Professional Documents
Culture Documents
Part 1 Algebra
Dr Tang Wee Kee
Division of Mathematical Sciences,
School of Physical and Mathematical Sciences,
Nanyang Technological University,
Singapore
Semester 1 2016/17
Contents
Contents
-1
1 Complex Numbers
1.4.1
1.4.2
1.4.3
Multiplication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.4.4
Division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
11
1.5.1
11
14
+ i sin ) . . . . . . . . . . . . . . . . . . .
15
1.7.1
18
1.7.2
19
1.7.3
20
2 Vectors
21
21
21
23
2.3.1
23
2.3.2
Vectors in 3-space . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23
CONTENTS
-1
32
2.4.1
Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
2.4.2
Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
3 Matrices
39
39
43
3.2.1
43
3.2.2
Scalar Multiplication. . . . . . . . . . . . . . . . . . . . . . . . . . . .
43
3.2.3
Matrix Multiplication. . . . . . . . . . . . . . . . . . . . . . . . . . .
45
3.3 Transpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
47
48
3.4.1
Invertible Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4.2
The inverse A
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49
3.4.3
Invertible 2 by 2 matrices . . . . . . . . . . . . . . . . . . . . . . . .
50
3.5 Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
53
3.6 Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
53
48
3.6.1
Cofactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
54
3.6.2
58
60
61
64
CONTENTS
Chapter 1
Complex Numbers
1.1
1.2
Complex Numbers
Does the quadratic equation x2 + 1 = 0 have a real root? That is, are there real numbers x
at which x2 = 1?
The answer is NO, because we have x2
To deal with the above irreducible quadratic equation, a new symbol iis introduced,
where i2 = 1. Thus, x2 + 1 = 0 has two distinct roots namely i and i.
Powers of i
i2 =
1;
i3 = (i2 )(i) =
i;
1; i4k+3 =
i:
i;
3 + i; + 9i;
Remark 1.2.6.
(1) We identify every real number x as a complex number x = x + 0i. In view of this we
may think of the set of real number as a subset of the set of complex numbers, i.e.,
R C.
(2) We say that a complex number z = x + iy is purely imaginary if the real part of z,
namely x, is zero.
1.3
Argand Diagram.
1. Since each complex number z is determined uniquely by its real and imaginary parts,
we can represent each complex number by a unique point on the xy-place, i.e., by
identifying each complex number z = x + yi by the point with coordinate (x; y):
We can also view each z = x + iy on the Argand diagram as a vector with initial point
(0; 0) and terminal point (x; y).
Im
x + iy
iy
Re
x = jzj cos
and
y = jzj cos :
<
2. Polar representation.
Using the modulus and argument we can express a complex number z = x + iy as
z = r(cos + i sin ) or z = rcis , where r = jzj and is an argument of z. This
representation is known as the polar form (also known as trigonometric form ) of z.
3. Exponential form of z
3i:
(a) Find the modulus and principal argument of z, and hence nd its polar representation.
(b) Write down the exponential form of z.
5
3
in rectangular form.
1.4
In this section, we discuss taking conjugation of a complex number, and four arithmetic
operations between two complex numbers. We shall also look at the representation of these
operations in Argand diagram. We also state and discuss related properties.
1.4.1
Example 1.4.2. Write down the conjugates of each of the following complex numbers:
z
3 + 5i
10
3:5
p i
3+i
p
+ 9i
7i
= r (cos(
= re i :
) + i sin(
)) ; or
z.
(d) For every complex number z, note that jzj = jzj and arg(z)) =
arg(z).
1.4.2
Complex numbers can be added or subtracted by adding or subtracting the real parts and
the imaginary parts separately. Formally we dene the two operations as follows:
Denition 1.4.4. Given two complex numbers z1 = x1 + y1 i and z2 = x2 + y2 i, we dene
z1
z2 = (x1 + y1 i)
(x2 + y2 i) = (x1
x2 ) + (y1
y2 )i:
Example 1.4.5.
(a) (3 + 5i) + (3:5 i) = 6:5 4i
p
p
(b) (
3 + i) ( + 9i) = (
3
) + ( 8)i
Proposition 1.4.6.
(i) z + 0 = z = 0 + z, where 0 = 0 + 0i is the real number zero.
(ii) For every z = x + iy, the complex number
( z1 ) + z.
z=
x + ( y)i satises z + ( z) = 0 =
1.4.3
z2 = z1
z2 .
Multiplication
Proposition 1.4.8.
i)
(i) z 1 = z = 1 z.
(iv) z1 z2 = z1 z2 .
(v) z z = jzj2 . In particular, if z 6= 0, then z z > 0.
(vi) z1 (z2 + z3 ) = z1 z2 + z1 z3 (Distributive Property)
Product in Polar Form
From the above algebraic expression, it is not clear there is a geometrical relationship
between z1 , z2 and their product. We shall express the product in polar representation to
deduce a geometrical relation.
Let us express the two complex numbers in polar form:
z1 = r1 (cos
+ i sin
1)
and z2 = r2 (cos
+ i sin
2 ).
We have
z1 z2 = r1 (cos
= r1 r2 [(cos
cos
sin
= r1 r2 (cos(
+ i sin
1
sin
1)
2)
2)
r2 (cos
+ (cos
sin
+ i sin(
+ i sin
2
2)
+ sin
cos
2 )i]
2 )) :
Proposition 1.4.9.
1. Modulus of the product is the product of moduli:
jz1 z2 j = r1 r2 = jz1 j jz2 j
2. Argument of the product is the sum of arguments:
arg(z1 z2 ) =
= arg(z1 ) + arg(z2 ):
This implies the complex number z1 z2 lies on the line obtained by rotating the line
segment representing z1 by the angle arg(z2 ).
Represent the product of z1 and z2 on an Argand diagram:
Remark For a complex number z, the complex number z (cos + i sin ) is represented
on the Argand diagram by by rotating z through .
10
1.4.4
Division
p
1
p in the form a + b 5, we perform the following
3+2 5
3
1
p
3+2 5 3
p
p
2 5
3 2 5
3
p =
p
=
29
2 5
32 + (2 5)2
2p
5:
29
z1 z2
z1 z2
=
z2 z2
jz2 j2
3+5i
:
2 i
+ i sin
1)
and z2 = r2 (cos
z1
jz1 j
=
z2
jz2 j
arg(
z1
) = arg(z1 )
z2
arg(z2 )
Thus, we have
z1
r1
= (cos(
z2
r2
2)
+ i sin(
2 )) :
+ i sin
2 ),
such that
1.5
11
1
z
= z.
Theorem 1.5.1 (The Fundamental Theorem of Algebra). Every polynomial equation of the
form
an x n + an 1 x n 1 +
+ a1 x + a0 = 0;
in which the coe cients a0 ; a1 ; : : : ; an 1 ; an are any complex numbers, whose degree n is
greater than or equal to one, and whose leading coe cient an is not zero, has exactly n roots
in the complex number system, provided each multiple root of multiplicity m is counted as m
roots.
Proof (Omitted): Textbook on theory of complex analysis.
1.5.1
Recall that for a quadratic equation ax2 + bx + c = 0 (or simply a quadratic expression), its
discriminant ,D, is dened as D = b2 4ac.
(i) If D > 0, the quadratic equation ax2 + bx + c = 0 has two distinct real roots given by
x=
b2
2a
4ac
12
(ii) If D = 0, the quadratic equation ax2 + bx + c = 0 has repeat real roots given by
x=
b
:
2a
(iii) If D < 0, the quadratic equation ax2 + bx + c = 0 has two distinct complex roots given
by
p
(b2 4ac)
b i
:
x=
2a
Note that the two complex roots are conjugate of each other.
When D < 0, the quadratic equation or expression is said to be irreducible.
Example 1.5.2. Solve the quadratic equation 2x2
3x + 5 = 0.
In fact there is a more general result, stated below, which we are ready to prove.
Polynomial with Real Coe cients
Theorem 1.5.3. Suppose p(x) = an xn + an 1 xn 1 + + a1 x + a0 is a polynomial in x with
real coe cients ak s. If z is a solution to p (x) = 0, then the conjugate z is also a solution
of p (x) = 0:
For example: suppose z0 is a complex root of 9x5 +7x2 6x+ = 0, then
is
5
2
also a complex root of 9x + 7x 6x + = 0. Therefore,
is a quadratic factor of 9x5 + 7x2 6x + . Moreover, (x z0 )(x z0 ) = x2 (z0 + z0 )x + z z0
is a real coe cient quadratic factor.
As a consequence of the Fundamental Theorem of Algebra and the above result, we have
the following useful result.
13
Theorem 1.5.4. Every odd degree polynomial p(x) with real coe cients has at least one
real root.
For example: 9x5 + 7x2
6x +
14
1.6
De Moivres Theorem
In this section, we shall state and prove de Moivres Theorem. For a positive integer n, we
use De Moivres Theorem to nd n-th root of a complex number. In particular, we apply it
to determine all distinct n-th roots of unity.
Theorem 1.6.1 (De Moivres Theorem).
For every integer n,
(cos + i sin )n = cos n + i sin n :
We look at some examples to illustrate the theorem before we discuss its proof.
Example 1.6.2. Express each of the following complex numbers in the form cos n +i sin n .
(a) (cos + i sin )9
(b) (cos + i sin )
+ i sin 4 )
(b) (cos
+ i sin 3 )9
Example 1.6.4. Express each of the following complex numbers in the form (cos +i sin )n .
(a) cos 7 + i sin 7 .
(b) cos 5
i sin 5
1.7
15
+ I SIN )
+ i sin )
We begin with an example to have a geometrical idea of nding roots of a complex number.
Example 1.7.1. Find all distinct cube roots of z = cos
+ i sin
+ i sin
are:
7
7
13
13
+ i sin ); (cos
+ i sin ) & (cos
+ i sin
):
9
9
9
9
9
9
16
p
n
r cos
2k +
n
+ i sin
2k +
n
<
+ i sin ) are
; k = 0; 1; 2; 3; : : : ; n
1:
p
n
r ei
+2k
n
; k = 0; 1; 2; 3; : : : ; n
1:
17
+ I SIN )
2k
2k
2k
) = cos( +
) + i sin( +
); k = 0; 1; 2; : : : ; n
n
n
n
n
n
+ i sin
= i are
+ 4k
=2 2k
+
) = cis (
); k = 0; 1; 2; 3; : : : ; 8;
9
9
18
They are
cis (
18
); cis (
5
9
13
17
21
25
29
33
); cis ( ) = cis = i; cis ( ); cis (
); cis (
); cis (
); cis ( ) & cis ( )
18
18
2
18
18
18
18
18
18
18
1.7.1
2k
2k
+ i sin
; k = 0; 1; 2; 3; : : : ; n
n
n
1:
Eulers Formula
Here, we introduce the Eulers Formula which help us in representing a complex number. In
advanced courses in mathematics, a formal denition of ez is discussed via series.
If we represent the unit complex number cos + i sin by ei , i.e.,
ei = cos + i sin ;
then for each integer n, we have
cos n + i sin n = ein :
Treating ei as exponential in , we note that
ei(n ) = ei
= (cos + i sin )n :
1.7.2
19
+ I SIN )
We can use De Moivres Theorem to express cos n , sin n and tan n in terms of powers of
cos , sin and tan .
Tools:
cos n = Re(cos n + i sin n ) = Re (cos + i sin )n ;
sin n = Re(cos n + i sin n ) = Im (cos + i sin )n ;
Apply binomial expansion to (cos + i sin )n
Notation used: c
cos , s
sin , t
tan .
=
=
=
=
=
=
Im(cos 3 + i sin 3 )
Im (cos + i sin )3 (why?)
Im (c + is)3
Im (c3 + 3c2 is + 3ci2 s2 + i3 s3 )
Im (c3 3cs2 + i(3c2 s s3 ))
3c2 s s3
Using c2 + s2 = 1, we have
sin 3
=
=
=
=
3c2 s s3
3(1 s2 )s s3
3s 4s3
3 sin
4 sin3 :
3cs2 = c3
3c(1
c2 ) = 4c3
3c
Using the expression for both sin 3 and cos 3 , we obtain a similar expression for tan 3 :
sin 3
3c2 s s3
= 3
cos 3
c
3cs2
3c2 s s3 1=c3
3t t3
= 3
:
=
c
3cs2 1=c3
1 3t2
tan 3 =
20
1.7.3
or sinn
z+
Thus we have cos
and group z k and z1k together.
1
z
and sin
= 2i1
1
z
i sin .
= cos
1
z
1
= cos k
zk
i sin k
which gives
zk +
1
= 2i sin k :
zk
1
= 2 cos k and z k
zk
cos
1
8
= (cos ) =
1
1
1
z 3 + 3z 2 + 3z( )2 + ( )3
z
z
z
=
1
1
(z + )
2
z
=
1
8
(z 3 +
1
1
) + 3(z + )
3
z
z
1
1
[2 cos 3 + 3(2 cos )] = (cos 3 + 3 cos ) :
8
4
Chapter 2
Vectors
Vectors are quantities which have both magnitude and direction. Examples of vectors include
acceleration, displacement, force, momentum and velocity.
2.1
We may use a directed segment to represent a vector graphically. The directed segment
!
from a point A to a point B is denoted by AB: The length jABj of the segment represents
the magnitude of the vector.
We may also use lower case bold letters u, v, w to denote vectors.
Two vectors are equal if they have the same magnitude and direction.
Given a vector v, its negative
direction.
2.2
22
CHAPTER 2. VECTORS
A
B
v
u
-u
u
C
u+v
Vector Addition
Vector Subtraction
!
!
!
If u = OA and v = OB; then AB = v
!
!
u: Or simply AB = OB
!
OA:
jvj :
Conversely, any vector that has the same direction as v must be equal to v for some
> 0:
Properties of vector addition and scalar multiplication
u + v = v + u;
u + (v + w) = (u + v) + w;
( v) = (
) v; ;
2 R,
( + ) v = v+ v; ;
(u + v) = u + v;
2 R,
2 R:
2.3
2.3.1
23
v2
j
i
2.3.2
u = (u1 ; u2 ) = ( u1 ; u2 ) ;
2 R:
Vectors in 3-space
Consider a rectangular coordinate system, with three mutually perpendicular lines, called
coordinate axes, passing through a point called the origin.
z
(0, y 0, z 0)
z0
P(x 0, y 0, z 0)
j
i
y0
y
x0
(x 0, y 0, 0)
x
v1
24
CHAPTER 2. VECTORS
Like vectors in 2-space, a vector in 3-space from the point P1 = (x1 ; y1 ; z1 ) to the point
!
P2 = (x2 ; y2 ; z2 ) will be the vector P1 P2 = (x2 x1 ; y2 y1 ; z2 z1 ). In particular, if the
!
starting point is the origin O = (0; 0; 0) and the end point P = (x0 ; y0 ; z0 ), then OP =
(x0 0; y0 0; z0 0) = (x0 ; y0 ; z0 ).
The numbers x0 ; y0 and z0 are called the components of v.
We also write
v =x0 i+y0 j+z0 k;
where i, j and k are unit vectors in the positive direction of0x, y 1
and z axis respectively.
x0
The vector v may be written in the column vector form v = @ y0 A.
z0
Suppose u = (u1 ; u2 ; u3 ) and v = (u1 ; v2 ; v3 ) are two vectors in 3-space. Then we have
the following:
1. u = (u1 ; u2 ; u3 ) is the zero vector if and only if u1 = 0, u2 = 0 and u3 = 0.
2. Two vectors u = (u1 ; u2 ; u3 ) and v = (u1 ; v2 ; v3 ) are equal if and only if u1 = u1 ,
u2 = v2 and u3 = v3 .
3. u + v = (u1 ; u2 ; u3 ) + (u1 ; v2 ; v3 ) = (u1 + u1 ; u2 + v2 ; u3 + v3 ):
4. ku = k(u1 ; u2 ; u3 ) = (ku1 ; ku2 ; ku3 ).
Solution
p1 )
2
in 3-space.
p
u21 + u22 :
25
Remark The distance d between two points P = (u1 ; u2 ; u3 ) and Q = (u1 ; v2 ; v3 ) is given
!
!
by the norm kP Qk of the vector P Q,
d=
p
(v1
u1 )2 + (v2
u2 )2 + (v3
u3 )2 :
p
p
A vector of length 1 is called a unit vector. The vector u = ( 1= 2; 0; 1= 2) is a unit
vector.
In particular, the following vectors
i = (1; 0; 0); j = (0; 1; 0) and k = (0; 0; 1)
are unit vectors along the positive direction of x-, y- and z-axes respectively.
It follows that:
1
u is the unit vector along the direction u.
For a nonzero vector u, the vector u
^ = kuk
26
CHAPTER 2. VECTORS
5. ( u) ( v) = (
6. u (v + c) = u v + u c:
7. (u + v) c = u c + v c:
Proposition 2.3.3. If u =
u1
u2
and v =
v1
v2
; then u v =u1 v1 + u2 v2 :
Solution We have
u v = (2; 3; 1) (1; 0; 5) = 2(1) + 3(0) + ( 1)(5) =
) u v:
3:
27
Example 2.3.5. Find the angle between u = i+2j 2k and v = 2i+3j 6k:
Solution Let
cos =
=
u v
jjujj jjvjj
20
:
21
20
= 0:310 rad
21
Example 2.3.6. Find the angle between u = i 2j+2k and v = 2i+3j 6k:
= cos
Solution Let
cos =
=
=
=
u v
jjujj jjvjj
20
( cos < 0 implies that is
21
20
cos 1
=
rad
21
28
CHAPTER 2. VECTORS
Work Done
One of the many applications of dot product in science and engineering is to compute
work done. Recall that force and displacement are vector quantities, they have magnitude
and direction.
The work done by a constant force F acting through a displacement D is the dot product
W = F D:
F
D
!
!
!
If u = P Q and v =P R, then P S
is the (perpendicular) projection of u onto v:
v
S
!
P Q cos
= jjujj
u v
u v
=
:
jjujj jjvjj
jjvjj
v
^=
u v
jjvjj
v
^=
u v
v
:
jjvjj jjvjj
Example 2.3.8. Find the vector projection of u = i 2j+2k onto v = 2i+3j 6k:
29
Solution
u v=
16
q
jjujj = (
q
jjvjj = (
)2 + (
)2 + (
)2 = 3
)2 + (
)2 + (
)2 = 7
v
2i+3j 6k
=
:
jjvjj
7
Then the vector projection of u onto v =
Therefore v
^=
projv u =
=
u v
jjvjj
v
^
16
(2i+3j 6k) :
49
Cross product
The cross product (or vector product) of
two vectors u and v is dened as
u v = (jjujj jjvjj sin ) n
^;
where is the angle between u and v,
and n
^ is the vector perpendicular to both u and v,
governed by the right hand rule
(u-index nger, v-middle nger, n-thumb, of right hand).
v=
4. u
(v + w) = u
5. ( u)
Example 2.3.10. i
v = 0:
u (anti-commutative).
( v) = (
Example 2.3.9. i
= 0 and thus u
v+u
) (u
i=j
v) :
j=k
j = k; j
k =0
k = i; k
i=j
n
v
30
CHAPTER 2. VECTORS
j
u2
v2
u2
v2
u3
v3
k
u3
v3
u1
v1
u3
v3
j+
u1
v1
u2
v2
k:
Proof.
u
u2
v2
u3
v3
u1
v1
u3
v3
j+
u1
v1
u2
v2
u2 v 1 ) k
k:
Solution
u
v = (1; 1; 1) (1; 0; 1)
1 1
1 1
1 1
=
;
;
0 1
1 1
1 0
31
v = (1; 0; 1):
Thus, we have
1
^ = p (1; 0; 1):
n
2
2. Area of parallelogram ABCD
=base height
= jjBCjj h
= jjBCjj jjABjj cos
!
!
= AB BC :
3. Area of triangle ABC =
1
2
C
b
1
2
!
AB
!
BC :
Example 2.3.14. Find the area of triangle whose vertices are A (1; 0; 0) ; B (0; 1; 0) and
C (0; 0; 1).
!
!
Solution Area of triangle ABC = 21 AB BC :
0 1 0 1 0
1
0
1
1
!
!
!
AB = OB OA = @ 1 A @ 0 A = @ 1 A
0
0
0
0 1 0 1 0
1
0
1
1
!
!
!
AC = OC OA = @ 0 A @ 0 A = @ 0 A
1
0
1
0
1 0
1 0 1
1
1
1
!
! @
A
@
A
@
1
0
1 A
AB BC =
=
0
1
1
0 1
p
p
1
2 + 12 + 12
1
3
1
Area of triangle ABC = 2 @ 1 A =
=
:
2
2
1
1
0
1
0
1
u1
v1
w1
u = @ u2 A ; v = @ v2 A and w = @ w2 A in R3 ;
u3
v3
w3
32
CHAPTER 2. VECTORS
the dot product u (v
v2 v3
w2 w3
w) = u1
u2
v1 v3
w1 w3
v1 v2
w1 w2
+ u3
w) = v (w
u) = w (u
v)
3
Theorem 2.3.15. (a) For three non-coplanar u, v and w in R , the absolute value ju (v w)j
of the scalar triple product is the volume of the parallelepiped which is a three-dimensional
gure formed (by six parallelograms) whose sides are the three given vectors.
(b) For general three vectors u; v and w, if the scalar triple product u (v w) = 0, we
may conclude that either at least one of the vectors u; v and w is a zero vector or the three
vectors are coplanar (they lie on the same plane).
Example 2.3.16. Find the volume of the parallelepiped formed by vectors u = (1; 2; 3),
v = ( 1; 1; 0) and w = (1; 2; 1).
Solution We compute the scalar triple product:
u (v
w) = 1
=1
1 0
2 1
1 0
+3
1 1
2( 1) + 3( 3) =
1 1
1 2
2.4
2.4.1
A line on a Cartesian plane may be determined by a point on the line and the gradient of
the line. In a space, line is uniquely determined by its direction vector and a point on
the line.
v
r0
r0+v
r0+2v
r0+3v
33
If a line ` is parallel to a vector v and passes through a point with position vector r0 ,
then the vector equation of ` is
` : r (t) = r0 +tv, t 2 R:
Example 2.4.1. If A; B and C have coordinates (1; 1; 0) ; (0; 1; 0) and (0; 0; 1) respectively.
Find a vector equation for the line that is parallel to BC and passes through A.
Solution
1
1
!
r0 = OA = @ 1 A
0
! B
v = BC = B
@
1
C
C
A
0
B
B
@
1
0
C
C = @ 1 A:
A
1
r (t)= r0 +tv
0
0 1
1
B
@
1 A + tB
=
@
0
Equation of a line: Cartesian form
1
C
C
A
1
0
1
x0
v1
Suppose that a line ` passes through r0 = @ y0 A and is parallel to v = @ v2 A. If
z0
v3
(x; y; z) is a point on `; then
0 1 0
1 0
1
x
x0
v1
r = @ y A = @ y0 A +t @ v2 A , t 2 R.
z
z0
v3
Therefore,
Thus
1 0
1
x
x0 + tv1
@ y A = @ y0 + tv2 A :
z
z0 + tv3
(2.1)
34
CHAPTER 2. VECTORS
Suppose that v1 ; v2 ; v3 6= 0; we obtain the Cartesian Equation of ` :
x
x0
v1
y0
v2
z0
v3
1=
y+1
z
= :
2
3
(i) Find a vector equation for `: (ii) Find the parametric equation for `:
Solution The Cartesian equation for ` :
x
1
1
( 1)
z 0
=
:
2
3
(i) The line ` is parallel to (1; 2; 3) and passes through (1; 1; 0) : Thus the a vector equation
for ` is
0
1 0 1
1
1
r = @ 1 A +t @ 2 A , t 2 R.
0
3
(ii) The parametric equation for ` :
x = 1 + t; y =
1 + 2t; z = 3t; t 2 R:
respectively. Let
B
v1 = B
@
0 1
0
C
C and v2 = @ 0 A
A
1
v1 v2
= cos
jjv1 jj jjv2 jj
1
= or 60 :
(2) (1)
3
35
r0
O
Example 2.4.4. Find the distance from a point S(1; 3; 2) to the line ` : r = (0; 1; 2) +
t (1; 0; 1) ; t 2 R:
!
PS v
!
P S sin =
jvj
0
1
=
B
B
@
C
C
A
0
B
B
@
jjvjj
1
C
C
A
0
=
B
B
@
1
p
C
C
A
3
= p units.
2
be the
36
CHAPTER 2. VECTORS
2.4.2
Plane
A plane in a space is determined by a point and its "inclination". This inclination can be
specied by a vector n that is normal, or perpendicular, to the plane.
P(x, y, z)
P 0(x0,y 0,z 0 )
M
a (x
x0 ) + b (y
y0 ) + c (z
z0 ) = 0
This is called
0 the1 scalar equation of the plane through P0 = (x0 ; y0 ; z0 ) with normal
a
@
b A : It can be simplied to
vector n =
c
ax + by + cz = d;
37
Example
0 2.4.5.
1 Find the scalar equation for the plane through P0 (1; 2; 3) and perpendicular
4
to n = @ 5 A :
6
Solution The required equation is
4 (x
1) + 5 (y
2) + 6 (z
3) = 0:
Simplifying, we have
4x + 5y + 6z = 32:
Example 2.4.6. Find the scalar equation of the plane passing through A (1; 0; 0) ; B (0; 1; 0)
and C (0; 0; 1) :
!
!
Solution We rst determine a vector normal to the plane. The vector AB and AC are
vectors parallel to the plane. Therefore their cross product is a vector normal to the plane.
!
!
Let n =AB AC:
0
1
0
1 0
1
1
1
B
C
C:
@
A
@
1
0 A=B
n=
@
A
0
1
The scalar equation of the plane is
1 (x
1) + 1 (y
0) + 1 (z
0) = 0;
or
x + y + z = 1:
Distance from a point to a plane
If P is a point of the plane with normal n, then the distance from any point S to the
!
plane is the length of the vector projection of P S to n; which is equal to
!
PS n
=
jjnjj
Example 2.4.7. Find the distance from the point S (1; 0; 1) to the plane x
Solution
A normal vector of the plane x
x
2y + z = 1.
p
6.
We nd a point P on the given plane by nding (x; y; z) which satises the equation
2y + z = 1:
Let x = 0 and y = 0, we have z =
38
CHAPTER 2. VECTORS
P0
1
1
(0; 0; 1)) p (1; 2; 1) = p :
6
6
2y + 2z = 1 and 6x
4y + 3z = 7:
1
0
1
1
6
Solution. The vectors n1 = @ 2 A and n2 = @ 4 A are the normal vectors of the
2
3
respective planes.
If
n1 n2
= cos
jn1 j jn2 j
n1 n2
jjn1 jjjjn2 jj
j6 + 8 + 6j
p
= cos
1 + 4 + 4 36 + 16 + 9
: Therefore
:
Chapter 3
Matrices
A matrix is simply a rectangular array of numbers or other mathematical objects. Rectangular arrays of real numbers (or complex numbers) arise in many contexts and in sciences and
engineering as well as social sciences. Beside providing a neat representation of data from
which we can obtain further information, matrices have wide applications in this modern
generation.
3.1
n, is
0
B
B
B
@
ain
a1j
a2j
..
.
amj
39
C
C
C:
A
40
CHAPTER 3. MATRICES
Notation and Terminology
1. Capital letters A; B; C; : : : are used to denote matrices, and lowercase letters to denote
numerical quantities. Some examples:
0
1
a b c
1 3 5
z1 z2 z3 z4
A=
;T = @ d e f A;Z =
9 6 3
y1 y2 y3 y4
x y z
Both square brackets or round brackets are used to enclose the array of entries.
2. The number m of rows and the number n of columns describe the size of a matrix.
We write it as m n, and read as m by n.
3. The matrix
is sometimes written as
B
B
A=B
@
a11
a21
..
.
a12
a22
..
.
a1n
a2n
..
.
am1 am2
amn
[aij ]m
C
C
C:
A
or simply [aij ] :
To refer to the (i; j)th-entry of the matrix A, we may use the notation Aij . So,
Aij = aij .
2
1
4
9
For A =
2
3
6
4
3
5
7
3 p0 5, we have A11 = 1; A21 = 9 and A34 =??.
2
8
4. Usually, we match the letter denoting a matrix with the letter denoting its entries. For
a matrix B, its ij-entry is bij .
5. When m = 1, the matrix has only one row,i.e.
A=
a1 a2 a3
an
41
7. The m n matrix with zeros as its entries is called the zero matrix and we denote
it by 0.
8. When m = n, we call A a square matrix of size n. (The rectangular array now looks
like a square.)
For an n
n square matrix
2
6
6
A=6
4
a11
a21
..
.
a12
a22
..
.
a1n
a2n
..
.
am1 am2
ann
7
7
7:
5
The entries a11 ; a22 ; a33 ; : : : ; ann are called the diagonal entries. They are on the
main diagonal of A.
9. The n n square matrix where all the entries along the diagonal from the top left to
the bottom right are 1, and 0 elsewhere, is called the identity matrix. It is often
denoted as In . E.g.,
0
1
1 0 0
1 0
I2 =
; I3 = @ 0 1 0 A :
0 1
0 0 1
10. The n n square matrix where all the o diagonal entries (i.e. entries below and above
the main diagonal) are 0 is called an diagonal matrix.
1
0
0
1
3 0 0 0
a 0 0
B 0 3 0 0 C
C
A=@ 0 e 0 A B=B
@ 0 0 0 0 A:
0 0 g
0 0 0
11. The n n square matrix where all the entries below the main diagonal are 0 is called
an upper triangular matrix.
1
0
0
1
1 3 5
0
a b c
B 0 3
2 9 C
C:
@
0 e f A B=B
A=
A
@ 0 0 0
0 0 g
0 0 0 0:8
In a similar way, a lower triangular matrix is a square matrix where all the entries
above the main diagonal are 0. E.g.
0
1
53 0
0
0 0
B 4 3
0
0 0 C
B
C
a 0
B
C:
0
0
C=
B = B 1=2 0
C
p
c d
@ 0
1 0:4
2 0 A
1
3
5
7 9
42
CHAPTER 3. MATRICES
12. Lastly, the term scalars refer to real numbers or complex numbers in discussing matrices (and vectors).
Denition 3.1.1. Two matrices are dened to be equal if they have the same size and
their corresponding entries are equal.
Matrices A and B are equal if they have the same size (same number of rows and same
number of columns) and Aij = Bij for all i and j. (Here, note that if the size of both matrix
is m n, then 1 i m and 1 j n.)
Example 3.1.2. Solve the following matrix equation for a; b; c and d.
a b
b+c
3d + c 2a 4d
8 1
7 6
[Solution] Note that both matrices are 2 by 2. For matrices to be equal, each corresponding entries must be equal. Therefore we have
a b
b+c
3d + c
2a 4d
=
=
=
=
8
1
7
6
[Solution]
3; c = 4; d = 1.
8
i+j
>
>
<
0
3 matrix where bij =
j
>
>
:
6
B=6
4
0
0
0
7
7:
5
if i > j
if i = j
if i < j
3.2
43
3.2.1
n matrices, then
(a) the addition (or sum) A + B is the matrix obtained by adding entries in the same
positions, i.e.,
(A + B)ij = (A)ij + (B)ij :
(b) the di erence A B is the matrix obtained by subtracting entries of B from the
corresponding entries of B, i.e.,
(A
B)ij = (A)ij
(B)ij :
1 2 3
4 5 6
and B =
A+B =
3.2.2
B=
7 8 9
10 11 12
. Then
1 7 2 8 3 9
4 10 5 11 6 12
8 10 12
14 16 18
6
6
6
6
6
6
Scalar Multiplication.
1 3 5
7 9 11
2 6 10
14 18 22
. Then
; ( 3)A =
3
21
9
27
15
33
; and
1
A=
3
1
3
7
3
1
3
5
3
11
3
44
CHAPTER 3. MATRICES
Let A =
4 3 2
1 3 1
Then
2A
,B=
1
B+ C=
3
7 2
5 3
B = A + ( 1)B. It is common
0
1
and C =
3
3
6 9
0 12
2 2 7
8 3 7
.
For matrices, when we perform arithmetic operations like addition, dierence and scalar
multiplication, we are basically performing similar arithmetic operations sum, dierence and
multiplication on numbers on entry-level. Thus, we would expect properties such as associativity, commutativity and distributive to hold for such matrix operations.
Theorem 3.2.6. Assuming the sizes of matrices are such that the indicated operations can
be performed, the following rules of matrix arithmetic are valid.
1. A + B = B + A (Commutative Law for addition)
2. (A + B) + C = A + (B + C) (Associative Law for addition)
3. A + 0 = 0 + A = A (Additive Identity)
4. A + ( A) = 0 (Additive Inverse)
5. 0
6.
7. (
8.
A=
(A
B) = A
)A = A
( A) = (
)A
3.2.3
45
Matrix Multiplication.
Can we multiply two matrices in the similar way i.e. by multiplying corresponding entries
like what we have done for addition, subtraction or scalar multiplication?
It turns out that such a denition is not very helpful for most problems. Experience has
led mathematicians to the following more useful denition of matrix multiplication.
Denition 3.2.7. If A is an m r matrix and B is an r n matrix, then the matrix product
(or simply the product) AB is the m n matrix whose (i; j)th entry is determined by
(AB)ij = Ai1 B1j + Ai2 B2j +
+ Air Brj =
r
X
Aik Bkj :
k=1
Note that the (i; j)th entry of AB is the value obtained by taking the dot product of the
vector formed by the ith row of A and that formed by the jth column of B.
0
A11
A21
..
.
A12
A22
..
.
A1r
A2r
..
.
B
B
B
B
AB = B
B Ai1 Ai2
B .
..
@ ..
.
Am1 Am2
Air
..
.
Amr
0
C B11 B12
C
C B B21 B22
CB
C B ..
..
C@ .
.
C
A Br1 Br2
B1j
B2j
..
.
Brj
1
B1n
B2n C
C
C:
A
Brn
Example 3.2.8.
0
1
2
1 2
1
(a) Let A =
and B = @ 4 A. Note that A is 2
3 1
4
2
3 1, the product AB will be a 2 1 matrix.
To nd the entries of AB:
(1; 1)-th entry:
(2; 1)-th entry:
(3; 1)-th entry:
Thus, we have AB =
4
.
6
3 matrix and B is
46
CHAPTER 3. MATRICES
(b) Let A =
1 2
3 1
Note that A is 2
(c) Let A =
1 2
3 1
1
3
B 0 C
C
and C = B
@ 2 A.
p
3
1
4
4
6
; DA = @
1
2
16 A :
Remark
1. Matrix multiplication is not commutative.
2. It is not true that if AB = 0, then A = 0 or B = 0. (Exercise: Find two non-zero
2 2 matrices A and B with AB = 0.)
Without computing the entire product, we may compute a particular row or column of
a matrix product AB as follows:
Proposition 3.2.9 (Optional). Let A and B be m
r and r
(a) the jth column of (AB) is the matrix product of A and the jth column of B;
jth column of (AB) = A[jth column of B]:
(b) the ith row of (AB) is the matrix product of the ith row of A and the matrix B;
ith row of (AB) = [ith row of A]B:
Example 3.2.10.
Let A =
4 3 2
1 3 1
,B=@
1
2
0
0
1
3
1
1
0 A.
2
Find (a) the 2nd column of AB and (b) the last row of AB.
3.3. TRANSPOSE
47
Identity Matrix
Recall that the identity matrix In is the square matrix In of size n with (i; j)th-entry
1
0
(In )ij =
Thus,
1
I1 =
1 0
0 1
; I2 =
if i = j;
if i =
6 j
0
1
1 0 0
; I3 = @ 0 1 0 A ; : : :
0 0 1
The role of identity matrices in matrix multiplication is like the number 1 in usual
multiplication. Some algebraic properties on matrix multiplication are recorded in the next
theorem.
Theorem 3.2.11. Assuming the sizes of matrices are such that the indicated operations can
be performed, the following rules of matrix arithmetic are valid.
1. AIn = A and Im A = A if A is m
n. (Identity)
3.3
Transpose
Example 3.3.2.
Let A =
4 3 2
1 3 1
0
1
0
4 1
1
T
T
@
A
@
3 3 ;B =
0
A =
2 1
1
1
1
0 A. Then
2
1
1
2
0
0
1
3
2
1
0
0
3 A
2
48
CHAPTER 3. MATRICES
Proposition 3.3.3. If the sizes of the matrices are such that the stated operations can be
performed, then
(a) (AT )T = A
B)T = AT
(b) (A
BT
is a scalar.
(d) (AB)T = B T AT
Proof. (Exercise. )
3.4
Matrix Inverse
In this section, we shall discuss matrix invertibility with respect to matrix multiplication.
We shall discuss uniqueness, properties of matrix inverses, nding inverses.
3.4.1
Invertible Matrices
2
5
and B =
5
3
2
1
1
0 0 0
Example 3.4.3. Let C = @ a b c A. Is there a matrix D such that CD = I?
d e f
Solution
1
a b c
Example 3.4.4. Let E = @ 2a 2b 2c A. Is there a matrix F such that EF = I?
d e f
49
Solution
Note that matrix invertibility is dened for square matrices only. It is clear that zero
square matrices are singular. By Proposition 3.2.9, we deduce that the following are singular.
Proposition 3.4.5.
(a) A square matrix with a row (or a column) of zeros is singular.
(b) A square matrix with a row (or a column) which is a multiple of another row (or
column) is singular. In particular, a square matrix with identical rows (or columns) is
singular.
A row Ri is a linear combination of other rows Rj , where j 6= i, means that there are
scalars j s such that
X
Ri =
+ i 1 Ri 1 + i+1 Ri+1 +
j Rj = 1 R1 + 2 R2 +
n R2 n:
j6=i
Proposition 3.4.6. A square matrix in which a row (or a column) is a linear combination
of other rows (or columns) is singular.
0
1
a
b
c
B
d
e
f C
C.
Example 3.4.7. Consider G = B
@ 2a 5d
A
2b 5e& 2c 5f
Note that R3 = 2R1 + ( 5)R2 . The matrix G singular.
3.4.2
The inverse A
Now we prove that an invertible matrix cannot have more than one inverses. In other words,
if an inverse exists, it is unique.
b are both inverses of A, then B = B.
b
Proposition 3.4.8. If B and B
Proof. Note that
b = I&BA
b = I:
AB = I&BA = I; and AB
50
CHAPTER 3. MATRICES
The above proposition says that the inverse of an invertible matrix A is unique. In view
of this, we shall denote the inverse of A by A 1 . Thus, we have
1
AA
= I and A 1 A = I:
It follows immediately from the above equation and denition of matrix invertibility that
the inverse A 1 of a matrix A is invertible.
Proposition 3.4.9. If A is an invertible matrix, then the inverse A
A
3.4.3
is invertible and
= A:
Invertible 2 by 2 matrices
1
ad
d
c
bc
Proof.
(=: Suppose ad
bc
d
c
b
a
A = I2 & A
1
ad
d
c
bc
b
a
= I2 :
1
ad
a b
c d
bc
d
c
b
a
bc 6= 0.
is not invertible.
Suppose ad
is
is invertible, then ad
a b
c d
51
For the case (a) where a = 0, note that either b = 0 or c = 0. If b = 0, then A has a
row of zero and this it is not invertible. If c = 0, then the rst row and second row of
A are multiple of each other. Hence A is not invertible.
For The case (b) where a 6= 0, we have d = bca . This the second row of A is a scalar
multiple of the st row of A. ( ac of the rst row ) Thus, A is not invertible.
Therefore, if ad
ad bc 6= 0.
(a) A =
5 3
7 9
(b) B =
8
6
(c) C =
cos
sin
2 matrices is invertible. If
4
3
sin
cos
Remark 3.4.13.
(a) For matrices of other sizes, there is similar result on invertibility via determinant
which will be discussed in latter section. There is also a formula for the inverse of an
invertible matrix. However, the formula is more complicated than the case for 2 2
matrices.
(b) We may use Gaussian method to determine whether a square matrix is invertible and
also to nd its inverse. This will be dealt with in another mathematics Course.
However, for a diagonal matrix, there is an easy way to determine its invertibility and
its inverse.
52
CHAPTER 3. MATRICES
Example 3.4.14. Find the inverse of the following diagonal matrix, if abcd 6= 0.
0
1
a 0 0 0
B 0 b 0 0 C
B
C
@ 0 0 c 0 A
0 0 0 d
More generally, we have the following result for diagonal matrices.
Proposition 3.4.15. Suppose A = [aij ] is a diagonal matrix. Then A is invertible if and
only of its diagonal entries are non-zero, i.e., aii 6= 0 for every i. When A is invertible, A 1
1
is a diagonal matrix whose ith diagonal entry is
for each i.
aii
Proposition 3.4.16 (Optional). Suppose A a triangular matrix. Then A is invertible if
and only if all diagonal entries are non zero. Moreover, the inverse of invertible upper (resp.
lower) triangular matrix will be an upper (resp. lower) triangular matrix.
(The inverse of a general invertible triangular matrix is quite messy to include here.)
Now, we study some properties of invertible matrices.
Proposition 3.4.17. Let A be an invertible matrix. Then we have
1. AB = AC =) B = C and
2. BA = CA =) B = C.
Proposition 3.4.18. Let A and B be invertible matrices. Then the matrix product AB is
invertible and
(AB) 1 = B 1 A 1 :
Proposition 3.4.19. Suppose A is invertible.
(a) Then its transpose AT is invertible. In this case,
AT
= A
1 T
A 1:
3.5. POWER
3.5
53
Power
Denition 3.5.1. Let A be a square matrix. We dene the nonnegative integer powers of
A to be
A0 = I An = AA
| {z A}(n > 0):
= A
1 n
1
1
=A
| A {z
A }1 (n > 0)
1 1
0 1
. Find A2 ; A3 ; A4 ; A 2 ; A 3 .
Solution
3.6
Determinants
2 matrix A =
a b
c d
bc 6= 0. This
a b
.
c d
It is denoted by the symbol det(A). The determinant is a function dened on square matrices.
For a general n n square matrix A, where n 3, we shall compute the det(A) inductively
via Cofactor Expansion.
special number ad bc is known as the determinant of the 2 2 square matrix A =
54
CHAPTER 3. MATRICES
3.6.1
Cofactors
n matrix.
(1) The (i; j)-minor of A is dened to be the determinant of the submatrix that remains
after the ith-row and the jth-column are deleted from A. It is denoted by Mij .
(2) The (i; j)-cofactor of A is the number ( 1)i+j Mij . It is denoted by Cij .
Let Mij be the determinant of the submatrix that remains after the ith-row and the jthcolumn are deleted from A.
The (i; j)th cofactor of A is the number ( 1)i+j Mij . It is denoted by Cij .
Note The value Mij is called (i; j)th minor of A.
Example 3.6.2. Consider the matrix
2
A=4
3
1 5 0
3 2 1 5
1 2 1
(a) Find the (1; 1)th cofactor and (2; 3)th cofactor of A.
(b) Calculate C12 and C31 .
Solution
3.6. DETERMINANTS
55
2 matrix A =
a b
c d
= ad
a b
c d
is ad
bc. We write it as
bc; or det(A) = ad
bc;
which can be obtained by computing the sum of the products on the rightward arrow
and subtracting the products in the leftward arrow.
To compute det(A), where A is n
n with n
3, we proceed as follows:
n
X
k=1
is det(A).
For an n
n matrix A, we have
det(A) = ai1 Ci1 + ai2 Ci2 +
ain Cin =
n
X
aij Cij
j=1
pansion.
3
1 5 0
3 2 1 5 by cofactor ex1 2 1
1 5 0
3 2 1
1 2 1
= ( 3)( 1)2+1
=
= 20:
5 0
1 0
1 5
+ (2)( 1)2+2
+ (1)( 1)2+3
2 1
1 1
1 2
56
CHAPTER 3. MATRICES
Note that using minors, we have
n
X
i=1
in which the sign of each consecutive terms alternates, with j being held at constant.
The "checkerboard matrix" S is a matrix with entries Sij = ( 1)i+j . The following are
examples of 3 by 3 and 4 by 4 checkerboard matrices.
3
2
2
3
+1
1 +1
1
+1
1 +1
6 1 +1
1 +1 7
7:
1 5 and S = 6
S = 4 1 +1
4 +1
1 +1
1 5
+1
1 +1
1 +1
1 +1
We may refer to the checkerboard matrix for the sign of ( 1)i+j when computing determinants.
For instance, if we compute determinant along third row, the checkerboard gives us the
signs f+; ; +g:
det(A) =
= 5
1 5 0
5 0
3 2 1 = (1)
2 1
1 2 1
2 + 17 = 20:
(2)
1 0
+ (1)
3 1
1 5
3 2
1. Select a column of A, say jth column. (So, we say that we perform cofactor expansion
along the jth-column.)
2. Multiply each entry akj of the selected row by its corresponding cofactor Ckj , i.e.,
akj Ckj .
3. Add all the resulting products obtained in the last step gives us the determinant of A,
i.e.,
n
X
det(A) = a1j C1j + a2j C2j +
+ anj Cnj =
akj Ckj :
k=1
3.6. DETERMINANTS
57
1 5 0
3 2 1
1 2 1
Solution Referring to the checkerboard, which gives us the signs f ; +; g along the second
column:
1
3
det(A) =
1
= ( 5)(
Determinant of 3
5 0
3 1
1 0
2 1 = 5
+2
1 1
1 1
2 1
4) + (2)(1) 2(1) = 20:
3 matrices
1
a11 a12 a13
3 matrix A = @ a21 a22 a23 A is
a31 a32 a33
1 0
3 1
(a)
2 1
3 5
1 6
(Ans
65)
(b)
2 7
5 1
3 8
(Ans 0)
4
7
2
6
2
4
58
CHAPTER 3. MATRICES
2
6
Example 3.6.8. Find the determinant of the matrix B = 6
4
1
0
0
1
1
3
1
0
expansion.
5
2
2
0
3
0
1 7
7 by cofactor
1 5
1
(Ans 12).
In general, one strategy for evaluating a determinant by cofactor expansion is to expand
along a row or column having the largest number of zeros.
(Ans
3.6.2
3
1
1
2
1
0
0
0
2
2
5
0
2
1
.
1
1
1).
2. If A has two rows (or columns) such that one of which is a multiple of the other, then
det(A) = 0.
3. Suppose A is a triangular matrix. Then det(A) = product of diagonal entries of A. In
particular if A is a diagonal matrix, then det(A) = product of diagonal entries of A.
Example 3.6.11. The determinants of the following matrices are zero.
2
6
6
4
4
5
0
3
3
9 8 5
2 0 7 7
7;
0 0 0 5
1
4
2
2
6
6
4
3
9 8 5
2 0 7 7
7:
8 2 1 5
1
4
2
4
5
6
3
(b)
4
0
0
0
0
2
0
0
0 5
0 7
3 0
0 21
(c)
a
0
0
0
0
b
0
0
0
0
c
0
0
0
0
d
3.6. DETERMINANTS
Theorem 3.6.13. For two n
59
n matrices A and B,
det(AB) = det(A)det(B):
Proof. From AA
1
:
det(A)
= I, we have
det(AA 1 ) = det(I); i.e. , det(A)det(A 1 ) = 1:
1
.
det(A)
60
CHAPTER 3. MATRICES
3.7
Adjoint of A (Optional)
In this section, we dene an adjoint of a square matrix A, and obtain an important result
relating A and its adjoint. From this result, we derive the equivalent statement for invertibility of a matrix A and nonzero determinant. It also gives a formula for the inverse of an
invertible matrix.
Denition 3.7.1. Let A be an n n matrix, and Cij be its (i; j)th cofactor. Then the matrix
2
3
C11 C12
C1n
6 C21 C22
C2n 7
6
7
6
7
6
7
6
7
6
7
4
5
Cn1 Cn2
Cnn
is called the matrix of cofactors from A.
3
1 5 0
3 2 1 5
1 2 1
Solution
(a) C11 = ( 1)1+1
2 1
2 1
= 0; C12 = ( 1)1+2
3 1
1 1
= 4; C13 = ( 1)1+3
3 2
1 2
8:::
So we have
2
C=4
(b)
0
5
5
4
1
1
3
2
8
3 5 & adj(A) = 4
17
0
4
8
5
1
3
3
5
1 5:
17
3
20 0 0
Aadj(A) = 4 0 20 0 5 = 20 I:
0 0 20
Denition 3.7.3. Let A be an n n matrix, and Cij be the cofactor. The transpose of the
matrix of cofactors is called the adjoint of A and is denoted by adj(A).
That is,
(adj(A))ij = Cji :
2
3
1 5 0
Example 3.7.4. (a) Find the adjoint of A=4 3 2 1 5. (b) Evaluate A(adj (A)) :
1 2 1
61
C=4
0
5
5
4
1
1
2
adj(A) = C T = 4
(b)
0
4
8
3
8
3 5
17
5
1
3
3
5
1 5:
17
3
20 0 0
Aadj(A) = 4 0 20 0 5 = 20 I:
0 0 20
Proof. (Optional. A proof is included at the end of this chapter for students who are keen
to read.)
The next result follows from the above theorem.
Theorem 3.7.6. Let A be an n
det(A) 6= 0.
3.8
1
adj(A):
det(A)
3
1 5 0
3 2 1 5 via adjoint.
1 2 1
Cramers Rule
3
2
62
CHAPTER 3. MATRICES
which can be expressed as a matrix equation:
0
1
x1
7
2
5 @
x2 A =
3
1
1
x3
3
2
For a linear system Ax = b whose coe cient matrix A is invertible, there is a formula
for its solution. The formula is known as Cramers rule. It is useful for studying the
mathematical properties of a solution without the need for solving the system.
Example 3.8.1. The linear system
2u
v + w = 3
u + v
3w = 5
5u
4v + 9w = 4
is equivalent to
2
@ 1
5
1
1
4
10
1 0 1
1
u
3
3 A@ v A = @ 5 A:
9
w
4
6
6
Aj = 6
4
a11 a12
a21 a22
..
..
.
.
an1 an2
det(Aj )
; j = 1; 2; : : : ; n
det(A)
a1j
a2j
..
.
b1 a1j+1
b2 a2j+1
..
..
.
.
anj
bn anj+1
3
a1n
a2n 7
7
7;
5
ann
the matrix obtained by replacing the entries in the jth column of A by the entries in the
matrix
2
3
b1
6 b2 7
6
7
b = 6 .. 7
4 . 5
bn
Example 3.8.3. For each of the following linear system, determine whether Cramers Rule
is applicable. If it is determine, use it to solve the linear system. If it is not, use other
method to solve the system.
(a)
7x1
2x2 = 3
3x1 + x2 = 5
63
(b)
2a + 4b = 3
3a + 6b = 5
(c)
2u
v + w = 3
u + v
3w = 5
5u
4v + 9w = 4
64
CHAPTER 3. MATRICES
3.9
Proofs (Optional)
n
X
aik Cik :
k=1
What happens if we have selected two dierent rows, and compute the product of entries
(of a selected row) and cofactors from dierent row(the other selected row)? That is, when
i 6= j, what is the value of
n
X
aik Cjk ?
k=1
We shall see from the next example that the sum of such products will be zero.
2
3
a11 a12 a13
6 a21 a22 a23 7
7
Example 3.9.1. Consider the matrix 3 3 matrix A = 6
4 a31 a32 a33 5.
Consider cofactors of entries along second row:
C21 =
(a12 a33
(a11 a32
a12 a31 )
aik Cjk = 0:
k=1
Proof. (Optional.) Suppose i 6= j. Let A0 be the matrix which has the same row as A except
the jth row. The jth row of A0 is the ith row of A. That is, A0 has two identical rows,
namely ith row and jth row.
65
Then we have
n
X
k=1
6
6
6
6
Aadj(A) = 6
6
6
4
a11 a12
a21 a22
..
..
.
.
ai1 ai2
..
..
.
.
an1 an2
a1n
a2n
..
.
2
7 C11 C21
7
7 6 C12 C22
76
76
..
ain 7 4 ...
.
7
.. 5 C
C
1n
2n
.
ann
Cj1
Cj2
..
.
Cn1
Cn2
..
.
Cjn
Cnn
3
7
7
7
5
+ ain Cjn
1
(A)adj(A).
det
1
(A)(jth-row
det
1
(A)adj(A)b:
det
of adj(A))b.
determinant of Aj where
3
a1j+1
a1n
7
a2j+1
a2n
7
..
.. 7 ;
.
. 7
7
5
anj+1
ann
66
CHAPTER 3. MATRICES
the matrix obtained by replacing the entries in the jth column of A by the entries in the
matrix
2
3
b1
6 b2 7
6
7
b = 6 .. 7 :
4 . 5
bn