3x3 = 0
2x1 + 5x2 + 3x3 = 0
x1 + 8x3 = 0.
Solution.
The coefficient matrix of the given system is invertible by the previous ex-
ample. Thus, by Theorem 11.1 the system has only the trivial solution
How can we tell when a square matrix A is singular? i.e., when does the
algorithm of finding A−1 fail? The answer is provided by the following theo-
rem
104 MATRICES
Theorem 11.2
An n × n matrix A is singular if and only if A is row equivalent to a matrix
B that has a row of zeros.
Proof.
Suppose first that A is singular. Then by Theorem 11.1, A is not row equiv-
alent to In . Thus, A is row equivalent to a matrix B 6= In which is in reduced
row echelon form. By Theorem 11.1, B must have a row of zeros.
Conversely, suppose that A is row equivalent to matrix B with a row consist-
ing entirely of zeros. Then rank(B) < n so that B is singular by Theorem
11.1. Now, B = Ek Ek−1 . . . E2 E1 A. If A is nonsingular then B is nonsingular,
a contradiction. Thus, A must be singular
Corollary 11.1
If A is a square matrix with a row consisting entirely of zeros then A is
singular.
Example 11.3
Solve the following system.
x1 + 2x2 + 3x3 = 5
2x1 + 5x2 + 3x3 = 3
x1 + 8x3 = 17.
Solution.
The coefficient matrix is invertible by Example 11.1. Hence, the unique
solution is given by x = A−1 b. That is,
x1 −40 16 9 5
x2 = 13 −5 −3 3
x3 5 −2 −1 17
1
= −1 .
2
Hence, x1 = 1, x2 = −1, and x3 = 2
11 FINDING A−1 USING ELEMENTARY MATRICES 105
Practice Problems
Problem 11.1
Determine if the following matrix is invertible.
1 6 4
2 4 −1 .
−1 2 5
Problem 11.2
For what values of a does the following homogeneous system have a nontrivial
solution? (
(a − 1)x1 + 2x2 = 0
2x1 + (a − 1)x2 = 0.
Problem 11.3
Find the inverse of the matrix
1 1 1
0 2 3 .
5 5 1
Problem 11.4
Prove that if A is symmetric and non-singular then A−1 is symmetric.
Problem 11.5
If
4 0 0
D = 0 −2 0 ,
0 0 3
find D−1 .
Problem 11.6
Prove that a square matrix A is non-singular if and only if A is a product of
elementary matrices.
Problem 11.7
Prove that two m × n matrices A and B are row equivalent if and only if
there exists a non-singular matrix P such that B = P A.
106 MATRICES
Problem 11.8
Let A and B be two n×n matrices. Suppose A is row equivalent to B. Prove
that A is non-singular if and only if B is non-singular.
Problem 11.9
Show that a 2 × 2 lower triangular matrix is invertible if and only if the
product of the diagonal entries is not zero and in this case the inverse is also
lower triangular.
Problem 11.10
Let A be an n × n matrix and suppose that the system Ax = 0 has only
the trivial solution. Show that Ak x = 0 has only the trivial solution for any
positive integer k.
Problem 11.11
Show that if A and B are two n × n invertible matrices then A is row equiv-
alent to B.
Problem 11.12
(a) Find the inverse of the matrix
1 −2 1
A= 0 2 −8 .
−4 5 9
(b) Solve the following system.
x1 − 2x2 + x3 = 0
2x2 − 8x3 = 8
−4x1 + 5x2 + 9x3 = −9.
Problem 11.13
(a) Find the inverse of the matrix
5 −5 −15
A = 4 −2 −6 .
3 −6 −17
(b) Solve the following system.
5x1 − 5x2 − 15x3 = 40
4x1 − 2x2 − 6x3 = 19
3x1 − 6x2 − 17x3 = 41.
11 FINDING A−1 USING ELEMENTARY MATRICES 107
Problem 11.14
Let A and B be two square matrices. Show that AB is non-singular if and
only if both A and B are non-singular.
Problem 11.15
If P is an n×n matrix such that P T P = In then the matrix H = In −2P P T is
called the Householder matrix. Show that H is symmetric and H T H = In .
Problem 11.16
Find the inverse of the matrix
1 0 4
A = −1 1 −1 .
−1 0 −3
Problem 11.17
Find all 2 × 2 matrices
a 1
1 b
such that A−1 = A.
Problem 11.18
Find the inverse of the matrix
3 5 7
A = 1 2 3 .
2 3 5
Problem 11.19
Find a 3 × 3 matrix A such that
1 3 0
(AT + 5I3 )−1 = 0 −1 0 .
2 2 1
Problem 11.20
Show that if A and B are two n × n matrices such that AB = In then
B −1 = A and BA = In .
108 MATRICES
Determinants
With each square matrix we can associate a real number called the determi-
nant of the matrix. Determinants have important applications to the theory
of systems of linear equations. More specifically, determinants give us a
method (called Cramer’s method) for solving linear systems. Also, determi-
nant tells us whether or not a matrix is invertible.
Throughout this chapter we use only square matrices.
109
110 DETERMINANTS
is the number
a11 a12
|A| = = a11 a22 − a21 a12 .
a21 a22
The determinant of a 3 × 3 matrix can be found using the determinants of
2 × 2 matrices and a cofactor expansion which we discuss next.
If A is a square matrix of order n then the minor of the entry aij , denoted
by Mij , is the determinant of the submatrix obtained from A by deleting the
ith row and the j th column. The cofactor of the entry aij is the number
Cij = (−1)i+j Mij .
Example 12.1
Let
3 1 −4
A = 2 5 6 .
1 4 8
Find the minor and the cofactor of the entry a32 = 4.
Solution.
The minor of the entry a32 is
3 −4
M32 = = 26
2 6
Example 12.2
Find the cofactors C11 , C12 , and C13 of the matrix
a11 a12 a13
a21 a22 a23 .
a31 a32 a33
12 DETERMINANTS BY COFACTOR EXPANSION 111
Solution.
We have
1+1
a22 a23
C11 = (−1) = a22 a33 − a32 a23
a32 a33
1+2 a21 a23
C12 = (−1) = −(a21 a33 − a31 a23 )
a31 a33
1+3 a21 a22
C13 = (−1) = a21 a32 − a31 a22
a31 a32
The determinant of a matrix A of order n can obtained by multiplying
the entries of a row (or a column) by the corresponding cofactors and adding
the resulting products. More precisely, the expansion along row i is
|A| = ai1 Ci1 + ai2 Ci2 + · · · + ain Cin .
The expansion along column j is given by
|A| = a1j C1j + a2j C2j + · · · + anj Cnj .
Any row or column chosen will result in the same answer.
Example 12.3
Find the determinant of the matrix
a11 a12 a13
a21 a22 a23 .
a31 a32 a33
Solution.
Using the previous example, we can find the determinant using the cofactor
along the first row to obtain
|A| =a11 C11 + a12 C12 + a13 C13
a22 a23 a21 a23 a21 a22
=a11 − a12
a31 a33 + a13 a31 a32
a32 a33
=a11 (a22 a33 − a32 a23 ) − a12 (a21 a33 − a31 a23 ) + a13 (a21 a32 − a31 a22 )
Remark 12.1
In general, the best strategy for evaluating a determinant by cofactor ex-
pansion is to expand along a row or a column having the largest number of
zeros.
112 DETERMINANTS
Example 12.4
If a square matrix has a row of zeros then its determinant is zero.
Solution.
Finding the determinant of the matrix by cofactor expansion along the row
of zeros, we find that the determinant is 0
A square matrix is called lower triangular if all the entries above the main
diagonal are zero. A square matrix is called upper triangular if all the
entries below the main diagonal are zero. A triangular matrix is one that
is either lower triangular or upper triangular. A matrix that is both upper
and lower triangular is called a diagonal matrix.
Example 12.5
Find the determinant of each of the following matrices.
(a)
0 0 a13
A = 0 a22 a23 .
a31 a32 a33
(b)
0 0 0 a14
0 0 a23 a24
B=
0 a32 a33
.
a34
a41 a42 a43 a44
(c)
a11 0 0 0
a21 a22 0 0
C=
a31 a32 a33
.
0
a41 a42 a43 a44
Solution.
(a) Expanding along the first column we find
Remark 12.2
Be aware that the matrices in (a) and (b) do not fall into the category of
triangular matrices whereas (c) does.
Example 12.6
Evaluate the determinant of the following matrix.
2
7 −3 8 3
0 −3 7 5 1
0 0 6 7 6
0 0 0 9 8
0 0 0 0 4
Solution.
The given matrix is upper triangular so that the determinant is the product
of entries on the main diagonal, i.e., equals to −1296
Example 12.7
Use cofactor expansion along the first column to find |A| where
3 5 −2 6
1 2 −1 1
A= 2
.
4 1 5
3 7 5 3
Solution.
Expanding along the first column we find
Practice Problems
Problem 12.1
Evaluate the determinant of each of the following matrices
(a)
3 5
A= .
−2 4
(b)
−2 7 6
A = 5 1 −2 .
3 8 4
Problem 12.2
Find all values of t for which the determinant of the following matrix is zero.
t−4 0 0
A= 0 t 0 .
0 3 t−1
Problem 12.3
Solve for x
1 0 −3
x 1
=
1 1 − x 2 x −6 .
1 3 x − 5
Problem 12.4
Evaluate the determinant of the following matrix
1 2 3
A = 4 5 6 .
0 0 0
Problem 12.5
Let
4 −1 1 6
0 0 −3 3
A=
4
.
1 0 14
4 1 3 2
Find M23 and C23 .
12 DETERMINANTS BY COFACTOR EXPANSION 115
Problem 12.6
Find all values of λ for which |A| = 0, where
λ−1 0
A= .
2 λ+1
Problem 12.7
Evaluate the determinant of the matrix
3 0 0
A = 2 −1 5 ,
1 9 −4
Problem 12.8
Evaluate the determinant of the matrix by a cofactor expansion along a row
or column of your choice.
3 3 0 5
2 2 0 −2
A= 4 1 −3
.
0
2 10 3 2
Problem 12.9
Evaluate the determinant of the following matrix by inspection.
1 2 7 −3
0 1 −4 1
A= 0 0
.
2 7
0 0 0 3
Problem 12.10
Evaluate the determinant of the following matrix.
sin θ cos θ 0
A = − cos θ sin θ 0 .
sin θ − cos θ sin θ + cos θ 1
116 DETERMINANTS
Problem 12.11
Find all values of λ such that |A| = 0.
(a)
λ − 1 −2
A= ,
1 λ−4
(b)
λ−6 0 0
A= 0 λ −1 .
0 4 λ−4
Problem 12.12
Find
3
0 0 −2 4
0 2 0 0 0
0 −1 0 5 −3 .
−4 0 1 0 6
0 −1 0 3 2
Problem 12.13
Solve the equation
1 1 1
1 2 t = 0.
1 4 t2
Problem 12.14
Find
1 2 3
|A| = 0 −1 4 .
0 0 π
Problem 12.15
Find
3
0 2 −1
1 2 0 −2
.
4
0 6 −3
5 0 2 0
12 DETERMINANTS BY COFACTOR EXPANSION 117
Problem 12.16
Find
a b c
|A| = ka kb kc .
e f g
Problem 12.17
Solve for λ.
λ − 5 −8 −16
−4 λ − 1 −8 = 0.
4 4 λ + 11
Problem 12.18
Solve for λ.
λ − 2 −1 −1
−2 λ − 1 2 = 0.
1 0 λ + 2
Problem 12.19
Solve for λ.
λ + 2
0 −1
6 λ+2 0 = 0.
−19 −5 λ + 4
Problem 12.20
Solve for λ.
λ − 2 −1 0
A = 0 λ−2 0 = 0.
−2 −3 λ − 1
118 DETERMINANTS
Theorem 13.1
If A is an n × n triangular matrix then |A| = a11 a22 . . . ann .
Example 13.1
Compute |A|.
(a)
1 2 3
A = 0 4 5 .
0 0 6
(b)
1 0 0
B = 2 3 0 .
4 5 6
Solution.
(a) Since A is upper triangular and hence triangular, |A| = (1)(4)(6) = 24.
(b) |B| = (1)(3)(6) = 18
Example 13.2
Compute the determinant of the identity matrix In .
Solution.
Since the identity matrix is triangular with entries equal to 1 on the main
diagonal, |In | = 1
Theorem 13.2
If a square matrix has two identical rows then its determinant is zero.
13 EVALUATING DETERMINANTS BY ROW REDUCTION 119
Proof.
We will prove this result by induction of the dimension of the matrix.
(i) Basis of induction: Suppose A = [aij ] is a 2 × 2 matrix with two indentical
rows. That is,
a b
.
a b
Then |A| = ab − ab = 0.
(ii) Induction hypothesis: Suppose the result holds for all numbers 1, 2, · · · , n.
(iii) Induction step: Let A be an (n + 1) × (n + 1) matrix with two identical
rows i and k with i < k. Choose a row m that is different from i and k.
Finding the determinant of A by cofactor expansion along the mth row, we
find n
X
|A| = am` (−1)m+` Mm` .
`=1
Theorem 13.3
Let A be an n × n matrix.
(a) Let B be the matrix obtained from A by multiplying a row of A by a
scalar c. Then |B| = c|A| or |A| = 1c |B|.
(b) Let B be the matrix obtained from A by adding c times a row of A to
another row. Then |B| = |A|.
(c) Let B be the matrix obtained from A by interchanging two rows of A.
Then |B| = −|A|.
Example 13.3
Use Theorem 13.3 to evaluate the determinant of the following matrix
0 1 5
A = 3 −6 9 .
2 6 1
120 DETERMINANTS
Solution.
We use Gaussian elimination as follows.
Step 1: r1 ↔ r2
3 −6 9
|A| = − 0 1 5 .
2 6 1
Step 2: r1 ← r1 − r3
1 −12 8
|A| = − 0 1 5 .
2 6 1
Step 3: r3 ← r3 − 2r1
1 −12 8
|A| = − 0 1 5 .
0 30 −15
Step 4: r3 ← r3 − 30r2
1 −12 8
|A| = − 0 1 5 = −(1)(1)(−165) = 165
0 0 −165
Example 13.4
Show that if a square matrix has two proportional rows then its determinant
is zero.
Solution.
Suppose that A is a square matrix such that row j is k times row i with k 6= 0.
By adding − k1 rj to ri then the ith row will consist of zeros. By Thereom 13.2
and Theorem 13.3(b), |A| = 0
Example 13.5
Find, by inspection, the determinant of the following matrix.
3 −1 4 −2
6 −2 5 2
A= 5
.
8 1 4
−9 3 −12 6
13 EVALUATING DETERMINANTS BY ROW REDUCTION 121
Solution.
Since the first and the fourth rows are proportional, the determinant is zero
by Example 13.4
Example 13.6
Show that if A is an n × n matrix and c is a scalar then |cA| = cn |A|.
Solution.
The matrix cA is obtained from the matrix A by multiplying the rows of A
by c 6= 0. By mutliplying the first row of A by c we obtain a matrix B1 such
that 1c |B1 | = |A| (Theorem 13.3(a)). Now, by multiplying the second row of
B1 by c we obtain a matrix B2 such that |B1 | = 1c |B2 |. Thus, |A| = c12 |B2 |.
Repeating this process, we find |A| = c1n Bn = c1n |cA| or |cA| = cn |A|
Example 13.7
(a) Let E1 be the elementary matrix corresponding to type I elementary row
operation, i.e., multiplying a row by a scalar. Find |E1 |.
(b) Let E2 be the elementary matrix corresponding to type II elementary
row operation, i.e., adding a multiple of a row to another row. Find |E2 |.
(c) Let E3 be the elementary matrix corresponding to type III elementary
row operation, i.e., swapping two rows. Find |E3 |.
Solution.
(a) The matrix E1 is obtained from the identity matrix by multiplying a row
of In by a nonzero scalar c. By Theorem 13.3(a), we have |E1 | = c|In | = c.
(b) E2 is obtained from In by adding a multiple of a row to another row. By
Theorem 13.3(b), we have |E2 | = |In | = 1.
(c) The matrix E3 is obtained from the matrix In by interchanging two rows.
By Theorem 13.3(c), we have |E3 | = −|In | = −1
122 DETERMINANTS
Practice Problems
Problem 13.1
Use the row reduction technique to find the determinant of the following
matrix.
2 5 −3 −2
−2 −3 2 −5
A= .
1 3 −2 2
−1 −6 4 3
Problem 13.2
Given that
a b c
d e f = −6,
g h i
find
(a)
d e f
g h i ,
a b c
(b)
3a 3b 3c
−d −e −f ,
4g 4h 4i
(c)
a + g b + h c + i
d e f ,
g h i
(d)
−3a −3b −3c
d e f .
g − 4d h − 4e i − 4f
13 EVALUATING DETERMINANTS BY ROW REDUCTION 123
Problem 13.3
Determine by inspection the determinant of the following matrix.
1 2 3 4 5
6 7 8 9 10
11 12 13 14 15
.
16 17 18 19 20
2 4 6 8 10
Problem 13.4
Let A be a 3 × 3 matrix such that |2A| = 6. Find |A|.
Problem 13.5
Find the determinant of the following elementary matrix by inspection.
1 0 0 0
0 1 0 0
0 0 −5 0 .
0 0 0 1
Problem 13.6
Find the determinant of the following elementary matrix by inspection.
1 0 0 0
0 0 1 0
.
0 1 0 0
0 0 0 1
Problem 13.7
Find the determinant of the following elementary matrix by inspection.
1 0 0 0
0 1 0 −9
.
0 0 1 0
0 0 0 1
Problem 13.8
Use the row reduction technique to find the determinant of the following
matrix.
2 1 3 1
1 0 1 1
A= 0
.
2 1 0
0 1 2 3
124 DETERMINANTS
Problem 13.9
Use row reduction to find the determinant of the following Vandermonde
matrix.
1 1 1
A = a b c.
a2 b 2 c 2
Problem 13.10
Let a, b, c be three numbers such that a + b + c = 0. Find the determinant of
the following matrix.
b+c a+c a+b
A= a b c .
1 1 1
Problem 13.11
Compute the following determinant using elementary row operations.
1 −4 2
−2 8 −9 .
−1 7 0
Problem 13.12
Compute the following determinant using elementary row operations.
2 −8 6 8
3 −9 5 10
.
−3 0 1 −2
1 −4 0 6
Problem 13.13
If
1 1 1
A = 0 1 1
1 0 1
and c 6= 0, then calculate |cA|.
Problem 13.14
Use row reduction to compute the determinant of
2 3 3 1
0 4 3 −3
A= 2 −1 −1 −3 .
0 −4 −3 2
13 EVALUATING DETERMINANTS BY ROW REDUCTION 125
Problem 13.15
Solve for x,
1 0 0 0
2 3 0 0
4 = 54.
5 x 0
6 7 8 9
Problem 13.16
Let A be an n × n matrix and E an elementary matrix. Show that |EA| =
|E||A|.
Problem 13.17
By induction on k, one can show that the previous result is true when A is
premultiplied by any number of elementary matrice. Recall that if a square
matrix A is invertible then A can be written as the product of elementary
matrices (Theorem 11.1). Use this theorem and the previous exercise to show
that if A is invertible then |A| =
6 0. Taking the contrapositive, if |A| = 0 then
A is singular.
Problem 13.18
Using determinants, show that the matrix
1 0 1 1
−1 5 −7 4
A= −3 5 −9
2
4 4 6 −1
is singular.
Problem 13.19
Find the determinant of
−1 0 0 0 0
0 3 0 0 0
0
A= 0 2 0 0
.
0 0 0 4 0
0 0 0 0 −2
Problem 13.20
Let A = [aij ] be a triangular matrix of size n × n. Use induction on n ≥ 2,
to show that |A| = a11 a22 · · · ann .
126 DETERMINANTS
Problem 13.21
Let A be an n × n matrix. Let B be the matrix obtained from A by multi-
plying a row of A by a scalar c. Show that |B| = c|A| or |A| = 1c |B|.
Problem 13.22
Let A = [aij ], B = [bij ], and C = [cij ] be n × n matrices such that akj =
bkj + ckj for all 1 ≤ j ≤ n and bij = cij = aij for all i 6= k. That is, the k th
row of A is the sum of the k th row of B and the k th row of C and all the
rows of B and C different from row k are the same as the rows of A. Show
that |A| = |B| + |C|.
Problem 13.23
Let A be an n × n matrix. Let B be the matrix obtained from A by adding
c times a row of A to another row of A. Show that |B| = |A|.
Problem 13.24
Let A be an n × n matrix. Let B be the matrix obtained from A by inter-
changing rows i and k of A with i < k. Show that |B| = −|A|.
14 ADDITIONAL PROPERTIES OF THE DETERMINANT 127
Theorem 14.3
If A is an n × n matrix then |AT | = |A|.
Proof.
(i) Suppose first that A is singular. Then |A| = 0. By Theorem 9.3(c), AT is
singular and by Theorem 14.1, |AT | = 0.
(ii) Suppose that A is non-singular. Then A can be written as the product of
elementary matrices A = E1 E2 · · · Ek . By Example 10.4 and Theorem 13.3,
|EjT | = |Ej for 1 ≤ j ≤ k. Hence, |AT | = |EkT · E2T E1T | = |EkT | · · · |E2T ||E1T | =
|Ek | · · · |E2 ||E1 | = |E1 E2 · · · Ek | = |A|
It is worth noting here that the above Theorem says that every property
about determinants that contains the word “row” in its statement is also
true when the word ‘column” is substituted for ‘row”.
Example 14.2
(a) Show that a square matrix with two identical columns has a zero deter-
minant.
(b) Show that a square matrix with a column of zeros has a zero determinant.
(c) Show that if a square matrix has two proportional columns then its de-
terminant is zero.
Solution.
(a) If A has two identical columns then AT has two identical rows so that by
Theorem 13.2, we have |AT | = 0. By Theorem 14.3, |A| = 0.
(b) If A has acolmun of zeros then AT has a row of zeros so that by Example
12.4, we have |AT | = 0. By Theorem 14.3, |A| = 0.
(c) If A has two proportional columns then AT has two proportional rows so
that by Example 13.4, we have |AT | = 0. By Theorem 14.3, |A| = 0
Our next major result in this section concerns the determinant of a product
of matrices.
Theorem 14.4
If A and B are n × n matrices then |AB| = |A||B|.
Example 14.3
Is it true that |A + B| = |A| + |B|?
14 ADDITIONAL PROPERTIES OF THE DETERMINANT 129
Solution.
No. Consider the following matrices.
1 0 −1 0
A= ,B = .
0 −1 0 1
Example 14.4
Show that if A is invertible then |A−1 | = 1
|A|
.
Solution.
If A is invertible then A−1 A = In . Taking the determinant of both sides
we find |A−1 ||A| = 1. That is, |A−1 | = |A|
1
. Note that, since A is invertible,
|A| =
6 0
Example 14.5
Let A and B be two similar square matrices , i.e. there exists a nonsingular
matrix P such that A = P −1 BP. Show that |A| = |B|.
Solution.
Using Theorem 14.4 and Example 14.4 we have, |A| = |P −1 BP | = |P −1 ||B||P | =
1
|P |
|B||P | = |B|. Note that since P is nonsingular, |P | =
6 0
Example 14.6
Find all values of x such that the matrix
2 0 10
A = 0 7 + x −3
0 4 x
is invertible.
Solution.
Expanding along the first column, we find |A| = 2[x(7 + x) + 12] = 2(x2 +
7x + 12). Now, A is invertible if and only if 2(x2 + 7x + 12) 6= 0. This is
equivalent to x2 + 7x + 12 6= 0. Factoring the quadratic expression, we find
x2 + 7x + 12 = (x + 3)(x + 4). Hence, A is invertible if and only if x 6= −3
and x 6= −4
130 DETERMINANTS
Practice Problems
Problem 14.1
Show that if n is any positive integer then |An | = |A|n .
Problem 14.2
Show that if A is an n × n skew-symmetric and n is odd then |A| = 0.
Problem 14.3
Show that if A is orthogonal, i.e. AT A = AAT = In then |A| = ±1. Note
that A−1 = AT .
Problem 14.4
If A is a non-singular matrix such that A2 = A, what is |A|?
Problem 14.5
Find out, without solving the system, whether the following system has a
non-trivial solution
x1 − 2x2 + x3 = 0
2x1 + 3x2 + x3 = 0
3x1 + x2 + 2x3 = 0.
Problem 14.6
For which values of c does the matrix
1 0 −c
A = −1 3 1
0 2c −4
have an inverse?
Problem 14.7
If |A| = 2 and |B| = 5, calculate |A3 B −1 AT B 2 |.
Problem 14.8
Show that |AB| = |BA|.
Problem 14.9
Show that |A + B T | = |AT + B| for any n × n matrices A and B.
14 ADDITIONAL PROPERTIES OF THE DETERMINANT 131
Problem 14.10
Let A = [aij ] be a triangular matrix. Show that |A| =
6 0 if and only if aii 6= 0,
for 1 ≤ i ≤ n.
Problem 14.11
Using determinant, show that the rank of the matrix
3 0 2 −1
1 2 0 −2
A=
4 0 6 −3
5 0 2 0
is 4.
Problem 14.12
Show, by inspection, that the matrix
1 2 3
A = 4 8 9
10 20 21
is singular.
Problem 14.13
Let A be a 3 × 3 matrix with |A| = 5. Find each of the following
(a) |A4 | (b) |AT | (c) |5A| (d) |A−1 |.
Problem 14.14
Let A be a 3 × 3 matrix such that |A| = 3. Find |(3A)−1 |.
Problem 14.15
Show that the matrix 2
2x + 3 x2 1
4x 2x 0
4 2 0
is singular for all values of x.
Problem 14.16
Using determinants, show that the rank of the following matrix is less than
4.
1 4 2 3
0 1 4 4
A= −1 0 1 0
2 0 4 1
132 DETERMINANTS
Problem 14.17
Let A be a 3 × 3 invertible matrix such that | − 8A2 | = |A−1 |. Find the value
of |A|.
Problem 14.18
Show that the matrix
3 2 −2 0
5 −6 −1 0
−6 0 3 0
4 7 0 −3
is singular.
Problem 14.19
Let
2 4 2 1
4 3 0 −1
A=
−6
.
0 2 0
0 1 1 2
Find |AT |.
Problem 14.20
Show that |AB| = |A||B| if A is singular.
Problem 14.21
Suppose that A and B are two n×n matrices such that A is invertible. Using
this fact and Problem 13.16, show that |AB| = |A||B|.
15 FINDING A−1 USING COFACTORS 133
Proof.
Let B be the matrix obtained by replacing the j th row of A by the ith row of
A. Then B has two identical rows and therefore |B| = 0(See Theorem 13.2
(a)). Expand |B| along the j th row. The elements of the j th row of B are
ai1 , ai2 , . . . , ain . The cofactors are Cj1 , Cj2 , . . . , Cjn . Thus
The following theorem states that the product A · adj(A) is a scalar mul-
tiple of the identity matrix.
Theorem 15.2
If A is an n × n matrix then A · adj(A) = |A|In .
Proof.
The (i, j) entry of the matrix
a11 a12 . . . a1n C11 C21 . . . Cn1
a21 a22 . . . a2n C12 C22 . . . Cn2
A · adj(A) = ..
.. .. .. .. ..
. . . . . .
an1 an2 . . . ann C1n C2n . . . Cnn
if i = j and 0 if i 6= j. Hence,
|A| 0 ... 0
0 |A| . . . 0
A · adj(A) = .. .. = |A|In .
..
. . .
0 0 ... |A|
The following theorem provides a way for finding the inverse of a matrix
using the notion of the adjoint.
15 FINDING A−1 USING COFACTORS 135
Theorem 15.3
adj(A)
6 0 then A is invertible and A−1 =
If |A| = |A|
. Hence, adj(A) = A−1 |A|.
Proof.
By the previous
theorem we have that A(adj(A)) = |A|In . If |A| =6 0 then
adj(A)
A |A| = In . By Theorem 9.2, A is invertible with inverse A = adj(A)
−1
|A|
Example 15.2
Let
3 2 −1
A = 1 6 3 .
2 −4 0
Use Theorem 15.3 to find A−1 .
Solution.
First we find the determinant of A given by |A| = 64. By Theorem 15.3 and
Example 15.1, we find
3 1 3
1 16 16 16
A−1 = 3
adj(A) = 32 1
32
5
− 32
|A|
− 41 14 1
4
Theorem 15.4
Let A and B denote invertible n × n matrices. Then,
(a) adj(A−1 ) = (adj(A))−1 .
(b) adj(AT ) = (adj(A))T .
(c) adj(AB) = adj(B)adj(A).
Proof.
(a) Since A(adj(A)) = |A|In , adj(A) is invertible ( Theorem 9.2) and (adj(A))−1 =
A
|A|
= (A−1 )−1 |A−1 | = adj(A−1 ).
(b) adj(AT ) = (AT )−1 |AT | = (A−1 )T |A| = [A−1 |A|]T = (adj(A))T .
Example 15.3
Show that if A is singular then A · adj(A) = 0, the zero matrix.
Solution.
If A is singular then |A| = 0. But then A · adj(A) = |A|In = 0
15 FINDING A−1 USING COFACTORS 137
Practice Problems
Problem 15.1
Let
3 −2 1
A = 5 6 2 .
1 0 −3
(a) Find adj(A).
(b) Compute |A|.
Problem 15.2
Let A be an n × n matrix. Show that |adj(A)| = |A|n−1 .
Problem 15.3
If
3 0 1
A−1 = 0 2 3 ,
3 1 −1
find adj(A).
Problem 15.4
If |A| = 2, find |A−1 + adj(A)|.
Problem 15.5
Show that adj(αA) = αn−1 adj(A).
Problem 15.6
Consider the matrix
1 2 3
A = 2 3 4 .
1 5 7
(a) Find |A|.
(b) Find adj(A).
(c) Find A−1 .
Problem 15.7
Prove that if A is symmetric then adj(A) is also symmetric.
138 DETERMINANTS
Problem 15.8
Prove that if A is a non-singular triangular matrix then adj(A) is a lower
triangular matrix.
Problem 15.9
Prove that if A is a non-singular triangular matrix then A−1 is also triangular.
Problem 15.10
Let A be an n × n matrix.
(a) Show that if A has integer entries and |A| = 1 then A−1 has integer
entries as well.
(b) Let Ax = b. Show that if the entries of A and b are integers and |A| = 1
then the entries of x are also integers.
Problem 15.11
Let
1 −1 2
A = 4 0 6 .
0 1 −1
(a) Find |A|.
(b) Find adj(A).
(c) Find A−1 .
Problem 15.12
Let
1 1 1
A = 3 2 1 .
4 5 1
(a) Find |A|.
(b) Find the (3, 2)−entry of A−1 without finding A−1 .
Problem 15.13
Let
1 −1 2
A = 2 −1 1 .
1 1 −1
(a) Find the cofactor matrix of A.
(b) Find adj(A).
(c) Find |A|.
(d) Find A−1 .
15 FINDING A−1 USING COFACTORS 139
Problem 15.14
Find the inverse of
1 0 1
A = 2 −1 2
2 1 0
using the adjoint method.
Problem 15.15
Let A be an n × n non-singular matrix. Show that adj(adj(A)) = |A|n−2 A.
Problem 15.16
Let D = [dii ] be an invertible n × n diagonal matrix. Find the (i, i) entry of
adj(D).
Problem 15.17
Decide whether the given matrix is invertible, and if so, use the adjoint
method to find its inverse.
2 −3 5
A = 0 1 −3 .
0 0 2
Problem 15.18
Let A be a 3 × 3 matrix such that |A| = 5. Find |adj(A)|.
Problem 15.19
Let A be a 3 × 3 matrix such that |A| = 5. Find |A · adj(A)|.
Problem 15.20
Show that adj(An ) = [adj(A)]n , where n is a positive integer.
140 DETERMINANTS
Theorem 16.1
Let Ax = b be a matrix equation with A = [aij ], x = [xi ], b = [bi ]. We have
|A|xi = |Ai |, i = 1, 2, · · · , n
(1) If |A| =
6 0 then the system Ax = b has a unique solution given by
|Ai |
xi = ,
|A|
where 1 ≤ i ≤ n.
(2) If |A| = 0 and |Ai | =
6 0 for some i then the system Ax = b has no solution.
(3) If |A| = |A1 | = · · · = |An | = 0 then the system Ax = b has an infinite
number of solutions.
Proof.
First of all, from the equation Ax = b, we have,
Now, let Ii (x) be the matrix obtained by replacing the ith column of In by
the vector x. Expanding along the ith row of Ii (x), we find |Ii (x)| = xi for
i = 1, 2, · · · , n.
Next, we have
16 APPLICATION OF DETERMINANTS TO SYSTEMS: CRAMER’S RULE141
a11 a12 · · · a1,i−1 (a11 x1 + a12 x2 + · · · + a1n xn ) a1,i+1 · · · a1n
a21 a22 · · · a2,i−1 (a21 x1 + a22 x2 + · · · + a2n xn ) a2,i+1 · · · a1n
AIi (x) = ..
.. .. .. .. .. .. ..
. . . . . . . .
an1 an2 · · · an,i−1 (an1 x1 + an2 x2 + · · · + ann xn ) an,i+1 · · · ann
a11 a12 · · · a1,i−1 b1 a1,i+1 · · · a1n
a21 a22 · · · a2,i−1 b2 a2,i+1 · · · a1n
= .. .. = Ai .
.. .. .. .. .. ..
. . . . . . . .
an1 an2 · · · an,i−1 bn an,i+1 · · · ann
Hence,
|Ai | = |AIi (x)| = |A||Ii (x)| = |A|xi , i = 1, 2, · · · , n.
Now, (1), (2), and (3) follow easily. This ends a proof of the theorem
Example 16.1
Use Cramer’s rule to solve
−2x1 + 3x2 − x3 = 1
x1 + 2x2 − x3 = 4
−2x1 − x2 + x3 = −3.
Solution.
By Cramer’s rule we have
−2 3 −1
A= 1 2 −1 , |A| = −2.
−2 −1 1
1 3 −1
A1 = 4 2 −1 , |A1 | = −4.
−3 −1 1
−2 1 −1
A2 = 1 4 −1 , |A2 | = −6.
−2 −3 1
−2 3 1
A3 = 1 2 4 , |A3 | = −8.
−2 −1 −3
|A1 | |A2 | |A3 |
Thus, x1 = |A|
= 2, x2 = |A|
= 3, x3 = |A|
=4
142 DETERMINANTS
Example 16.2
Use Cramer’s rule to solve
5x1 − 3x2 − 10x3 = −9
2x1 + 2x2 − 3x3 = 4
−3x1 − x2 + 5x3 = 1.
Solution.
By Cramer’s rule we have
5 −3 −10
A= 2 2 −3 , |A| = −2.
−3 −1 5
−9 −3 −10
A1 = 4 2 −3 , |A1 | = 66.
1 −1 5
5 −9 −10
A2 = 2 4 −3 , |A2 | = −16.
−3 1 5
5 −3 −9
A3 = 2 2 4 , |A3 | = 36.
−3 −1 1
|A1 | |A2 | |A3 |
Thus, x1 = |A|
= −33, x2 = |A|
= 8, x3 = |A|
= −18
16 APPLICATION OF DETERMINANTS TO SYSTEMS: CRAMER’S RULE143
Practice Problems
Problem 16.1
Use Cramer’s Rule to solve
x1 + 2x3 = 6
−3x1 + 4x2 + 6x3 = 30
−x1 − 2x2 + 3x3 = 8.
Problem 16.2
Use Cramer’s Rule to solve
5x1 + x2 − x3 = 4
9x1 + x2 − x3 = 1
x1 − x2 + 5x3 = 2.
Problem 16.3
Use Cramer’s Rule to solve
4x1 − x2 + x3 = −5
2x1 + 2x2 + 3x3 = 10
5x1 − 2x2 + 6x3 = 1.
Problem 16.4
Use Cramer’s Rule to solve
3x1 − x2 + 5x3 = −2
−4x1 + x2 + 7x3 = 10
2x1 + 4x2 − x3 = 3.
Problem 16.5
Use Cramer’s Rule to solve
−x1 + 2x2 + 3x3 = −7
−4x1 − 5x2 + 6x3 = −13
7x1 − 8x2 − 9x3 = 39.
144 DETERMINANTS
Problem 16.6
Use Cramer’s Rule to solve
3x1 − 4x2 + 2x3 = 18
4x1 + x2 − 5x3 = −13
2x1 − 3x2 + x3 = 11.
Problem 16.7
Use Cramer’s Rule to solve
5x1 − 4x2 + x3 = 17
6x1 + 2x2 − 3x3 = 1
x1 − 4x2 + 3x3 = 15.
Problem 16.8
Use Cramer’s Rule to solve
2x1 − 3x2 + 2x3 = 1
3x1 + 2x2 − x3 = 16
x1 − 5x2 + 3x3 = −7.
Problem 16.9
Use Cramer’s Rule to solve
x1 − 2x2 + 2x3 = 5
3x1 + 2x2 − 3x3 = 13
2x1 − 5x2 + x3 = 2.
Problem 16.10
Use Cramer’s Rule to solve
5x1 − x2 + 3x3 = 10
6x1 + 4x2 − x3 = 19
x1 − 7x2 + 4x3 = −15.
Problem 16.11
Using Cramer’s Rule, determine whether the system is consistent or incon-
sistent. If the system is consistent, determine whether it is independent or
dependent.
x1 + 2x2 + x3 = 3
x1 + x2 + 3x3 = 1
3x1 + 4x2 + 7x3 = 1.
16 APPLICATION OF DETERMINANTS TO SYSTEMS: CRAMER’S RULE145
Problem 16.12
Using Cramer’s Rule, determine whether the system is consistent or incon-
sistent. If the system is consistent, determine whether it is independent or
dependent.
3x1 + x2 + 7x3 + 2x4 = 13
2x − 4x + 14x − x = −10
1 2 3 4
5x1 + 11x2 − 7x3 + 8x4 = 59
2x1 + 5x2 − 4x3 − 3x4 = 39.
Problem 16.13
Using Cramer’s Rule, determine whether the system is consistent or incon-
sistent. If the system is consistent, determine whether it is independent or
dependent.
x1 − 2x2 − 6x3 = 12
2x1 + 4x2 + 12x3 = −17
x1 − 4x2 − 12x3 = 22.
Problem 16.14
Use Cramer’s rule to solve for x3 only.
x1 + x2 + 2x3 = 8
−x1 − 2x2 + 3x3 = 1
3x1 − 7x2 + 4x3 = 10.
Problem 16.15
Use Cramer’s rule to express cos α in terms of a, b, and c.
b cos α + a cos β = c,
c cos α + a cos γ = b,
c cos β + b cos γ = a.
Problem 16.16
Solve the following linear system by using Cramer’s rule.
11 13 17 x1 0
5 3 2 x2 = 1 .
7 2 1 x3 0
146 DETERMINANTS
Problem 16.17
True or false: One can find the value of a variable in a square linear system
without using the values of the remaining variables.
Problem 16.18
True or false: Cramer’s Rule can be applied to overdetermined and under-
determined linear systems.
Problem 16.19
Consider the system
a11 x1 + a12 x2 = b1 ,
a21 x1 + a22 x2 = b2 .
Problem 16.20
List three techniques that you have learned so far for solving square linear
systems.
The Theory of Vector Spaces
147
148 THE THEORY OF VECTOR SPACES
Theorem 17.1
The following properties hold, for u, v, w in Rn and α, β scalars:
(a) u + v = v + u
(b) u + (v + w) = (u + v) + w
(c) u + 0 = 0 + u = u where 0 = (0, 0, . . . , 0)
(d) u + (−u) = 0
(e) α(u + v) = αu + αv
(f) (α + β)u = αu + βu
(g) α(βu) = (αβ)u
(h) 1u = u.
17 VECTOR SPACES AND SUBSPACES 149
The set Rn with the above operations and properties is called the Euclidean
space.
Example 17.1
Let Mmn be the collection of all m × n matrices. Show that Mmn is a vector
space using matrix addition and scalar multiplication.
Solution.
See Problem 17.17
Example 17.2
Let V = {(x, y) : x ≥ 0, y ≥ 0}. Show that the set V fails to be a vector
space under the standard operations on R2 .
Solution.
For any (x, y) ∈ V with x, y > 0, we have −(x, y) = (−x, −y) 6∈ V. Thus, V
is not a vector space
The following theorem exhibits some properties which follow directly from
the axioms of the definition of a vector space and therefore hold for every
vector space.
Theorem 17.2
Let V be a vector space, u a vector in V and α is a scalar. Then the following
properties hold:
(a) (−1)u = −u
(b) 0u = 0.
150 THE THEORY OF VECTOR SPACES
(c) α0 = 0
(d) If αu = 0 then α = 0 or u = 0.
Proof.
See Problem 17.18
Vector spaces may be formed from subsets of other vectors spaces. These
are called subspaces. A non-empty subset W of a vector space V is called a
subspace of V if the following two properties are satisfied:
Every vector space V has at least two subspaces: V itself and the subspace
consisting of the zero vector of V. These are called the trivial subspaces of
V.
Example 17.3
Let S be the subset of Mnn consisting of symmetric matrices. Show that S
is a subspace of Mnn .
Solution.
Since InT = In , In ∈ S and S is non-empty. Now, if A and B belong to S then
(A + B)T = AT + B T = A + B and (αA)T = αAT = αA. That is, A + B ∈ S
and αA ∈ S. By the previous theorem, S is a subspace of Mnn
Example 17.4
Show that a subspace of a vector space is itself a vector space.
Solution.
The elements of the subspace inherits all the axioms of a vector space
Theorem 17.3
Show that W 6= ∅ is a subspace of V if and only if αu + v ∈ W for all
u, v ∈ W and α ∈ R.
17 VECTOR SPACES AND SUBSPACES 151
Proof.
Suppose that W is a non-empty subspace of V. If u, v ∈ W and α ∈ R then
αu ∈ W and therefore αu + v ∈ W. Conversely, suppose that for all u, v ∈ W
and α ∈ R we have αu + v ∈ R. In particular, if α = 1 then u + v ∈ W. If
v = 0 then αu + v = αu ∈ W. Hence, W is a subspace
Example 17.5
Let M22 be the collection of 2 × 2 matrices. Show that the set W of all 2 × 2
matrices having zeros on the main diagonal is a subspace of M22 .
Solution.
The set W is the set
0 a
W = : a, b ∈ R .
b 0
0 a0 αa + a0
0 a 0
α + 0 = ∈W
b 0 b 0 αb + b0 0
Practice Problems
Problem 17.1
Let D([a, b]) be the collection of all differentiable functions on [a, b]. Show
that D([a, b]) is a subspace of the vector space of all functions defined on
[a, b].
Problem 17.2
Let A be an m × n matrix. Show that the set S = {x ∈ Rn : Ax = 0} is a
subspace of Rn .
Problem 17.3
Let P be the collection of polynomials in the indeterminate x. Let p(x) =
a0 + a1 x + a2 x2 + · · · and q(x) = b0 + b1 x + b2 x2 + · · · be two polynomials
in P. Define the operations:
(a) Addition: p(x) + q(x) = (a0 + b0 ) + (a1 + b1 )x + (a2 + b2 )x2 + · · ·
(b) Multiplication by a scalar: αp(x) = αa0 + (αa1 )x + (αa2 )x2 + · · · .
Show that P is a vector space.
Problem 17.4
Let F (R) be the set of functions f : R → R. Define the operations
and
(αf )(x) = αf (x).
Show that F (R) is a vector space under these operations.
Problem 17.5
Define on R2 the following operations:
(i) (x, y) + (x0 , y 0 ) = (x + x0 , y + y 0 );
(ii) α(x, y) = (αy, αx).
Show that R2 with the above operations is not a vector space.
Problem 17.6
Let U = {p(x) ∈ P : p(3) = 0}. Show that U is a subspace of P.
Problem 17.7
Let Pn denote the collection of all polynomials of degree n. Show that Pn is
a subspace of P.
17 VECTOR SPACES AND SUBSPACES 153
Problem 17.8
Show that the set S = {(x, y) : x ≤ 0} is not a vector space of R2 under the
usual operations of R2 .
Problem 17.9
Show that the collection C([a, b]) of all continuous functions on [a, b] with
the operations:
is a vector space.
Problem 17.10
Let S = {(a, b, a + b) : a, b ∈ R}. Show that S is a subspace of R3 under the
usual operations.
Problem 17.11
Let V be a vector space. Show that if u, v, w ∈ V are such that u + v = u + w
then v = w.
Problem 17.12
Let H and K be subspaces of a vector space V.
(a) The intersection of H and K, denoted by H ∩ K, is the subset of V
that consists of elements that belong to both H and K. Show that H ∩ K is
a subspace of V.
(b) The union of H and K, denoted by H ∪ K, is the susbet of V that
consists of all elements that belong to either H or K. Give, an example of
two subspaces of V such that H ∪ K is not a subspace.
(c) Show that if H ⊂ K or K ⊂ H then H ∪ K is a subspace of V.
Problem 17.13
a a−2
Show that S = is not a subspace of M22 , the vector space of
b c
2 × 2 matrices.
Problem 17.14
Show that V = {(a, b, c) : a + b = 2c} with the usual operations on R3 is a
vector space.
154 THE THEORY OF VECTOR SPACES
Problem 17.15
Show that S = {(a, b) ∈ R2 : a2 − b2 = 0} is not a subspace of R2 .
Problem 17.16
Prove Theorem 17.1.
Problem 17.17
Let Mmn be the collection of all m × n matrices. Show that Mmn is a vector
space using matrix addition and scalar multiplication.
Problem 17.18
Prove Theorem 17.2
Problem 17.19
Let V be a vector space over R and v1 , v2 , · · · , vn be elements of V. Let
Problem 17.20
Show that the set S all n × n diagonal matrices is a subspace of Mnn .
18 BASIS AND DIMENSION 155
ay 00 + by 0 + cy = 0 a < t < b.
Example 18.1
Show that the vector w ~ = (9, 2, 7) is a linear combination of the vectors
~ 0 = (4, −1, 8) is not.
~u = (1, 2, −1) and ~v = (6, 4, 2) whereas the vector w
156 THE THEORY OF VECTOR SPACES
Solution.
We must find numbers s and t such that
Solving the first two equations one finds s = −3 and t = 2 both values satisfy
the third equation.
Turning to (4, −1, 8), the question is whether s and t can be found such that
(4, −1, 8) = s(1, 2, −1) + t(6, 4, 2). Equating components gives
s + 6t = 4
2s + 4t = −1
−s + 2t = 8.
Theorem 18.1
Let W = {v1 , v2 , . . . , vn } be a subset of a vector space V. Let span(W ) be
the collection of all linear combinations of elements of W. Then span(W ) is
a subspace of V.
Proof.
Let u = α1 v1 + α2 v2 + · · · + αn vn and v = β1 v1 + β2 v2 + · · · + βn vn . Then for
any scalar α, we have
αu + v =α(α1 v1 + α2 v2 + · · · + αn vn ) + (β1 v1 + β2 v2 + · · · + βn vn )
=(αα1 + β1 )v1 + (αα2 + β2 )v2 + · · · + (ααn + βn )vn ∈ W.
Hence, W is a subspace of V
18 BASIS AND DIMENSION 157
Example 18.2
Show that Pn = span{1, x, x2 , · · · , xn }.
Solution.
Clearly, span{1, x, x2 , · · · , xn } ⊆ Pn . If p(x) ∈ Pn then there are scalars
a0 , a1 · · · , an such that p(x) = a0 + a1 x + · · · + an xn ∈ span{1, x, · · · , xn }.
That is, Pn ⊆ span{1, x, x2 , · · · , xn }
Example 18.3
Show that Rn = span{e1 , e2 , · · · , en } where ei is the vector with 1 in the ith
component and 0 otherwise.
Solution.
Clearly, span{e1 , e2 , · · · , en } ⊆ Rn . We must show that if u ∈ Rn then u is a
linear combination of the e0i s. Indeed, if u = (x1 , x2 , · · · , xn ) ∈ Rn then
u = x1 e1 + x2 e2 + · · · + xn en
Hence u lies in span{e1 , e2 , · · · , en }. That is, Rn ⊆ span{e1 , e2 , · · · , en }
The problem is reduced of showing that the above system is consistent. This
system will be consistent if and only if the coefficient matrix A
1 1 2
A = 1 0 1
2 1 3
158 THE THEORY OF VECTOR SPACES
Next, we introduce a concept which guarantees that any vector in the span
of a set S ⊆ V has only one representation as a linear combination of vectors
in S. Spanning sets with this property play a fundamental role in the study
of vector spaces as we shall see later in this section.
If v1 , v2 , . . . , vn are vectors in a vector space with the property that
α1 v1 + α2 v2 + · · · + αn vn = 0
Example 18.5
Show that the set S = {1, x, x2 , · · · , xn } is a linearly independent set in Pn .
Solution.
Suppose that a0 +a1 x+a2 x2 +· · ·+an xn = 0 for all x ∈ R. By the Fundamental
Theorem of Algebra, a polynomial of degree n has at most n roots. But by
the above equation, every real number is a root of the equation. This forces
the numbers a0 , a1 , · · · , an to be 0
Example 18.6
Let u be a nonzero vector. Show that {u} is linearly independent.
Solution.
Suppose that αu = 0. If α 6= 0 then we can multiply both sides by α−1 and
obtain u = 0. But this contradicts the fact that u is a nonzero vector. Hence,
we must have α = 0
Example 18.7
(a) Show that the vectors v~1 = (1, 0, 1, 2), v~2 = (0, 1, 1, 2), and v~3 = (1, 1, 1, 3)
are linearly independent.
(b) Show that the vectors v~1 = (1, 2, −1), v~2 = (1, 2, −1), and v~3 = (1, −2, 1)
are linearly dependent.
18 BASIS AND DIMENSION 159
Solution.
(a) Suppose that s, t, and w are real numbers such that sv~1 + tv~2 + wv~3 = 0.
Then equating components gives
s + w=0
t+ w =0
s+ t+ w =0
2s + 2t + 3w = 0.
The second and third equation leads to s = 0. The first equation gives w = 0
and the second equation gives t = 0. Thus, the given vectors are linearly
independent.
(b) These vectors are linearly dependent since v~1 + v~2 −2v~3 = 0. Alternatively,
the coefficient matrix of homogeneous system
s + 2t − w = 0
s + 2t − w = 0
s − 2t + w = 0.
is singular
Example 18.8
Show that the unit vectors e1 , e2 , · · · , en in Rn are linearly independent.
Solution.
Suppose that x1 e1 + x2 e2 + · · · + xn en = (0, 0, · · · , 0). Then (x1 , x2 , · · · , xn ) =
(0, 0, · · · , 0) and this leads to x1 = x2 = · · · = xn = 0. Hence the vectors
e1 , e2 , · · · , en are linearly independent
Example 18.9
Let ei be the vector of Rn whose ith component is 1 and zero otherwise. Show
that the set S = {e1 , e2 , . . . , en } is a basis for Rn . This is called the standard
basis of Rn .
160 THE THEORY OF VECTOR SPACES
Solution.
By Example 18.3, we have Rn = span{e1 , e2 , · · · , en }. By Example 18.8,
the vectors e1 , e2 , · · · , en are linearly independent. Thus {e1 , e2 , · · · , en } is a
basis of R3
Example 18.10
Show that {1, x, x2 , · · · , xn } is a basis of Pn .
Solution.
By Example 18.2, Pn = span{1, x, x2 , · · · , xn } and by Example 18.5, the set
S = {1, x, x2 , · · · , xn } is linearly independent. Thus, S is a basis of Pn
Example 18.11
Determine a basis and the dimension for the solution space of the homoge-
neous system
2x1 + 2x2 − x3
+ x5 + = 0
−x − x + 2x − 3x + x
=0
1 2 3 4 5
x1 + x2 − 2x3 − x5 =0
x3 + x4 + x5 = 0.
Solution.
By Example 6.1, we found that x1 = −s − t, x2 = s, x3 = −t, x4 = 0, x5 = t.
So if S is the vector space of the solutions to the given system then S =
{(−s − t, s, −t, 0, t) : s, t ∈ R} = {s(−1, 1, 0, 0, 0) + t(−1, 0, −1, 0, 1) : s, t ∈
R} = span{(−1, 1, 0, 0, 0), (−1, 0, −1, 0, 1)}. Moreover, if s(−1, 1, 0, 0, 0) +
t(−1, 0, −1, 0, 1) = (0, 0, 0, 0, 0) then s = t = 0. Thus, the set
The following theorem will indicate the importance of the concept of a basis
in investigating the structure of vector spaces. In fact, a basis for a vector
space V determines the representation of each vector in V in terms of the
vectors in that basis.
Theorem 18.2
If S = {v1 , v2 , . . . , vn } is a basis for V then any element of V can be written
in one and only one way as a linear combination of the vectors in S.
Proof.
Suppose v ∈ V has the following two representations v = α1 v1 + α2 v2 + · · · +
αn vn and v = β1 v1 + β2 v2 + · · · + βn vn . We want to show that αi = βi for
1 ≤ i ≤ n. But this follows from (α1 −β1 )v1 +(α2 −β2 )v2 +· · ·+(αn −βn )vn = 0
and the fact that S is a linearly independent set
Remark 18.1
A vector space can have different bases; however, all of them have the same
number of elements.
The following theorem tells us that one of the conditions of the definition of
a basis is sufficient in the case the vector space is finite dimensional.
Theorem 18.3
Let V be a finite dimensional vector space with dim(V ) = n.
(i) Any set of n linearly independent vectors of V is a basis of V.
(ii) Any span of V that consists of n vectors is a basis of V.
Proof.
See Problem 18.22
162 THE THEORY OF VECTOR SPACES
Practice Problems
Problem 18.1
Let W = span{v1 , v2 , · · · , vn }, where v1 , v2 , · · · , vn are vectors in V. Show
that any subspace U of V containing the vectors v1 , v2 , · · · , vn must con-
tain W, i.e. W ⊂ U. That is, W is the smallest subspace of V containing
v1 , v2 , · · · , vn .
Problem 18.2
Show that the polynomials p1 (x) = 1 − x, p2 (x) = 5 + 3x − 2x2 , and p3 (x) =
1 + 3x − x2 are linearly dependent vectors in P2 .
Problem 18.3
Express the vector ~u = (−9, −7, −15) as a linear combination of the vectors
v~1 = (2, 1, 4), v~2 = (1, −1, 3), v~3 = (3, 2, 5).
Problem 18.4
(a) Show that the vectors v~1 = (2, 2, 2), v~2 = (0, 0, 3), and v~3 = (0, 1, 1) span
R3 .
(b) Show that the vectors v~1 = (2, −1, 3), v~2 = (4, 1, 2), and v~3 = (8, −1, 8)
do not span R3 .
Problem 18.5
Show that
1 0 0 1 0 0 0 0
M22 = span , , , .
0 0 0 0 1 0 0 1
Problem 18.6
Show that the vectors v~1 = (2, −1, 0, 3), v~2 = (1, 2, 5, −1), and v~3 = (7, −1, 5, 8)
are linearly dependent.
Problem 18.7
Show that the vectors v~1 = (4, −1, 2) and v~2 = (−4, 10, 2) are linearly inde-
pendent.
Problem 18.8
Show that the {u, v} is linearly dependent if and only if one is a scalar
multiple of the other.
18 BASIS AND DIMENSION 163
Problem 18.9
Find a basis for the vector space M22 of 2 × 2 matrices.
Problem 18.10
Show that B = {1 + 2x, x − x2 , x + x2 } is a basis for P2 .
Problem 18.11
Determine the value(s) of a so that {p1 (x), p2 (x), p3 (x)} is linearly indepen-
dent in P2 where
Problem 18.12
How many different representations can the vector ~b = (2, −1, 3) be written
as a linear combination of the vectors ~v1 = (1, −2, 0), ~v2 = (0, 1, 1), and
~v3 = (5, −6, 4)?
Problem 18.13
Based on the previous problem, explain why {~v1 , ~v2 , ~v3 } can not be a basis
of R3 .
Problem 18.14
Let W be the subspace of M22 :
a 0
W = : a, b ∈ R .
0 b
Problem 18.15
Suppose that V = span{v1 , v2 , · · · , vn }. Show that if vk is a linear combina-
tion of the remaining v 0 s then V = span{v1 , · · · , vk−1 , vk+1 , · · · , vn }.
Problem 18.16
Let {v1 , v2 , · · · , vn } be a basis for a vector space V and S = {w1 , w2 , · · · , wm }
be a subset of V with m > n. I want to show that S is linearly dependent.
(a) Show that there is a matrix A = [aij ] of dimension m × n such that
164 THE THEORY OF VECTOR SPACES
(c) Show that the homogeneous system in (b) has a non-trivial solution.
Hence, concluding that S is linearly dependent.
Problem 18.17
Let {v1 , v2 , · · · , vn } be a basis for a vector space V and S = {w1 , w2 , · · · , wm }
be a subset of V with m < n. I want to show that S can not span V.
(a) Suppose that V = span(S). Show that there is a matrix A = [aij ] of
dimension n × m such that Aw = v, where v = [vi ] is n × 1 and w = [wi ] is
m × 1.
(b) Show that the homogeneous system AT α = 0, where
α1 0
a11 a21 ... an1
a12 a22 ... an2 α2
0
=
... ... ...
... ... ...
a1m a2m ... anm αn 0
Problem 18.18
Show that if {v1 , v2 , · · · , vn } is a basis of a vector space V then every basis
of V has exactly n elements.
Problem 18.19
Let V be a vector space with dim(V ) = n. I want to show that any subset
of independent vectors of V can be extended to a basis of V.
(a) Let S = {v1 , v2 , · · · , vm } be a linearly independent subset of V with
V 6= span(S). Show that there is a vector v ∈ V and v 6∈ span(S) such that
18 BASIS AND DIMENSION 165
Problem 18.20
Let V be a vector space with dim(V ) = n. I want to show that any spanning
subset of V can be extended to a basis of V.
(a) Suppose that V = {span}{v1 , v2 , · · · , vm }. Show that if {v1 , v2 , · · · , vm }
is linearly dependent then there is a subset S1 ⊆ {v1 , v2 , · · · , vm } such that
V = {span}(S1 ).
(b) Continuing the process in (a), show that there is a subset S of {v1 , v2 , · · · , vm }
that is a basis of V.
Problem 18.21
Let V be a vector space of dimension n and W a subspace of V. Show that
if dim(W ) = dim(V ) then W = V.
Problem 18.22
Use Problem 18.21 to prove Theorem 18.3.
166 THE THEORY OF VECTOR SPACES
Eigenvalues and Eigenvectors
167
168 EIGENVALUES AND EIGENVECTORS
λeλt y = eλt Ay
where
1 −2
A=
3 −4
or
Ay = λy.
Thus, we need to find λ and y from this matrix equation.
Example 19.1
1
Show that x = is an eigenvector of the matrix
2
3 0
A=
8 −1
corresponding to the eigenvalue λ = 3.
19 THE EIGENVALUES OF A SQUARE MATRIX 169
Solution.
The value λ = 3 is an eigenvalue of A with eigenvector x since
3 0 1 3
Ax = = = 3x
8 −1 2 6
Eigenvalues can be either real numbers or complex numbers. To find the
eigenvalues of a square matrix A we rewrite the equation Ax = λx as
Ax = λIn x
or equivalently
(λIn − A)x = 0.
For λ to be an eigenvalue, there must be a nonzero solution to the above
homogeneous system. But, the above system has a nontrivial solution if and
only if the coefficient matrix (λIn − A) is singular, that is , if and only if
|λIn − A| = 0.
This equation is called the characteristic equation of A.
Example 19.2
Find the characteristic equation of the matrix
0 1 0
A = 0 0 1 .
4 −17 8
Solution.
The characteristic equation of A is the equation
λ −1 0
0
λ −1 = 0.
−4 17 λ − 8
Example 19.3
Show that the constant term in the characteristic polynomial of a matrix A
is (−1)n |A|.
Solution.
The constant term of the polynomial p(λ) corresponds to p(0). It follows that
p(0) = constant term = | − A| = (−1)n |A|
Example 19.4
Find the eigenvalues of the matrix
3 2
A= .
−1 0
Solution.
The characteristic equation of A is given by
λ − 3 −2
= 0.
1 λ
Expanding the determinant and simplifying, we obtain
λ2 − 3λ + 2 = 0
or
(λ − 1)(λ − 2) = 0.
Thus, the eigenvalues of A are λ = 2 and λ = 1
Example 19.5
Find the eigenvalues of the matrix
0 1 0
A = 0 0 1 .
4 −17 8
Solution.
According to Example 19.2, the characteristic equation of A is λ3 − 8λ2 +
17λ − 4 = 0. Using the rational root test we find that λ = 4 is a solution to
this equation. Using synthetic division of polynomials we find
(λ − 4)(λ2 − 4λ + 1) = 0.
The eigenvalues√of the matrix A√are the solutions to this equation, namely,
λ = 4, λ = 2 + 3, and λ = 2 − 3
19 THE EIGENVALUES OF A SQUARE MATRIX 171
Example 19.6
(a) Show that the eigenvalues of a triangular matrix are the entries on the
main diagonal.
(b) Find the eigenvalues of the matrix
1
2
0 0
2
A = −1 3
0 .
5 −8 − 41
Solution.
(a) Suppose that A is an upper triangular n × n matrix. Then the matrix
λIn − A is also upper triangular with entries on the main diagonal are λ −
a11 , λ − a22 , · · · , λ − ann . Since the determinant of a triangular matrix is just
the product of the entries of the main diagonal, the characteristic equation
of A is
(λ − a11 )(λ − a22 ) · · · (λ − ann ) = 0.
Hence, the eigenvalues of A are a11 , a22 , · · · , ann .
(b) Using (a), the eigenvalues of A are λ = 21 , λ = 23 , and λ = − 14
Example 19.7
Find the algebraic multiplicity of the eigenvalues of the matrix
2 1 0
A = 0 2 0 .
2 3 1
Solution.
The characteristic equation of the matrix A is
λ − 2 −1 0
0
λ−2 0 = 0.
−2 −3 λ − 1
(λ − 2)2 (λ − 1) = 0.
172 EIGENVALUES AND EIGENVECTORS
There are many matrices with real entries but with no real eigenvalues. An
example is given next.
Example 19.8
Show that the following matrix has no real eigenvalues.
0 1
A= .
−1 0
Solution.
The characteristic equation of the matrix A is
λ −1
1 λ = 0.
Example 19.10
Show that the following matrices are not similar.
1 2 1 1
A= ,B = .
2 1 1 1
Solution.
The eigenvalues of A are λ = 3 and λ = −1. The eigenvalues of B are λ = 0
and λ = 2. According to Example 19.9 (c), these two matrices cannot be
similar
Example 19.11
Let A be an n×n matrix with eigenvalues λ1 , λ2 , · · · , λn including repetitions.
Show the following.
(a) tr(A) = λ1 + λ2 + · · · + λn .
(b) |A| = λ1 λ2 · · · λn .
Solution.
Factoring the characteristic polynomial of A we find
p(λ) =(λ − λ1 )(λ − λ2 ) · · · (λ − λn )
=λn − (λ1 + λ2 + · · · + λn )λn−1 + · · · + (−1)n λ1 λ2 · · · λn .
(a) By Equation 19.1, tr(A) = λ1 + λ2 + · · · + λn .
(b) | − A| = p(0) = (−1)n λ1 λ2 · · · λn . But | − A| = (−1)n |A|. Hence, |A| =
λ1 λ2 · · · λn
Example 19.12
(a) Find the characteristic polynomial of
1 2
A= .
3 4
(b) Find the matrix A2 − 5A − 2I2 .
(c) Compare the result of (b) with (a).
Solution.
λ − 1 −2
(a) p(λ) = = λ2 − 5λ − 2.
−3 λ − 4
(b) Simple algebra shows A2 − 5A − 2I2 = 0.
(c) A satisfies p(A) = 0. That is, A satisfies its own characteristic equation
Proof.
Let p(λ) = λn + an−1 λn−1 + · · · + a1 λ + a0 be the characteristic polynomial
corresponding to A. We will show that p(A) = 0. The cofactors of A are
polynomials in λ of degree at most n − 1. Thus,
That is
Bn−1 λn + (Bn−2 − ABn−1 )λn−1 + (Bn−3 − ABn−2 )λn−2 + · · · + (B0 − AB1 )λ − AB0
=In λn−1 + an−1 In λn−1 + an−2 In λn−2 + · · · a1 In λ + a0 In .
Bn−1 =In
Bn−2 − ABn−1 =an−1 In
Bn−3 − ABn−2 =an−2 In
..
.
B0 − AB1 =a1 In
−AB0 =a0 In
0 = An + an−1 An−1 + · · · + a1 A + a0 In .
Example 19.13
Use the Cayley-Hamilton theorem to find the inverse of the matrix
1 2
A= .
3 4
19 THE EIGENVALUES OF A SQUARE MATRIX 175
Solution.
Since |A| = 4 − 6 = −2 6= 0, A−1 exists. By Cayley-Hamilton Theorem we
have
A2 − 5A − 2I2 =0
2I2 =A2 − 5A
2A−1 =A − 5I2
−1 1 2 5 0
2A = −
3 4 0 5
−4 2
=
3 −1
−1 −2 1
A = 3
2
− 12
176 EIGENVALUES AND EIGENVECTORS
Practice Problems
Problem 19.1
Find the eigenvalues of the matrix
5 8 16
A= 4 1 8 .
−4 −4 −11
Problem 19.2
Find the eigenvalues of the matrix
3 0
A= .
8 −1
Problem 19.3
Find the eigenvalues of the matrix
2 1 1
A = 2 1 −2 .
−1 0 −2
Problem 19.4
Find the eigenvalues of the matrix
−2 0 1
A = −6 −2 0 .
19 5 −4
Problem 19.5
1
Show that if λ is a nonzero eigenvalue of an invertible matrix A then λ
is an
eigenvalue of A−1 .
Problem 19.6
Show that if λ is an eigenvalue of a matrix A then λm is an eigenvalue of Am
for any positive integer m.
Problem 19.7
Show that if A is similar to a diagonal matrix D then Ak is similar to Dk .
19 THE EIGENVALUES OF A SQUARE MATRIX 177
Problem 19.8
Show that the identity matrix In has exactly one eigenvalue.
Problem 19.9
Let A be an n × n nilpotent matrix, i.e. Ak = 0 for some positive integer
k.
(a) Show that λ = 0 is the only eigenvalue of A.
(b) Show that p(λ) = λn .
Problem 19.10
Suppose that A and B are n × n similar matrices and B = P −1 AP. Show
that if λ is an eigenvalue of A with corresponding eigenvector x then λ is an
eigenvalue of B with corresponding eigenvector P −1 x.
Problem 19.11
Let A be an n × n matrix with n odd. Show that A has at least one real
eigenvalue.
Problem 19.12
Consider the following n × n matrix
0 1 0 0
0 0 1 0
A= 0
.
0 0 1
−a0 −a1 −a2 −a3
Problem 19.13
Show that if D is a diagonal matrix then Dk , where k is a positive integer, is
a diagonal matrix whose entries are the entries of D raised to the power k.
Problem 19.14
Show that A and AT have the same characteristic polynomial and hence the
same eigenvalues.
178 EIGENVALUES AND EIGENVECTORS
Problem 19.15
Find the eigenvalues of the matrix
1 −3 3
A = 3 −5 3 .
6 −6 4
Problem 19.16
Find the eigenvalues of the matrix
−2 −1
B= .
−5 2
Problem 19.17
Show that λ = 0 is an eigenvalue of a matrix A if and only if A is singular.
Problem 19.18
Show that
Problem 19.19
Consider the matrix
6 −2
.
6 −1
Using Cayley-Hamilton theorem, show that A6 = 665A − 1266I2 .
Problem 19.20
Use the Cayley-Hamilton theorem to find the inverse of the matrix
2 0 1
A = −2 3
4 .
−5 5 6
20 FINDING EIGENVECTORS AND EIGENSPACES 179
Theorem 20.1
Let Vλ denote the set of eigenvectors of a matrix corresponding to an eigen-
value λ. The set V λ = Vλ ∪ {0} is a subspace of Rn . This subspace is called
the eigenspace of A corresponding to λ.
Proof.
Let Vλ = {x ∈ Rn and x 6= {0} : Ax = λx}. We will show that V λ = Vλ ∪{0}
is a subspace of Rn .
(i) Let u, v ∈ V λ . If u = 0 or v = 0 then the sum is either u, v, or 0
which belongs to V λ . So assume that both u, v ∈ Vλ . We have A(u + v) =
Au + Av = λu + λv = λ(u + v). That is u + v ∈ V λ .
(ii) Let u ∈ V λ and α ∈ R. Then A(αu) = αAu = λ(αu) so αu ∈ V λ . Hence,
V λ is a subspace of Rn
Example 20.1
Find the eigenspaces of the matrix
2 1 0
A = 0 2 0 .
2 3 1
Solution.
The characteristic equation of the matrix A is
λ−2 −1 0
0 λ−2 0 .
−2 −3 λ − 1
180 EIGENVALUES AND EIGENVECTORS
(λ − 2)2 (λ − 1) = 0.
x1 = 0, x2 = 0, x3 = s.
1
x1 = s, x2 = 0, x3 = s.
2
20 FINDING EIGENVECTORS AND EIGENSPACES 181
Practice Problems
Problem 20.1
Show that λ = −3 is an eigenvalue of the matrix
5 8 16
A= 4 1 8
−4 −4 −11
and then find the corresponding eigenspace V −3 .
Problem 20.2
Find the eigenspaces of the matrix
3 0
A= .
8 −1
Problem 20.3
Find the eigenspaces of the matrix
2 1 1
A = 2 1 −2 .
−1 0 −2
Problem 20.4
Find the bases of the eigenspaces of the matrix
−2 0 1
A = −6 −2 0 .
19 5 −4
Problem 20.5
Find the eigenvectors and the eigenspaces of the matrix
−1 4 −2
A = −3 4 0 .
−3 1 3
Problem 20.6
Find the eigenvectors and the eigenspaces of the matrix
0 0 −2
A = 1 2 1 .
1 0 3
20 FINDING EIGENVECTORS AND EIGENSPACES 185
Problem 20.7
Find the eigenvectors and the eigenspaces of the matrix
2 4 3
A = −4 −6 −3 .
3 3 1
Problem 20.8
Find the eigenvectors and the eigenspaces of the matrix
−1 1 1 −2
−1 1 3 2
A= .
1 1 −1 −2
0 −1 −1 1
Problem 20.9
Find the geometric and algebraic multiplicities of the eigenvalues of the ma-
trix
1 1 0
A = 0 1 1 .
0 0 1
Problem 20.10
When an n × n matrix has a eigenvalue whose geometric multiplicity is less
than the algebraic multiplicity, then it is called a defective matrix. Is A
defective?
1 1 −1
A = 0 2 −1 .
0 0 1
Problem 20.11
Find bases for the eigenspaces of the matrix
0 0 −2
A= 1 2 1 .
1 0 3
Problem 20.12
Find the eigenspaces of the matrix
3 −2 0
A = −2 3 0 .
0 0 5
186 EIGENVALUES AND EIGENVECTORS
Problem 20.13
Find the eigenspaces of the matrix
2 −1
A= .
5 −2
Problem 20.14
Suppose that λ1 and λ2 are two distinct eigenvalues of a square matrix A
with corresponding eigenvectors v1 and v2 . Show that {v1 , v2 } is linearly
independent.
Problem 20.15
Show that if λ is an eigenvalue of A with corresponding eigenvector v then for
any scalar r, λ − r is an eigenvalue of A − rIn with corresponding eigenvector
v.
Problem 20.16
Let
2 0
A= .
2 3
Show that A and AT do not have the same eigenspaces.
Problem 20.17
Let A be a 2 × 2 matrix such that A2 = I2 . Suppose that the eigenvectors of
A are x + Ax and x − Ax. Find the corresponding eigenvalue.
Problem 20.18
Show that if A is invertible then it can not have 0 as an eigenvalue.
Problem 20.19
Consider the following matrix
−9 15 3
A = −12 18 3 .
24 −30 −3
Problem 20.20
Show that if A2 is the zero matrix, then the only eigenvalue of A is zero.
Problem 20.21
Suppose A is 3, and v is an eigenvector of A corresponding to an eigenvalue
of 7. Is v an eigenvector of 2I3 − A? If so, find the corresponding eigenvalue.
Problem 20.22
If each row of A sums to the same number s, what is one eigenvalue and
eigenvector?
188 EIGENVALUES AND EIGENVECTORS
21 Diagonalization of a Matrix
In this section, we shall discuss a method for finding a basis of Rn consisting
of the eigenvectors of a given n × n matrix A. It turns out that this is
equivalent to finding an invertible matrix P such that P −1 AP = D is a
diagonal matrix. That is, A is similar to D. In this case, the columns of P
are the eigenvectors of A and the diagonal entries of D are the eigenvalues
of A. Also, the columns of P form a basis of the Euclidean space Rn . We say
that A is a diagonalizable matrix.
The computations of diagonalizable matrices is one of the most frequently
applied numerical processes in applications such as quantum mechanics or
differential equations.
The next theorem gives a characterization of diagonalizable matrices. In
fact, this theorem and Theorem 18.3 support our statement mentioned at the
beginning of this section that the problem of finding a basis of Rn consisting
of eigenvectors of A is equivalent to diagonalizing A.
Theorem 21.1
If A is an n × n square matrix, then the following statements are equivalent.
(a) A is diagonalizable.
(b) A has n linearly independent eigenvectors.
Proof.
(a) ⇒ (b) : Suppose A is diagonalizable. Then there are an invertible matrix
P and a diagonal matrix D such that A = P DP −1 . That is
p11 p12 . . . p1n λ1 0 0 . . . 0
p21 p22 . . . p2n 0 λ2 0 . . . 0
AP = P D = .. .. .
.. ..
. . . .
pn1 pn2 . . . pnn 0 0 0 . . . λn
For 1 ≤ i ≤ n, let
p1i
p2i
pi = .
..
.
pni
Then the columns of AP are λ1 p1 , λ2 p2 , . . . , λn pn . But AP = [Ap1 , Ap2 , . . . , Apn ].
Hence, Api = λi pi , for 1 ≤ i ≤ n. Thus λ1 , λ2 , . . . , λn are eigenvalues of
21 DIAGONALIZATION OF A MATRIX 189
Then the columns of AP are Ap1 , Ap2 , . . . , Apn . But Ap1 = λ1 p1 , Ap2 =
λ2 p2 , . . . , Apn = λn pn . Hence
λ1 p11 λ2 p12 . . . λn p1n
λ1 p21 λ2 p22 . . . λn p2n
AP =
.. ..
. .
λ1 pn1 λ2 pn2 . . . λn pnn
p11 p12 . . . p1n λ1 0 0 . . . 0
p21 p22 . . . p2n 0 λ2 0 . . .
0
= = P D.
.. .. .. ..
. . . .
pn1 pn2 . . . pnn 0 0 0 ... λn
How do we find P and D? From the above proof we obtain the following
procedure for diagonalizing a diagonalizable matrix.
Step 1. Find n linearly independent eigenvectors of A, say p1 , p2 , · · · , pn .
Step 2. Form the matrix P having p1 , p2 , · · · , pn as its column vectors.
Step 3. The matrix P −1 AP will then be diagonal with λ1 , λ2 , . . . , λn as its
diagonal entries, where λi is the eigenvalue corresponding to pi , 1 ≤ i ≤ n.
190 EIGENVALUES AND EIGENVECTORS
Example 21.1
Find a matrix P that diagonalizes
3 −2 0
A = −2 3 0 .
0 0 5
Solution.
By Problem 20.12, the eigenspaces corresponding to the eigenvalues λ = 1
and λ = 5 are
s 1
1
V = s : s ∈ R = span 1
0 0
and
−t
5
V = t : s, t ∈ R
s
−1 0
= t 1 + s 0 : s, t ∈ R
0 1
−1 0
=span 1 , 0 .
0 1
Let v1 = [1, 1, 0]T , v2 = [−1, 1, 0]T , and v3 = [0, 0, 1]T . It is easy to verify that
these vectors are linearly independent. The matrices
1 −1 0
P = 1 1 0
0 0 1
and
1 0 0
D = 0 5 0
0 0 5
satisfy AP = P D or D = P −1 AP
21 DIAGONALIZATION OF A MATRIX 191
Example 21.2
Show that the matrix
−3 2
A=
−2 1
is not diagonalizable.
Solution.
The characteristic equation of the matrix A is
λ + 3 −2
= 0.
2 λ − 1
(λ + 1)2 = 0.
Theorem 21.2
If A is an n × n matrix with n distinct eigenvalues then A is diagonalizable.
Theorem 21.3
If v1 , v2 , . . . , vn are nonzero eigenvectors that correspond to distinct eigenval-
ues λ1 , λ2 , . . . , λn then the set {v1 , v2 , . . . , vn } is linearly indepedent.
Proof.
The proof is by induction on n. If n = 1 then {v1 } is linearly independent
(Example 18.6). So assume that the vectors {v1 , v2 , · · · , vn−1 } are linearly
independent. Suppose that
α1 v1 + α2 v2 + · · · αn vn = 0. (21.1)
Apply A to both sides of (21.1) and using the fact that Avi = λi vi to obtain
α1 λ1 v1 + α2 λ2 v2 + · · · αn λn vn = 0. (21.2)
Now, multiplying (21.1) by λn and subtracting the resulting equation from
(21.2) we obtain
α1 (λ1 − λn )v1 + α2 (λ2 − λn )v2 + · · · αn−1 (λn−1 − λn )vn−1 = 0. (21.3)
By the induction hypothesis, all the coefficients must be zero. Since the λi are
distinct, i.e. λi − λn 6= 0 for i 6= n, we arrive at α1 = α2 = · · · = αn−1 = 0.
Substituting this into (21.1) to obtain αn vn = 0, and hence αn = 0. This
shows that {v1 , v2 , · · · , vn } is linearly indepedent
Example 21.3
Show that the following matrix is diagonalizable.
1 0 0
A = 1 2 −3
1 −1 0
Solution.
The characteristic equation of the matrix A is
λ − 1 0 0
−1 λ − 2 3 = 0.
−1 1 λ
Expanding the determinant along the first row and simplifying we obtain
(λ − 1)(λ − 3)(λ + 1) = 0.
The eigenvalues are 1, 3 and −1, so A is diagonalizable by Theorem 21.1
The converse of Theorem 21.2 is false as illustrated in the next example.
21 DIAGONALIZATION OF A MATRIX 193
Example 21.4
(a) Find a matrix P that diagonalizes
0 0 −2
A = 1 2 1
1 0 3
Solution.
(a) By Problem 20.11, the eigenspaces corresponding to the eigenvalues λ = 1
and λ = 2 are
−2
1
V = span 1
1
and
−1 0
2
V = span 0 , 1
1 0
Let v1 = [−2, 1, 1]T , v2 = [−1, 0, 1]T , and v3 = [0, 1, 0]T . It is easy to verify
that these vectors are linearly independent. The matrices
−2 −1 0
P = 1 0 1
1 1 0
and
1 0 0
D = 0 2 0
0 0 2
satisfy AP = P D or D = P −1 AP.
(b) Since dim(R3 ) = 3 and the eigenvectors v1 , v2 , and v3 are linearly inde-
pendent, Theorem 18.3(i), {v1 , v2 , v3 } is a basis of R3
194 EIGENVALUES AND EIGENVECTORS
Practice Problems
Problem 21.1
Recall that a matrix A is similar to a matrix B if and only if there is an
invertible matrix P such that P −1 AP = B. In symbol, we write A ∼ B.
Show that if A ∼ B then
(a) AT ∼ B T .
(b) A−1 ∼ B −1 .
Problem 21.2
If A is invertible show that AB ∼ BA for all B.
Problem 21.3
Show that the matrix A is diagonalizable.
0 1 0
A = 0 0 1
4 −17 8
Problem 21.4
Show that the matrix A is not diagonalizable.
2 1 1
A = 2 1 −2
−1 0 −2
Problem 21.5
Show that the matrix
3 0
A=
0 3
is diagonalizable with only one eigenvalue.
Problem 21.6
Show that A is diagonalizable if and only if AT is diagonalizable.
Problem 21.7
Show that if A and B are similar then A is diagonalizable if and only if B is
diagonalizable.
21 DIAGONALIZATION OF A MATRIX 195
Problem 21.8
Give an example of two diagonalizable matrices A and B such that A + B is
not diagonalizable.
Problem 21.9
Find P and D such that P −1 AP = D where
1 1 −1
A = 0 2 −1
0 0 1
Problem 21.10
Find P and D such that P −1 AP = D where
−1 1 1 −2
−1 1 3 2
A=
1 1 −1 −2
0 −1 −1 1
Problem 21.11
Is the matrix
6 3 −8
A = 0 −2 0
1 0 −3
diagonalizable?
Problem 21.12
Is the matrix
2 −2 1
A = −1 3 1
2 −4 3
diagonalizable?
Problem 21.13
The matrix A is a 3×3 matrix with eigenvalues 1, 2, and −1, and eigenvectors
[1, 1, 0]T , [1, 2, 1]T , and [0, 1, 2]T . Find A.
Problem 21.14
1 1 5 0
Let A be a diagonalizable 2 × 2 matrix with P = and D = .
1 2 0 4
Find A5 .
196 EIGENVALUES AND EIGENVECTORS
Problem 21.15
Is the matrix
2 4 6
A = 0 2 2
0 0 4
diagonalizable?
Problem 21.16
Show that the eigenvectors p1 , p2 , · · · , pn of Theorem 21.1 are linearly inde-
pendent.
Problem 21.17
Given the diagonalized matrix
" 1 # " #
− √2 √1 − √12 √1
1 2 2 −1 0 2
A= = √1 √1 √1 √1
.
2 1 2 2
0 3 2 2
Find A10 .
Problem 21.18
Why is
2 0 0
A = 2 6 0
3 2 1
diagonalizable?
Problem 21.19
An n × n matrix A is said to be orthogonally diagonalizable if and only
if there exist an orthogonal matrix P (i.e., P T = P −1 ) and a diagonal ma-
trix D such that A = P DP T . Show that an orthogonally diagonalizable is
symmetric.
Problem 21.20
Show that the matrix 1 1
A= 2 2
1 1
2 2
is diagonalizable.
Linear Transformations
197
198 LINEAR TRANSFORMATIONS
Example 22.2
Let T : R2 → R3 be given by
x
x
T = y .
y
1
Solution.
We show that the first condition of the definition is violated. Indeed, for any
two vectors [x1 , y1 ]T and [x2 , y2 ]T we have
x1 + x2
x1 x x1 + x2
T + 2 =T = y1 + y2
y1 y2 y1 + y2
1
x1 x2
x1 x2
6= y1 + y2 = T
+T .
y1 y2
1 1
The next theorem collects four useful properties of all linear transformations.
Theorem 22.1
If T : V → W is a linear transformation then
(a) T (u − w) = T (u) − T (w)
(b) T (0) = 0
(c) T (−u) = −T (u)
(d) T (α1 u1 + α2 u2 + · · · + αn un ) = α1 T (u1 ) + α2 T (u2 ) + · · · + αn T (un ).
Proof.
See Problem 22.16
Theorem 22.2
A function T : V → W is linear if and only if T (αu + v) = αT (u) + T (v) for
all u, v ∈ V and α ∈ R.
Proof.
See Problem 22.17
Example 22.3
Let Mmn denote the vector space of all m × n matrices.
(a) Show that T : Mmn → Mnm defined by T (A) = AT is a linear transfor-
mation.
(b) Show that T : Mnn → R defined by T (A) = tr(A) is a linear functional.
Solution.
(a) For any A, B ∈ Mmn and α ∈ R we find T (αA + B) = (αA + B)T =
αAT + B T = αT (A) + T (B). Hence, T is a linear transformation.
(b) For any A, B ∈ Mnn and α ∈ R we have T (αA + b) = tr(αA + B) =
αtr(A) + tr(B) = αT (A) + T (B) so T is a linear functional
Example 22.4
Let S : V → W and T : V → W be two linear transformations. Show the
following:
(a) S + T and S − T are linear transformations.
(b) αT is a linear transformation where α denotes a scalar.
Solution.
(a) Let u, v ∈ V and α ∈ R then
(S ± T )(αu + v) =S(αu + v) ± T (αu + v)
=αS(u) + S(v) ± (αT (u) + T (v))
=α(S(u) ± T (u)) + (S(v) ± T (v))
=α(S ± T )(u) + (S ± T )(v).
(b) Let u, v ∈ V and β ∈ R then
(αT )(βu + v) =(αT )(βu) + (αT )(v)
=αβT (u) + αT (v)
=β(αT (u)) + αT (v)
=β(αT )(u) + (αT )(v).
22 LINEAR TRANSFORMATION: DEFINITION AND ELEMENTARY PROPERTIES201
Theorem 22.3
Let V = span{v1 , v2 , · · · , vn }. If T and S are two linear transformations from
V into a vector space W such that T (vi ) = S(vi ) for each i then T = S.
Proof.
See Problem 22.18
The following very useful theorem tells us that once we say what a linear
transformation does to a basis for V, then we have completely specified T.
Theorem 22.4
Let V be an n−dimensional vector space with basis {v1 , v2 , · · · , vn }. If T :
V −→ W is a linear transformation then for any v ∈ V , T v is completely
determined by {T v1 , T v2 , · · · , T vn }.
Proof.
Indeed, if v ∈ V then v = α1 v1 + α2 v2 + · · · + αn vn . Thus, T v = α1 T (v1 ) +
α2 T (v2 ) + · · · + αn T (vn ). Hence, if the vectors T (v1 ), T (v2 ), · · · , T (vn ) are
known then T (v) is known
Example 22.5
Let V and W be n−dimensional vector spaces and {v1 , v2 , · · · , vn } be a basis
of V. Given any vectors w1 , w2 , · · · , wn in W , there exists a unique linear
transformation T : V −→ W such that T (vi ) = wi for each i.
Solution.
Let v ∈ V. Then v = α1 v1 + α2 v2 + · · · + αn vn . Hence, we can define the
transformation T : V → W by
Example 22.6
Let T : Rn → Rm be a linear transformation. Show that there exists an
m × n matrix A such that T (x) = Ax for all x ∈ Rn . The matrix A is called
the standard matrix of T.
Solution.
Consider the standard basis of Rn , {e1 , e2 , · · · , en } where ei is the vector
with 1 at the ith component and 0 otherwise. Let x = [x1 , x2 , · · · , xn ]T ∈ Rn .
Then x = x1 e1 + x2 e2 + · · · + xn en . Thus,
Example 22.7
Find the standard matrix of T : R3 → R2 defined by
x
x − 2y + z
T y = .
x−z
z
Solution.
We have
1 −2 1
T (e1 ) = , T (e2 ) = , T (e3 ) =
1 0 −1
and
1 −2 1
A=
1 0 −1
22 LINEAR TRANSFORMATION: DEFINITION AND ELEMENTARY PROPERTIES203
Practice Problems
Problem 22.1
Let A be an m × n and let T : Rn → Rm be defined by T (x) = Ax where
x ∈ Rn . Show that T is a linear transformation.
Problem 22.2
(a) Show that the identity transformation defined by I(v) = v for all v ∈ V
is a linear transformation.
(b) Show that the zero transformation is linear.
Problem 22.3
Let {v1 , v2 , . . . , vn } be a basis for a vector space V and let T : V → W be a
linear transformation. Show that if T (v1 ) = T (v2 ) = · · · = T (vn ) = 0 then
T (v) = 0 for any vector v in V.
Problem 22.4
Let T : R2 → R3 be a linear transformation such that T (e1 ) = [1, 1, 1]T , T (e2 ) =
[0, 1, −1]T , where {e1 , e2 } is the standard basis of R2 . Find T (x) for any
x ∈ R2 .
Problem 22.5
Consider the matrix
0 1
E= .
1 0
Show that the transformation
x y
TE =
y x
is linear. This transformation is a reflection about the line y = x.
Problem 22.6
Consider the matrix
α 0
F = .
0 1
Show that
x αx
TF =
y y
is linear. Such a transformation is called an expansion if α > 1 and a
compression if α < 1.
204 LINEAR TRANSFORMATIONS
Problem 22.7
Consider the matrix
1 1
G=
0 1
Show that
x αx + y
TG =
y y
is linear. This transformation is called a shear.
Problem 22.8
Show that the function T : R2 → R3 defined by
x+y
x
T = x − 2y
y
3x
is a linear transformation.
Problem 22.9
(a) Show that D : Pn −→ Pn−1 given by D(p) =Rp0 is a linear transformation.
x
(b) Show that I : Pn −→ Pn+1 given by I(p) = 0 p(t)dt is a linear transfor-
mation.
Problem 22.10
3
If T : R3 −→ R is a linear transformation with T −1 = 5 and
2
1 −1
T 0 = 2. Find T 1 .
1 0
Problem 22.11
Let T : R3 −→ R2 be the transformation
x
x
T y = .
y
z
Problem 22.12
Show that the following transformation is not linear: T : Mnn −→ R where
T (A) = |A|.
Problem 22.13
If T1 : U −→ V and T2 : V −→ W are linear transformations, then the
composition T2 ◦ T1 : U −→ W is also a linear transformation.
Problem 22.14
Let T be a linear transformation on a vector space V such that T (v−3v1 ) = w
and T (2v − v1 ) = w1 . Find T (v) and T (v1 ) in terms of w and w1 .
Problem 22.15
Let T : V −→ W be a linear transformation. Show that if the vectors
T (v1 ), T (v2 ), · · · , T (vn )
are linearly independent then the vectors v1 , v2 , · · · , vn are also linearly in-
dependent.
Problem 22.16
Prove Theorem 22.1.
Problem 22.17
Prove Theorem 22.2.
Problem 22.18
Prove Theorem 22.3.
Problem 22.19
Define T : P2 → R3 , where P2 is the vector space of all polynomials of degree
at most 2, by
p(−1)
T (p) = p(0) .
p(1)
(a) Find the image of p(t) = 5 + 3t.
(b) Show that T is a linear transformation.
Problem 22.20
Let T : V → W be a linear transformation.
(a) Show that the set Ker(T ) = {v ∈ V : T (v) = 0} is a subspace of V.
(b) Show that the set Im(T ) = {w ∈ W : T (v) = w for some v ∈ V } is a
subspace of W.
206 LINEAR TRANSFORMATIONS
s
The general solution is given by s and therefore
0
1
Ker(T ) = span 1 .
0
Example 23.2
Let T : Rn → Rm be given by T x = Ax. Find Ker(T ) and Im(T ).
Solution.
We have
Ker(T ) = {x ∈ Rn : Ax = 0}
and
Im(T ) = {Ax : x ∈ Rn }
208 LINEAR TRANSFORMATIONS
Example 23.3
Let V be any vector space and α be a scalar. Let T : V → V be the
transformation defined by T (v) = αv.
(a) Show that T is linear.
(b) What is the kernel of T ?
(c) What is the range of T ?
Solution.
(a) Let u, v ∈ V and β ∈ R. Then T (βu + v) = α(βu + v) = αβu + αv =
βT (u) + T (v). Hence, T is linear
(b) If v ∈ Ker(T ) then 0 = T (v) = αv. If α = 0 then T is the zero transfor-
mation and Ker(T ) = V. If α 6= 0 then Ker(T ) = {0}.
(c) If α = 0 then Im(T ) = {0}. If α 6= 0 then Im(T ) = V since T ( α1 v) = v
for all v ∈ V
Since the kernel and the range of a linear transformation are subspaces of
given vector spaces, we may speak of their dimensions. The dimension of the
kernel is called the nullity of T (denoted nullity(T )) and the dimension of
the range of T is called the rank of T (denoted rank(T )).
The following important result is called the dimension theorem.
Theorem 23.2
If T : V → W is a linear transformation with dim(V ) = n, then
nullity(T ) + rank(T ) = n.
Proof.
See Problem 23.17
Example 23.4
Let T : R2 −→ R3 be given by
x1
x1
T = x1 + x2 .
x2
x2
Solution.
x1 y
(a) Let and 1 be two vectors in R2 . Then for any α ∈ R we have
x2 y2
x1 y1 αx1 + y1
T α + =T
x2 y2 αx2 + y2
αx1 + y1
= αx1 + y1 + αx2 + y2
αx2 + y2
αx1 y1
= α(x1 + x2 ) + y1 + y2
αx2 y2
x1 y1
=αT +T .
x2 y2
0 x1
x x1
(b) Let 1 ∈ Ker(T ). Then 0 = T = x1 + x2 and this leads
x2 x2
0 x2
0
to Ker(T ) = . Hence, nullity(T ) = dim(Ker(T )) = 0.
0
Now,
x1
Im(T ) = T : x1 , x2 ∈ R
x2
x1
= x1 + x2 : x1 , x2 ∈ R
x2
1 0
=span 1 , 1 .
0 1
Practice Problems
Problem 23.1
Let Mnn be the vector space of all n × n matrices. Let T : Mnn → Mnn be
given by T (A) = A − AT .
(a) Show that T is linear.
(b) Find ker(T ) and Im(T ).
Problem 23.2
Consider the linear transformation T : R3 → R3 defined by
x1 x1
1 1 1
T x2 = x2 .
1 2 3
x3 x3
Find a basis for Ker(T ).
Problem 23.3
Consider the linear transformation
T : M22 → M22 defined by T (X) =
0 2
AX − XA, where A = . Find the rank and nullity of T.
3 3
Problem 23.4
Consider the linear transformation T : R2 → R2 defined by
x1 0
T = .
x2 x1
Find Ker(T ) and Im(T ).
Problem 23.5
Let T : R3 → R5 be a linear transformation such that its kernel is of dimen-
sion 2. How many vectors needed to span the range of T ?
Problem 23.6
The nullity of a m × n matrix A is defined to be the nullity of the linear
transformation T : Rn → Rm with T (x) = Ax. Find the value(s) of a so that
the nullity of the matrix
0 1 1
1 a 1
0 2 a
is zero.
23 KERNEL AND RANGE OF A LINEAR TRANSFORMATION 211
Problem 23.7
Consider the linear transformation T : R3 → R3 defined by
x1 x1
T x2 = x2 .
x3 0
Find a basis for Ker(T ) and Im(T ).
Problem 23.8
Consider the linear transformation T : R2 → R3 defined by
x1
x1
T = x1 + x2 .
x2
x1 + 2x2
Find a basis for Ker(T ) and Im(T ).
Problem 23.9
Consider the linear transformation T : P2 → P3 defined by
Problem 23.10
Suppose that S : P12 → P15 and T : P15 → P12 are linear transformations.
(a) If T has rank 6, what is the nullity of T ?
(b) If S has nullity 6, what is the rank of S?
Problem 23.11
Consider the linear transformation T : P3 → R2 defined by T (p) = [p(0), p(1)]T .
Describe the kernel and range of T.
Problem 23.12
Consider the linear transformation T : Pn → Pn−1 defined by T (p) = p0 (t).
Find the nullity and rank of T.
Problem 23.13
Let V be the vector space of twice differentiable function over an interval I
and F(I) the vector space of functions defined on I. Define T : V → F(I)
by T (y) = y 00 + y. Show that T is linear and find its kernel.
212 LINEAR TRANSFORMATIONS
Problem 23.14
Let V be a finite dimensional vector space and T : V → W be a linear
transformation. Show that any basis of Ker(T ) can be extended to a basis
of V.
Problem 23.15
Let T : V → W be a linear transformation and dim(V ) = n. Suppose
that {v1 , v2 , · · · , vk } is a basis of Ker(T ). By the previous problem, we can
extend this basis to a basis {v1 , v2 , · · · , vk , vk+1 , · · · , vn } of V. Show that
Im(T ) = span{T (vk+1 ), · · · , T (vn )}.
Problem 23.16
Show that the set {T (vk+1 ), · · · , T (vn )} in Problem 23.15 is linearly indepen-
dent.
Problem 23.17
Using Problems 23.15 - 23.16, prove Theorem 23.2.
Problem 23.18
Let T : V → W and S : W → X be two linear transformations. Prove that
(a) Ker(T ) ⊆ Ket(S ◦ T ) and (b) Im(S ◦ T ) ⊆ Im(S).
Problem 23.19
Find the kernel and range of the linear transformation T : P2 → P2 defined
by T (p) = xp0 (x).
Problem 23.20
Let T : V → W be a linear transformation with the property that no two
distinct vectors of V have the same image in W. Show that Ker(T ) = {0}.
24 ISOMORPHISMS 213
24 Isomorphisms
Since linear transformations are functions, it makes sense to talk about one-
to-one and onto functions. We say that a linear transformation T : V → W
is one-to-one if T (v) = T (w) implies v = w. We say that T is onto if
Im(T ) = W. If T is both one-to-one and onto we say that T is an isomor-
phism and the vector spaces V and W are said to be isomorphic and we
write V ∼= W. The identity transformation is an isomorphism of any vector
space onto itself. That is, if V is a vector space then V ∼
= V.
The following theorem is used as a criterion for proving that a linear trans-
formation is one-to-one.
Theorem 24.1
Let T : V → W be a linear transformation. Then T is one-to-one if and only
if Ker(T ) = {0}.
Example 24.1
x1
3 2 x 1 + x 2
Consider the transformation T : R → R given by T x2 = .
x1 − x2
x3
(a) Show that T is linear.
(b) Show that T is onto but not one-to-one.
Solution.
x1 y1
(a) Let x2 and y2 be two vectors in R3 and α ∈ R. Then
x3 y3
x1 y1 αx1 + y1
T α x2 + y2 =T αx2 + y2
x3 y3 αx3 + y3
αx1 + y1 + αx2 + y2
=
αx1 + y1 − αx2 − y2
α(x1 + x2 ) y1 + y2
= +
α(x1 − x2 ) y1 − y2
x1 y1
=αT x2 + T y2 .
x3 y3
214 LINEAR TRANSFORMATIONS
0
(b) Since 0 ∈ Ker(T ), by Theorem 24.1 T is not one-to-one. Now, let
1
u u
v ∈ R3 be such that T v = x1 . In this case, x1 = 1 (u + v) and
x2 2
w w
x2 = 21 (u − v). Hence, Im(T ) = R2 so that T is onto
Example 24.2
Let T : V → W be a one-to-one linear transformation. Show that if
{v1 , v2 , . . . , vn } is a basis for V then {T (v1 ), T (v2 ), . . . , T (vn )} is a basis for
Im(T ).
Solution.
The fact that {T (v1 ), T (v2 ), · · · , T (vn )} is linearly independent follows from
Problem 24.6. It remains to show that Im(T ) = span{T (v1 ), T (v2 ), · · · , T (vn )}.
Indeed, let w ∈ Im(T ). Then there exists v ∈ V such that T (v) = w.
Since {v1 , v2 , · · · , vn } is a basis of V, v can be written uniquely in the form
v = αv1 + α2 v2 + · · · + αn vn . Hence, w = T (v) = α1 T (v1 ) + α2 T (v2 ) + · · · +
αn T (vn ). That is, w ∈ span{T (v1 ), T (v2 ), · · · , T (vn )}. We conclude that
{T (v1 ), T (v2 ), · · · , T (vn )} is a basis of Im(T )
Theorem 24.2
If W is a subspace of a finite-dimensional vector space V and dim(W ) =
dim(V ) then V = W.
Proof.
See Problem 18.21
Theorem 24.3
Let T : V → W be a linear transformation such that dim(V ) = dim(W ) = n.
Then
(a) if T is one-to-one, then T is onto;
(b) if T is onto, then T is one-to-one.
24 ISOMORPHISMS 215
Proof.
(a) If T is one-to-one then Ker(T ) = {0}. Thus, dim(Ker(T )) = 0. By
Theorem 23.2 we have dim(Im(T )) = n. Hence, by Theoren 24.2, we have
Im(T ) = W. That is, T is onto.
(b) If T is onto then dim(Im(T )) = dim(W ) = n. By Theorem 23.2, dim(Ker(T )) =
0. Hence, Ker(T ) = {0}, i.e., T is one-to-one
Theorem 24.4
Let T : V → W be an invertible linear transformation. Then
(a) T −1 is linear.
(b) (T −1 )−1 = T.
Proof.
(a) Suppose T −1 (w1 ) = v1 , T −1 (w2 ) = v2 and α ∈ R. Then αw1 + w2 =
αT (v1 ) + T (v2 ) = T (αv1 + v2 ). That is, T −1 (αw1 + w2 ) = αv1 + v2 =
αT −1 (w1 ) + T −1 (w2 ).
(b) Since T ◦ T −1 = idW and T −1 ◦ T = idV , T is the inverse of T −1 . That
is, (T −1 )−1 = T
Theorem 24.5
A linear transformation T : V → W is invertible if and only if T is one-to-one
and onto.
Example 24.3
Let T : R3 → R3 be given by T (x) = Ax where A is the matrix
1 1 1
A = 0 1 2
1 2 2
Solution.
x1
(a) We must show that T is one-to-one and onto. Let x = x2 ∈ Ker(T ).
x3
0
Then T x = Ax = 0 . Since |A| = −1 6= 0, A is invertible and therefore
0
0 0
x = 0 . Hence, Ker(T ) = 0 . Now, since A is invertible the system
0 0
Ax = b is always solvable. This shows that Im(T ) = R3 . Hence, by the above
theorem, T is invertible.
(b) T −1 x = A−1 x
24 ISOMORPHISMS 217
Practice Problems
Problem 24.1
Let T : Mmn → Mmn be given by T (X) = AX for all X ∈ Mmn , where A is
an m × m invertible matrix. Show that T is both one-one and onto.
Problem 24.2
x1
3 2
Show that the projection transformation T : R → R defined by T x2 =
x3
x1
is not one-one.
x2
Problem 24.3
Let T : V → W. Prove that T is one-one if and only if dim(Im(T )) = dim(V ).
Problem 24.4
Show that the linear transformation T : Mnn → Mnn given by T (A) = AT is
an isomorphism.
Problem 24.5
Consider the linear transformation T : R2 → R3 defined by
x1 + x2
x1
T = x1 + 2x2 .
x2
x2
Problem 24.6
Let T : V → W be a linear transformation. Suppose that T is one-to-one.
Show that if S is a finite linearly independent set of vectors then T (S) is also
linearly independent.
Problem 24.7
Let T : V → W be a linear transformation. Suppose that T (S) is linearly
independent set of vectors whenever S is a set of linearly independent vectors.
Show that T is one-to-one.
218 LINEAR TRANSFORMATIONS
Problem 24.8
x1 + x2
x1
Consider the transformation T : R2 → R3 given by T = x1 − x2 .
x2
x1
(a) Show that T is linear.
(b) Show that T is one-to-one but not onto.
Problem 24.9
Prove Theorem 24.1.
Problem 24.10
Prove Theorem 24.5.
Problem 24.11
2 3 x1
Consider the linear transformation T : R → R defined by T =
x2
5x1 − 3x2
−7x1 + 8x2 .
2x1
(a) Find the standard matrix of T.
(b) Find Ker(T ). Is T one-to-one?
Problem 24.12
Consider the linear transformation T : P2 → P1 defined by T (p) = p0 . Show
that T is onto but not one-to-one.
Problem 24.13
Let T : Rn → Rn be the linear transformation defined by T (x) = Ax with A
singular. Show that T is neither one-to-one nor onto.
Problem 24.14
x1
x2 =
Consider the linear transformation T : R4 → R4 defined by T
x3
x4
0
x1 + x2
x2 + x3 . Determine whether T is one-to-one and onto.
x3 + x4
24 ISOMORPHISMS 219
Problem 24.15
2 2
Find
the inverse
linear transformation T : R → R defined by
of the
x1 x + 3x2
T = 1 .
x2 x1 + 4x2
Problem 24.16
Let T : V → W and S : W → X be two one-to-one linear transformations.
Show that S ◦ T is also one-to-one.
Problem 24.17
Let T : R4 → M22 be the linear transfomation defined by
x1
x2 x1 x2
T =
.
x3 x3 x4
x4
Find T −1 .
Problem 24.18
Let T : V → W and S : W → X be two onto linear transformations. Show
that S ◦ T is also onto.
Problem 24.19
Consider the linear transformation T : M22 → M22 defined by
2 0 4 0
T (A) = A−A .
0 3 0 5
Problem 24.20
Suppose that T : V → W is an isomorphism where V and W are finite
dimensional. Show that dim(V ) = dim(W ).
220 LINEAR TRANSFORMATIONS
Answer Key
Section 1
1.4 (a) The two lines intersect at the point (3, 4) so the system is consis-
tent.
(b) The two equations represent the same line. Hence, x2 = s is a parameter.
Solving for x1 we find x1 = 5+3t
2
.The system is consistent.
(c) The two lines are parallel. So the given system is inconsistent.
221
222 ANSWER KEY
1.6 x1 = 1 + 5w − 3t − 2s, x2 = w, x3 = t, x4 = s.
1.7 (a)
x1 + x2 + 2x3 = 9
x1 + x2 + 2x3 = 2
3x1 + 6x2 − 5x3 = 0.
1.10 x − 2y = 5.
1.11 c = a + b.
1.13 (A) (a) Consistent (b) Independent (c) One solution. (B) (a) Inconsis-
tent (b) Does not apply (c) No solution.
(C) (a) Consistent (b) Dependent (c) Infinitely many solutions.
1.14 (
x1 + x2 = 5800
0.035x1 + 0.055x2 = 283.
1.19 x1 = x2 = 1.
1.20 (a) There is no point that lies on all three of the planes.
(b) The three planes have a single point in common.
(c) The three planes have an infinite number of points in common.
224 ANSWER KEY
Section 2
2.2 A = − 19 , B = − 59 , and C = 11
9
.
3
2.4 x1 = 2
, x2 = 1, x3 = − 25 .
2.5 x1 = 19 , x2 = 10
9
, x3 = − 73 .
2.7 a = 1, b = 2, c = −1.
2.9 x1 = 2, x2 = 1, x3 .
2.10 x1 = 2, x2 = −1, x3 = 1.
2.12
1.50x1 + 5.75x2 + 2.60x3 = 589.50
x1 + x2 + x3 = 200
x2 − x3 = −20.
2.13 (a) Swap rows 1 and 3. (b) This is not a valid row operation. You can’t
add or subtract a number from the elements in a row. (c) Add four times
row 2 to row 3. (d) Multiply row 4 by 5.
225
2.14 x1 = − 32 t + 1, x2 = − 32 t + 12 , x3 = t − 12 , x4 = t.
2.15 a = 3.
2.16 (a) True. The row reduced augmented matrix leads to a linear sys-
tem that is equivalent to the original system.
(b) False. They are row equivalent if you can get from one to the other using
elementary row operations.
(c) True by the definition of equivalent systems.
2.17 (a) The system is consistent and dependent if h = −2. In this case,
the parametric equations are: x1 = 2t − 3, x2 = t.
(b) The system is consistent and independent if h 6= −2. The unique solution
is x1 = −3, x2 = 0.
2.18 d − 3c 6= 0.
2.23 If two equations in a system are swapped then the new system is triv-
ially equivalent to the original system since it is just a rearrangement of the
equations in the old system.
If an equation in the original system is multiplied by a non-zero scalar then
the new system is equivalent to the original system by Problem 2.21.
If a multiple of an equation of a system is added to another equation then
the new system is equivalent to the original system by Problem 2.22.
227
Section 3
3.1 x1 = 2, x2 = −1, x3 = 1.
3.2 −5g + 4h + k = 0.
3.3 x1 = − 11
2
. x2 = −6, x3 = − 25 .
3.4 x1 = 19 , x2 = 10
9
, x3 = − 37 .
3.7 x1 = 3, x2 = 1, x3 = 2.
3.10
−1 − 1 0 0
0
−4 −2 1
0 0 2 3
0 0 0 0
3.12 a = −8.
3.13 x1 = 5 − t − 2s, x2 = 1 − t, x3 = s, x4 = t.
(c) True. This rows shows that ox1 + 0x2 + 0x3 + 0x4 = 0 and this equation
has infinite number of solutions.
(d) False. Any matrix can be reduced. The matrix does not know where it
came from.
3.17 (a) If k = 0 then the last row gives 0x1 + 0x2 + 0x3 = −4 which
does not have a solution. Hence, the system is inconsistent.
(b) If k 6= 0, the system has the unique solution x1 = 24−3k
k
, x2 = − 12
k
, x3 =
4
−k.
3.18 198.
3.19 x5 = t, x4 = 3t − 1, x3 = 0, x2 = s, x1 = 1 − 2s.
3.20 Inconsistent.
229
Section 4
4.1
1 −2 3 1 −3
.
0 1 −1 −1 2
4.2
1 0 −2 3
0 1 −7 9 .
0 0 1 −1
4.3
1 −1 −3 8
0 1 2 −4 .
0 0 1 − 52
4.4
1 2 1 0
0 1 1 1
.
3 3
7
0 0 1 −3
4.5 (a) No, because the matrix fails condition 1 of the definition. Rows of
zeros must be at the bottom of the matrix.
(b) No, because the matrix fails condition 2 of the definition. Leading entry
in row 2 must be 1 and not 2.
(c) Yes. The given matrix satisfies conditions 1 - 4.
4.6
1 −3 1 −1 0 −1
0 0 1 2 1 2
.
0 0 0 0 1 −1
0 0 0 0 0 0
4.7
1 0 0 0
0 1 0 0
.
0 0 1 0
0 0 0 1
230 ANSWER KEY
4.8
1 0 3 0 4
0
1 −2 0 5
.
0 0 0 1 −2
0 0 0 0 0
4.9
1 1 0 0
0
1 .5 −.25.
0 0 1 1.5
0 0 0 0
4.10
1 1 2 8
0 1 −5 −9 .
0 0 1 2
4.12
1 0 0 0
0 1 0 1
0 0 1 2
4.13
1 0 0 2
0 1 0 0 .
0 0 1 −1
4.14 3.
4.15 3.
4.16
1 0 0 3
0 1 0 4
0 0 1 −2
231
4.17
1 0 −1 0
0 1 −1 0
0 0 0 1
4.18
1 0 − 45 0
0 1 1 1
4 2
0 0 0 0
Section 5
5.1 x1 = 3, x2 = 1, x3 = 2.
5.5 x1 = 19 , x2 = 10
9
, x3 = − 73 .
5.6 x1 = − 11
2
, x2 = −6, x3 = − 52 .
5.7 x1 = 1, x2 = −2, x3 = 1, x4 = 3.
5.8 x1 = 2, x2 = 1, x3 = −1.
5.11 x1 = 3, x2 = 4, x3 = −2.
5.12 Inconsistent.
5.13 x1 = 74 t, x2 = 1
2
− 41 t, x3 = t.
5.14 x1 = 2s + t, x2 = s, x3 = 2t, x4 = t.
5.15 x1 = −8, x2 = 1, x3 = 5.
5.16 x1 = 3, x2 = 4, x3 = −2.
5.18 x1 = 74 t, x2 = 1
2
− 41 t, x3 = t.
233
5.19 x1 = −2 − s, x2 = s, x3 = t.
5.21 Suppose first that the reduced augmented matrix has a row of the
form (0, · · · , 0, b) with b 6= 0. That is, 0x1 + 0x2 + ... + 0xm = b. Then the
left side is 0 whereas the right side is not. This cannot happen. Hence, the
system has no solutions.
5.22 If the reduced augmented matrix has independent variables and no rows
of the form (0, 0, · · · , 0, b) for some b 6= 0 then these variables are treated as
parameters and hence the system has infinitely many solutions.
5.23 If the reduced augmented matrix has no row of the form (0, 0, · · · , 0, b)
where b 6= 0 and no indepedent variables then the system looks like
x 1 = c1
x 2 = c2
x 3 = c3
.. ..
.=.
x m = cm
Section 6
6.1 (a) This is true since the trivial solution is always a solution.
(b) The system will have free variables, so it either has infinitely many so-
lutions or no solutions. However, by part (a), it always has at least one
solution. Therefore, such a system will always have infinitely many solutions.
6.3 x1 = −s, x2 = s, x3 = s, x4 = 0.
6.4 x1 = x2 = x3 = 0.
6.6 x1 = −s + 3t, x2 = s, x3 = t.
6.7 x1 = − 73 t, x2 = − 32 t, x3 = − 13
3
t, x4 = t.
6.9 x1 = x2 = x3 = 0.
6.10 x1 = − 12 t, x2 = 32 t, x3 = x4 = t.
6.12 x1 = t − 32 s, x2 = t, x3 = 13 s, x4 = s.
6.14 The rank of the augmented matrix is 3 which is less than the num-
ber of unknowns. Hence, by Theorem 6.1(a), the system has infinitely many
solutions.
6.15 The augmented matrix of the first system is row equivalent to the
augmented matrix of the second system.
235
6.16 (a) True. The trivial solution is a common solution to all homoge-
neous linear systems.
(b) False. See Example 6.3.
(c) The system will have free variables, so it either has infinitely many so-
lutions or no solutions. However, by part (a), it always has at least one
solution. Therefore, such a system will always have infinitely many solutions.
and
ai1 y1 + ai2 y2 + · · · + ain yn = 0.
Thus,
ai1 (αx1 + βy1 ) + ai2 (αx2 + βy2 ) + · · · + ain (αxn + βyn ) = α(ai1 x1 + ai2 x2 + · · · + ain xn )
+ β(ai1 y1 + ai2 y2 + · · · + ain yn )
= 0 + 0 = 0.
6.19 Consider the equation x1 − 4x2 + 7x3 = 5 (see Example 1.2). Two solu-
tions to this equation are (5, 0, 0) and (2, 1, 1). However, (5, 0, 0) + (2, 1, 1) =
(7, 1, 1) is not a solution.
6.20 It is possible to have the reduced row echelon form of the augmented
matrix with a row of zeros and no free variables. For example,
1 0 0
0 1 0
0 0 0
It is possible to have free variables and yet the reduced row echelon form of
the augmented matrix has no rows of zeros. See Example 6.2.
236 ANSWER KEY
Section 7
4 −1
7.1
−1 −6
7.2 w = −1, x = −3, y = 0, and z = 5.
7.3 s = 0 and t = 3.
7.4 We have
1 0 0 1 0 0 0 0
a +b +c +d =
0 0 0 0 1 0 0 1
a 0 0 b 0 0 0 0
+ + + =
0 0 0 0 c 0 0 d
a b
= A.
c d
7.6
9 5 1
−4 7 6
237
7.9 4tr(7A) = 0.
7.11
12 2.5 0.5
1
S = (A + AT ) = 2.5 −4 −4
2
0.5 −4 2
and
0 4.5 0.5
1
K = (A − AT ) = −4.5 0 4 .
2
−0.5 −4 0
7.13 x = −4.
7.14
12 −12 21
.
21 12 −6
7.15
1 2 3 4 5
2 3 4 5 1
3 4 5 1 2
.
4 5 1 2 3
5 1 2 3 4
7.16 x1 == 2, x2 = 1, x3 = −1.
α · 0 + β · 0 = 0.
7.21 (i) A + B = [aij ] + [bij ] = [aij + bij ] = [bij + aij ] = [bij ] + [aij ] = B + A,
since addition of scalars is commutative.
(ii) (A + B) + C = ([aij ] + [bij ]) + [cij ] = [aij + bij ] + [cij ] = [(aij + bij ) + cij ] =
[aij + (bij + cij )] = [aij ] + [bij + cij ] = [aij ] + ([bij ] + [cij ]) = A + (B + C) since
addition of scalars is associative.
(iii) A + 0 = [aij ] + [0] = [aij + 0] = [aij ] = A since 0 is the 0 element for
scalars.
(iv) A + (−A) = [aij ] + [−aij ] = [aij + (−aij ] = [0] = 0.
(v) c(A + B) = c([aij + bij ]) = [c(aij + bij )] = [caij + cbij ] = [caij ] + [cbij ] =
c[aij ] + c[bij ] = cA + cB since multiplication of scalars is distributive with
respect to addition.
(vii) (c + d)A = (c + d)[aij ] = [(c + d)aij ] = [caij + daij ] = [caij ] + [daij ] =
c[aij ] + d[aij ] = cA + dA.
(viii) (cd)A = (cd)[aij ] = [(cd)aij ] = [c(daij )] = c[daij ] = c(dA).
239
Section 8
8.1
2x1 − x2
= −1
−3x1 + 2x2 + x3 = 0
x2 + x3 = 3.
8.2 (a) If A is the coefficient matrix and B is the augmented matrix then
2 3 −4 1 2 3 −4 1 5
A = −2 0 1 0 , B = −2 0 1 0 7 .
3 2 −4 0 3 2 −4 0 3
8.5 We have
100 −432
2
(AB) =
0 289
and
2 160 −460
2
AB = .
−5 195
8.8
1 1 −1 −1
, .
1 1 −1 −1
240 ANSWER KEY
8.9 k = −1.
8.10
3x1 − x2 + 2x3 = 2
4x1 + 3x2 + 7x3 = −1
−2x1 + x2 + 5x3 = 4.
8.11 (a) a = 0 (b) The system has exactly one solution if a 6= 0 (c)
x1 = −15, x2 = 6, and x3 = 2.
8.14
1 −8 −2
−4 11 4 .
−8 16 3
8.15 x1 = t − s, x2 = t, x3 = x4 = s.
8.16 AAT = AT A = I2 .
8.17
7 −8
, .
4 −5
1 −2 1 3 0 0
= .
2 −4 2 − 23 0 0
CA = [hij ]. Then
8.25 Let A = [aij ], B = [bij , and AB = [cij ]. Then (AB)T = [cji ]. Let
B T AT = [dij ].Then
Section 9
9.1 (a)
1 0 0 0
A= ,B = .
0 0 0 1
(b)
1 0 −1 0
A= ,B = .
0 1 0 −1
9.3
41 −30
.
−15 11
9.4
5 1
13 13
3 2
− 13 13
9.6
sin θ − cos θ
.
cos θ sin θ
9 1
− 13 13
9.7 A = 2 6 .
− 13
132
−5 1
9.8 A = .
− 15 3
5
9.10 B −1 .
244 ANSWER KEY
9.13 AT = A−1 = A.
9.16 A−1 = AT .
9.17
1 0 0 0
0 1 0 0
2 1
0 0 0
3
0 0 0 14
9.18 x1 = 3 and x2 = 7.
9.19 x1 = − 53 and x2 = − 37 .
Section 10
10.2 (a)
1 0 0 0
0 −2 0 0
.
0 0 1 0
0 0 0 1
(b)
1 0 0 0
0 1 0 0
.
0 0 1 0
0 0 3 1
(c)
0 0 1 0
0 1 0 0
.
1 0 0 0
0 0 0 1
10.3.(a) r1 ↔ r3 , E −1 = E.
(b) r2 ← r2 − 2r1
1 0 0
2 1 0 .
0 0 1
(c) r3 ← 5r3
1 0 0
0 1 0 .
0 0 51
10.4 (a)
0 0 1
E1 = 0 1 0 .
1 0 0
246 ANSWER KEY
(b) E2 = E1 .
(c)
1 0 0
E3 = 0 1 0 .
−2 0 1
(d)
1 0 0
E4 = 0 1 0 .
2 0 1
10.5
0 0 1
0 1 0 .
1 0 0
10.6
r2 ← 21 r2 , r1 ← −r2 + r1 , r2 ↔ r3 .
10.7
a 0 0 1 0 0 1 0 0
0 1 0 , 0 a 0 , 0 0 1 .
0 0 1 0 0 1 0 0 a
10.8
1 a 0 1 0 a 1 0 0 1 0 0 1 0 0 1 0 0
0 1 0 , 0 1 0 , a 1 0 , 0 1 a , 0 1 0 , 0 1 0 .
0 0 1 0 0 1 0 0 1 0 0 1 a 0 1 0 a 1
10.10
0 5 −3
−4 3 0 .
3 0 6
10.11
1 0 0 1 0 0 1 0 −1
0 1 0 0 1 1 0 1 0
0 0 −1 0 0 1 0 0 1
10.12 The third row is a multiple of the first row. If we perform r3 ← r3 −3r1
on A, we will get a row of all zeroes in the third row. Thus, there is no way
we can do elementary row operations on A to get I3 , because the third row
will never have a leading one. So, the reduced row elementary operations of
A is not I3 . So, A cannot be written as a product of elementary matrices.
(b) We have
1 0 −1
(E3 E2 E1 )−1 = E1−1 E2−1 E3−1 = 0 1 1 .
0 0 −1
10.14 (a)
1 0 0 1 0 2 1 0 0
E1 = 0 41 0 , E2 = 0 1 0 , E3 = 0 1 − 34 .
0 0 1 0 0 1 0 0 1
(b)
1 0 0 1 0 −2 1 0 0
A = 0 4 0 0 1 0 0 1 43
0 0 1 0 0 1 0 0 1
248 ANSWER KEY
10.15 (a)
1 0 0 0 1 0 0 0
0 1 0 0
, E1−1 = 0 1 0 0
E1 =
0
0 1 0 0 0 1 0
0 0 2 1 0 0 −2 1
(b)
0 0 0 1 0 0 0 1
0 1 0 0
, E2−1 = 0 1 0 0
E2 =
0
0 1 0 0 0 1 0
1 0 0 0 1 0 0 0
(c)
1 0 0 0 1 0 0 0
0 6 0 0 1
, E3−1 = 0 6 0 0 .
E3 =
0 0 1 0 0 0 1 0
0 0 0 1 0 0 0 1
10.16 (a) is obvious, since In can row reduce a matrix to itself by performing
the identity row operation.
(b) Supose that B = Ek Ek−1 · · · E1 A. Then A = E1−1 E2−1 · · · Ek−1 B. Since the
inverse of an elementary matrix is an elementary matrix, it follows that B
row reduces to A. That is, B ∼ A.
(c) Supose that B = Ek Ek−1 · · · E1 A and C = Fn Fn−1 · · · F1 B. Then C =
Fn Fn−1 · · · F1 Ek Ek−1 · · · E1 A. Hence, A ∼ C.
10.19 Note that EiT = Ei and Ei2 = In for each i. Thus, AAT = E1 E2 · · · Ek EkT Ek−1
T
· · · E1T =
2 T T T T
E1 E2 · · · Ek−1 Ek Ek−1 · · · E1 = E1 E2 · · · Ek−1 Ek−1 · · · E1 = · · · = In .
249
10.20
5
1 0 11
10 .
0 1 11
250 ANSWER KEY
Section 11
11.2 a = −1 or a = 3.
11.3
13
− 12 − 18
8
A−1 = − 15
8
1
2
3
8
.
5
4
0 − 14
11.7 Suppose that A is row equivalent to B. Then there exist elementary ma-
trices E1 , E2 , · · · , Ek such that B = Ek Ek−1 · · · E1 A. Let P = Ek Ek−1 · · · E1 .
Then by Theorem 10.2 and Theorem 9.3(a), P is nonsingular.
Conversely, suppose that B = P A, for some nonsingular matrix P. By The-
orem 11.1, P is row equivalent to In . That is, In = Ek Ek−1 · · · E1 P. Thus,
B = E1−1 E2−1 · · · Ek−1 A and this implies that A is row equivalent to B.
11.12 (a)
23
29 2
7
A−1 = 16 13
2
4
3
4 2
1
(b) x1 = 23 , x2 = 1, x3 = − 52 .
11.16
−3 0 −4
A−1 = −2 1 −3 .
1 0 1
11.19
−4 0 −2
A = 3 −6 −4 .
0 0 −4
11.20 We will show that Bx = 0 has only the trivial solution. Indeed, if x
is any solution of Bx = 0 then x = In x = (AB)x = A(Bx) = A0 = 0. It
follows from Theorem 11.1 that B is invertible and B −1 = A. Also, BA =
BB −1 = In .
253
Section 12
12.2 t = 0, t = 1, or t = 4.
12.3 x = 1 or x2 = 12 .
12.4 |A| = 0.
12.6 λ = ±1.
12.8 −240
12.9 |A| = 6.
12.10 |A| = 1.
12.12 −114.
12.13 t = 1 or t = 2.
12.15 20.
12.16 |A| = 0. Note that the second row is k times the first row.
12.17 λ = −3 or λ = 1.
12.18 λ = 3 or λ = −1.
254 ANSWER KEY
12.19 λ = −8.
12.20 λ = 2 or λ = 1.
255
Section 13
13.2 (a)
d e f
g h i = −6.
a b c
(b)
3a 3b 3c
−d −e −f = 72.
4g 4h 4i
(c)
a + g b + h c + i a b c
d e f = d e f = −6
g h i g h i
(d)
−3a −3b −3c
d e f = 18.
g − 4d h − 4e i − 4f
13.3 The determinant is 0 since the first and the fifth rows are proportional.
13.4 |A| = 34 .
13.11 15.
13.12 −36.
13.13 |cA| = c3 .
13.14 8.
13.15 x = 2.
13.19 48.
The first sum is the determinant of B using cofactor expansion along the k th
row and the second sum is the determinant of C using cofactor expansion
along the k th row. Hence, |A| = |B| + |C|.
problem, we have |B| = |A| + |C| = |A| + c|D|, where D is the matrix A
with two identical rows i and k. By Theorem 13.2, |D| = 0. Hence, |B| = |A|.
r1
r2
..
.
ri + rk
0 =
..
.
r + r
i k
..
.
rn
r1 r1
r2 r2
.. ..
. .
ri + rk ri + rk
= .. +
..
. .
ri rk
.
..
.
..
rn rn
r1 r1 r1 r1
r2 r2 r2 r2
.. .. .. ..
. . . .
ri rk ri rk
=
... + ... + ... + ...
r r r r
i i k k
. . . .
.. .. .. ..
rn rn rn rn
259
r1 r1
r2 r2
.. ..
. .
rk ri
=
... + ... .
r r
i k
. .
.. ..
rn rn
Hence,
r1 r1
r2 r2
.. ..
. .
rk r
. = − i. .
.. ..
r r
i k
. .
.. ..
rn rn
260 ANSWER KEY
Section 14
14.4 Taking the determinant of both sides to obtain |A2 | = |A| or |A|(|A| −
1) = 0. Hence, either A is singular or |A| = 1.
14.6 Finding the determinant we get |A| = 2(c + 2)(c − 3). The determi-
nant is 0 if c = −2 or c = 3.
14.12 The second column of A is twice the first column so that by Ex-
ample 13.4, |A| = 0. Hence, by Theorem 14.1, A is singular.
14.13 (a) |A4 | = |A|4 = 54 = 625 (b) |AT | = |A| = 5 (c) |5A| = 53 |A| = 54 =
625 (d) |A−1 | = 51 .
1
14.14 81
.
14.17 |A| = − 81 .
14.18 |A| = 0.
14.19 14.
Section 15
15.1 (a)
−18 −6 −10
adj(A) = 17 −10 −1 .
−6 −2 28
(b) |A| = −94.
15.2 Suppose first that A is invertible. Then adj(A) = A−1 |A| so that
n
|adj(A)| = ||A|A−1 | = |A|n |A−1 | = |A|
|A|
= |A|n−1 . If A is singular then
adj(A) is singular. Suppose the contrary, then there exists a square matrix
B such that B · adj(A) = adj(A) · B = In . Then A = AIn = A(adj(A)B) =
(A·adj(A))B = |A|B = 0 and this leads to adj(A) = 0, a contradiction to the
fact that adj(A) is non-singular. Thus, adj(A) is singular and consequently
|adj(A)| = 0 = |A|n−1 .
15.3 1 1
−7 0 − 21
adj(A) = |A|A−1 2
= 0 − 21 − 71 .
1 1 1
− 7 − 21 21
3n
15.4 |A−1 + adj(A)| = 2
.
(c)
1 1
− 12
adj(A) 2 2
A−1 = = −5 2 1 .
|A| 7
2
− 2 − 12
3
15.10 (a) If A has integer entries then adj(A) has integer entries. If |A| = 1
then A−1 = adj(A) has integer entries.
(b) Since |A| = 1, A is invertible and x = A−1 b. By (a), A−1 has integer
entries. Since b has integer entries, A−1 b has integer entries.
(c)
3 − 12
3
1
A−1 = − adj(A) = −2 1
2
−1 .
2 1
−2 2
−2
(b) We have
0 1 1
adj(A) = 3 −3 3 .
3 −2 1
(c) Expanding along the third row, we find
(d)
1 1
0 3 3
A−1 = 1 −1 1 .
1 − 23 13
15.14
1 1
−1
adj(A) 2 2
A−1 = = 2 −1 0 .
|A|
2 − 2 − 12
1
15.17 1 3
2
1 2
A−1 = 0 1 32 .
0 0 12
15.18 25.
15.19 125.
265
Section 16
|A1 | 10 |A2 | 18 |A3 | 38
16.1 x1 = |A|
= − 11 , x2 = |A|
= 11
, x3 = |A|
= 11
.
|A1 | |A2 | |A3 |
16.2 x1 = |A|
= − 43 , x2 = |A|
= 83
8
, x3 = |A|
= 21
8
.
|A1 | |A2 | |A3 |
16.3 x1 = |A|
= −1, x2 = |A|
= 3, x3 = |A|
= 2.
|A1 | 212 |A2 | 273 |A3 | 107
16.4 x1 = |A|
= 187
, x2 = |A|
= 187
, x3 = |A|
= 187
.
|A1 | |A2 | |A3 |
16.5 x1 = |A|
= 4, x2 = |A|
= −1, x3 = |A|
= − 13 .
|A1 | |A2 | |A3 |
16.6 x1 = |A|
= 2, x2 = |A|
= −1, x3 = |A|
= 4.
|A1 | |A2 | |A3 |
16.7 x1 = |A|
= 2, x2 = |A|
= −1, x3 = |A|
= 3.
|A1 | |A2 | |A3 |
16.8 x1 = |A|
= 4, x2 = |A|
= 1, x3 = |A|
= −2.
|A1 | |A2 | |A3 |
16.9 x1 = |A|
= 5, x2 = |A|
= 2, x3 = |A|
= 2.
|A1 | |A2 | |A3 |
16.10 x1 = |A|
= 1, x2 = |A|
= 4, x3 = |A|
= 3.
16.11 Inconsistent.
16.13 Inconsistent.
16.14 x3 = 2.
b2 +c2 −a2
16.15 cos α = 2bc
.
|A1 | 7 |A2 | |A3 |
16.16 x1 = |A|
= − 27 , x2 = |A|
= 43 , x3 = |A|
= − 23
27
.
(b) We have
x1 −1 −1 a22 −a12 b1 −1 b1 a22 − b2 a12
= A b = |A| = |A| .
x2 −a21 a11 b2 b2 a11 − b1 a21
(c) We have
b1 a12
b1 a22 − b2 a12 b2 a22
x1 = =
|A| |A|
and
a11 b1
b2 a11 − b1 a21 a21 b2
x2 = = .
|A| |A|
16.20 (i) Elimination (ii) The method of inverse matrix x = A−1 b and (iii)
Cramer’s Rule.
268 ANSWER KEY
Section 17
17.5 Let x 6= y. Then α(β(x, y)) = α(βy, βx) = (αβx, αβy) 6= (αβ)(x, y).
Thus, R2 with the above operations is not a vector space.
17.9 Since for any continuous functions f and g and any scalar α the function
αf + g is continuous, C([a, b]) is a subspace of F ([a, b]) and hence a vector
space.
since
17.15 We have (1, 1) ∈ S and (1, −1) ∈ S but (1, 1) + (1, −1) = (2, 0) 6∈ S.
Hence, S is not a vector space of R2 .
17.16 The prove of this theorem uses the properties of addition and scalar
multiplication of real numbers.
Let u = (u1 , u2 , · · · , un ), v = (v1 , v2 , · · · , vn ), and w = (w1 , w2 , · · · , wn ).
(a) We have
(b) We have
(c) We have
(d) We have
(e) We have
(f) We have
(g) We have
(h) We have
1 · u = 1 · (u1 , u2 , · · · , un ) = (1 · u1 , 1 · u2 , · · · , 1 · un ) = (u1 , u2 , · · · , un ).
(b) We have
A + (B + C) =[aij ] + ([bij ] + [cij ])
=[aij ] + [bij + cij ]
=[aij + (bij + cij )]
=[(aij + bij ) + cij ]
=[aij + bij ] + [cij ]
=([aij ] + [bij ]) + [cij ]
=(A + B) + C.
(c) We have
A + 0 =[aij ] + [0] = [aij + 0]
=[aij ] = A.
(d) We have
A + (−A) =[aij ] + [−aij ] = [aij + (−aij )]
=[0] = 0.
(e) We have
α(A + B) =α([aij ] + [bij ]) = α[aij + bij ]
=[α(aij + bij )] = [αaij + αbij ]
=[αaij ] + [αbij ]
=α[aij ] + α[bij ] = αA + αB.
(f) We have
(α + β)A =(α + β)[aij ] = [(α + β)aij ]
=[αaij + βaij ] = [αaij ] + [βaij ]
=α[aij ] + β[aij ]
=αA + βB.
(g) We have
α(βA) =α(β[aij ]) = α[βaij ]
=[α(βaij )] = [(αβ)aij ] = (αβ)[aij ]
=(αβ)A.
273
(h) We have
1 · A = 1 · [aij ] = [1 · aij ] = [aij ].
α(c1 v1 +c2 v2 +· · ·+cn vn )+d1 v1 +d2 v2 +· · ·+dn vn = (αc1 +d1 )v1 +(αc2 +d2 )v2 +· · ·+(αcn +dn )vn ∈ S.
Hence, S is a subspace of V.
αD + D0 = [αdii + d0ii ] ∈ S.
Section 18
18.7 Suppose that α(4, −1, 2) + β(−4, 10, 2) = (0, 0, 0) this leads to the
system
4α1 − 4α2 = 0
−α1 + 10α2 = 0
2α2 + 2α2 = 0
275
This system has only the trivial solution so that the given vectors are linearly
independent.
18.8 Suppose that {u, v} is linearly dependent. Then there exist scalars
α and β not both zero such that αu + βv = 0. If α 6= 0 then u = − αβ v, i.e. u
is a scalar multiple of v. A similar argument if β 6= 0.
Conversely, suppose that u = λv then 1u+(−λ)v = 0. This shows that {u, v}
is linearly dependent.
and B is a basis of P2 .
18.11 Suppose that αp1 (x) + βp2 (x) + γp3 (x) = 0 for all x ∈ R. This leads
to the system
aα − 2β + γ = 0
(a − 4)β + 2γ = 0
(a − 1)γ = 0.
The set {p1 (x), p2 (x), p3 (x)} is linearly independent if the above homogeneous
system has only the trivial solution. This is the case, if the coefficient matrix
276 ANSWER KEY
18.13 By Theorem 18.2, if {~v1 , ~v2 , ~v3 } is a basis of R3 then every vector
in R3 has a unique representation as a linear combination of ~v1 , ~v2 and ~v3 .
By the previous problem, this is not true for the vector ~b. Hence, {~v1 , ~v2 , ~v3 }
can not be a basis for R3 .
Thus,
1 0 0 0
W = span , .
0 0 0 1
Next, we prove linear independence. Suppose that
1 0 0 0 0 0
α +β = .
0 0 0 1 0 0
Then α = β = 0. Hence,
1 0 0 0
,
0 0 0 1
Hence, x ∈ span{v1 , v2 , · · · , vn }.
or in matrix notation, Av = w.
(b) The condition α1 w1 + α2 w2 + · · · + αm wm = 0 leads to the homogeneous
system
a11 α1 + a21 α2 + · · · + am1 αm = 0
a12 α1 + a22 α2 + · · · + am2 αm = 0
.. ..
.=.
a1n α1 + a2n α2 + · · · + amn αm = 0
or in matrix notation
α1 0
a11 a21 ... am1
a12 a22 ... am2
α2 0
= .
... ... ...
... ... ...
a1n a2n ... amn αm 0
278 ANSWER KEY
(c) The homogeneous system in (b) has more unknowns than equations so
that by Theorem 6.1(2), the system has a non-trivial solution. Thus, S is
linearly dependent.
Section 19
19.1 λ = −3 and λ = 1.
in this case can be either a complex number or a real number. But if a root
is complex then its conjugate is also a root. Since n is odd,there must be at
least one real root.
p(λ) = λ4 + a3 λ3 + a2 λ2 + a1 λ + a0 .
then
dk11 0 ··· 0
0 dk · · · 0
22
Dk = ..
.. ..
. . .
0 0 ··· dknn
Indeed, the result is true for k = 1. Suppose true up to k − 1 then
k−1
d11 0 ··· 0 d11 0 ··· 0
0 dk−1 ··· 0 0 d22 · · · 0
22
Dk =Dk−1 D =
.. .. .. .. .. ..
. . . . . .
k−1 0 · · · dnn
0 0 · · · dnn 0
dk11 0 ··· 0
0 dk · · · 0
22
= .. .. .
..
. . .
k
0 0 · · · dnn
19.14 We use the fact that a matrix and its transpose have the same deter-
minant. Hence,
Thus, A and AT have the same characteristic equation and therefore the
same eigenvalues.
A2 =5A − 6I2
A3 =5A2 − 6A = 25A − 6A − 30I2
=19A − 30I2
A =19A2 − 30I2 = 19(5A − 6I2 ) − 30I2
4
=65A − 114I2
A5 =65A2 − 114A = 65(5A − 6I2 ) − 114A
=211A − 190I2
6
A =211(5A − 6I2 ) − 190I2
=665A − 1266I2 .
−2 5 −3
19.20 A−1 = −8 17 −10 .
5 −10 6
284 ANSWER KEY
Section 20
20.1
−2t − s −2 −1
V −3 = s : s, t ∈ R = span 0 ,
1 .
t 1 0
20.2 1 1
3 s
V = 2 : s ∈ R = span 2
s 1
and
−1 0 0
V = : s ∈ R = span .
s 1
20.3
−5s −5
V 3 = −6s : s ∈ R = span −6
s 1
and
−s −1
V −1 = 2s : s ∈ R = span 2 .
s 1
20.4 1 1
−6s −6
−8
V = − 6 s : s ∈ R = span − 61 .
1
s 1
2
+ 5i
(10 + 5i)s 5
V −i = (−18 − 24i)s : s ∈ R = span − 18 25
− 2425
i .
25s 1
2 i
(10 − 5i)s 5
− 5
i 18 24
V = (−18 + 24i)s : s ∈ R = span − 25
+ 25
i .
25s 1
20.5
s 1
1
V = s : s ∈ R = span 1 ,
s 1
285
2 2
3s 3
2
V = s : s ∈ R = span 1 ,
s 1
and 1 1
4s 4
V 3 = 34 s : s ∈ R = span 34 .
s 1
20.6
−2s −2
V 1 = s : s ∈ R = span 1
s 1
and
−s −1 0
2
V = t : s, t ∈ R = span
0 , 1 .
s 1 0
20.7
s 1
1
V = −s : s ∈ R = span −1
s 1
and
−s −1
V −2 = s : s ∈ R = span 1 .
0 0
20.8
−s
−1
s 1 ,
V1 = −s : s ∈ R = span −1
s 1
s
1
−s −1
V −1 = s
: s ∈ R = span ,
1
0 0
−s
−1
0 : s ∈ R = span 0 ,
V2 = −s −1
s 1
286 ANSWER KEY
and
0
0
−s −1
V −2 = : s ∈ R = span .
s
1
0 0
20.11
−2s −2
V 1 = s : s ∈ R = span 1
s 1
20.12
s 1
1
V = s : s ∈ R = span 1 .
0 0
Note that the vectors [−1, 1, 0]T and [0, 0, 1]T form a basis of V 5 and dim(V 5 ) =
2.
20.13
2
+ 5i s
i 2+i
V = 5 : s ∈ R = span .
s 5
287
and
2
− 5i s
−i 2−i
V = 5 : s ∈ R = span .
s 5
20.14 Suppose the contrary. That is, {v1 , v2 } is linearly dependent. Then
there is a non-zero scalar α such that v1 = αv2 . Applying A to both sides to
obtain λ1 v1 = αλ2 v2 . Multiplying v1 = αv2 by λ1 to obtain λ1 v1 = αλ1 v2 .
Substracting this equation from λ1 v1 = αλ2 v2 to obtain α(λ2 − λ)v2 = 0.
Since {v2 } is linearly independent and λ1 6= λ2 , we arrive at α = 0. This
contradicts that α 6= 0. Hence, the statement {v1 , v2 } is dependent is false
and therefore {v1 , v2 } is independent.
and
−1
VA3 = span .
2
The eigenvalues of AT are λ = 2 and λ = 3, since the matrix is upper
triangular. One can easily show that
2 1
VAT = span
0
and
2
VA3T = span .
1
20.17 λ = 1.
20.18 Since A is invertible, the homogeneous system has only the trivial
solution. That is, 0 cannot be an eigenvalue of A.
288 ANSWER KEY
20.21 We have
(2I3 − A)v = 2v − Av = 2v − 7v = −5v.
Thus, v is an eigenvector of 2I3 − A with corresponding eigenvalue −5.
Section 21
21.1 (a) Suppose that A ∼ B and let P be an invertible matrix such that B =
P −1 AP. Taking the transpose of both sides we obtain B T = (P T )−1 AT P T ;
that is, AT ∼ B T .
(b) Suppose that A and B are invertible and B = P −1 AP. Taking the inverse
of both sides we obtain B −1 = P −1 A−1 P. Hence A−1 ∼ B −1 .
and
−5s −5
V 3 = −6s : s ∈ R = span −6 .
s 1
is, B is diagonalizable. For the converse, repeat the same argument using
A = (P −1 )−1 BP −1 .
21.9
1 0 1 1 0 0
P = 0 1 1 and D = 0 1 0 .
0 1 0 0 0 2
21.10
−1 1 −1 0 1 0 0 0
1 −1 0 −1 and D = 0 −1 0 0
P =
−1
.
1 −1 1 0 0 2 0
1 0 1 0 0 0 0 −2
21.13
1 2 −1
A = 6 7 −4 .
−6 6 −4
5226 4202
21.14 A5 = .
−2101 −1077
21.15 Not diagonalizable.
21.17
1 310 + 1 310 − 1
.
2 310 − 1 310 + 1
Section 22
and
x αx αy
TE α =TE =
y αy αx
y x
=α = αTE .
x y
Hence, TE is linear.
293
and
x βx
TF β =TF
y βy
βαx αx x
= =β = βTF .
βy y y
Hence, TF is linear.
and
x αx
TG α =TG
y αy
α(x + y) x+y x
= =α = αTG
αy y y
Hence, TG is linear.
294 ANSWER KEY
22.12 Since |A + B| =
6 |A| + |B| in general, the given transformation is
not linear.
Solving this system to find T (v) = 41 (3w1 − 2w) and T (v1 ) = 41 (w1 − 2w).
T (v) =T (α1 v1 + α2 v2 + · · · + αn vn )
=α1 T (v1 ) + α2 T (v2 ) + · · · + αn T (vn )
=α1 S(v1 ) + α2 S(v2 ) + · · · + αn S(vn )
=S(α1 v1 + α2 v2 + · · · + αn vn ) = S(v).
22.19 (a)
2
5 .
8
297
Section 23
23.6 a 6= 2.
0 1 0
23.7 Ker(T ) = span 0 and Im(T ) = span 0 , 1 .
1 0 0
0 1 0
23.8 Ker(T ) = 0 and Im(T ) = span 1 , 1 .
0 1 2
w = T (α1 v1 +α2 v2 +· · ·+αk vk +αk+1 vk+1 +· · ·+αn vn ) = αk+1 T (vk+1 )+· · ·+αn T (vn )
23.16 Suppose not. Then one of the vectors of {T (vk+1 ), · · · , T (vn )} is a lin-
ear combination of the remaining vectors. Without loss of generality, suppose
that T (vk+1 ) = αk+2 T (vk+2 ) + · · · + αn T (vn ). Then T (αk+2 vk+2 + · · · αn vn −
vk+1 ) = 0. Hence, αk+2 vk+2 +· · · αn vn −vk+1 ∈ Ker(T ) = span{v1 , v2 , · · · , vk }.
It follows that αk+2 vk+2 + · · · αn vn − vk+1 = β1 v1 + β2 v2 + · · · + βk vk which
can be written as
Since the coefficient of vk+1 is 1, the previous equality says that {v1 , v2 , · · · , vk , vk+1 , · · · , vn }
is linearly independent and this contradicts the fact that {v1 , v2 , · · · , vk , vk+1 , · · · , vn }
is a basis of V. Hence, {T (vk+1 ), · · · , T (vn )} is linearly independent.
23.18 (a) Let v ∈ Ker(T ). Then T (v) = 0 and S(T (v)) = S(0) = 0. Hence,
v ∈ Ket(S ◦ T ).
(b) Let x ∈ Im(S ◦ T ). Then there is a v ∈ V such that (S ◦ T )(v) = x.
Hence, S(T (v)) = x. Let w = T (v) ∈ W. Then S(w) = x. Thus, x ∈ Im(S).
T (0). But this contradicts the definition of T. Hence, 0 is the only vector in
Ker(T ).
301
Section 24
24.7 Suppose that T (S) is linearly independent for any linearly indepen-
dent set S. Let v be a nonzero vector of V. Since {v} is linearly independent,
{T (v)} is linearly independent. That is, T (v) 6= 0. Hence, Ker(T ) = {0} and
by Theorem 24.1, T is one-to-one.
x1 y
24.8 (a) Let and 1 be two vectors in R2 . Then for any α ∈ R
x2 y2
we have
x1 y1 αx1 + y1
T α + =T
x2 y2 αx2 + y2
αx1 + y1 + αx2 + y2
= αx1 + y1 − αx2 − y2
αx1 + y1
α(x1 + x2 ) y1 + y2
= α(x1 − x2 ) + y1 − y2
αx1 y1
x1 y1
=αT +T .
x2 y2
Hence, T is linear.
0 x1 + x2
x x1
(b) If 1 ∈ Ker(T ) then 0 = T = x1 − x2 and this leads to
x2 x2
0 x1
x1 0 0
= . Hence, Ker(T ) = so that T is one-to-one.
x2 0 0
0
To show that T is not onto, take the vector 0 ∈ R3 . Suppose that
1
0
x1 2 x1
∈ R is such that T = 0 . This leads to x1 = 1 and x1 = 0
x2 x2
1
which is impossible. Thus, T is not onto.
24.15
x1 4x1 − 3x2
T = .
x2 −x1 + x2
24.16 Suppose that (S ◦ T )(v1 ) = (S ◦ T )(v2 ). Then S(T (v1 )) = S(T (v2 )).
Since S is one-to-one, T (v1 ) = T (v2 ). Since T is one-to-one, v1 = v2 . Hence,
S ◦ T is one-to-one.
x1
x 1 x 2
x2
24.17 T −1 = .
x3 x4 x3
x4
24.18 Let x ∈ X. Since S is onto, there is a w ∈ W such that S(w) = x. Since
T is onto, there is a v ∈ V such that T (v) = w. Hence, S(T (v)) = S(w) = x.
This shows that S ◦ T is onto.
305
306 INDEX
Wronskian, 155
Zero matrix, 62