Complex Numbers - Introduction PDF
Complex Numbers - Introduction PDF
Complex Numbers - Introduction PDF
supplementary notes
ucsc
1.
One of the fundamental properties of the real numbers is that the square of a real
number is always nonnegative. I.e., if x is a real number, then x2 0. This implies,
among other things, that certain quadratic equations dont have real solutions. In
particular, the equation x2 = 1, has no solution in real numbers. This doesnt
mean however, that the equation cannot have a solution.
(1.1)
i = 1.
With the set of imaginary numbers in hand, we can find a square root for every real
number, positive or negative. Indeed, following the usual algebraic rules, we have
(bi)2 = b2 i 2 = b2 (1) = b2 ,
for any real number b. Now, if < 0 and b =
real square root), then b2 = , so
bi =
Thus, for example,
p
|| (remember, || > 0 so it has a
More is true. By combining real and imaginary numbers, we can solve any quadratic equation.
Example 1.1. Solve the equation x2 + 2x + 2 = 0. Using the quadratic formula,
we find the two solutions
2 + 4 8
2 4 8
z1 =
= 1 + 2i and z2 =
= 1 2i.
2
2
Note that the two solutions are neither real numbers, nor are they purely imaginary.
1
Definition 2.
A complex number is
a number of the form z = a + bi, , where a and b are
real numbers, and i = 1. The (real) numbers a and b are called the real and
imaginary parts of z, respectively, and we often use the notation
a = Re(z)
and
b = Im(z).
ax2 + bx + c = 0
has no real solutions. But, as in Example 1.1, there are always complex solutions
and these solutions come in conjugate pairs. Namely, the solutions to equation (1.2)
are given by + i and i, where (from the quadratic formula)
p
|b2 4ac|
b
and =
.
=
2a
2a
Observe that if z = a = a + 0 i is a real number, then z = a 0 i = a = z.
In other words, real numbers are their own complex conjugates. Another thing to
note is that z = z for any complex number z (check!), that is, the conjugate of the
conjugate is the original number. Well see a geometric interpretation of these two
observations a little later.
Comments:
a. Complex numbers first appeared explicitly in the work of the 16th century Italian
mathematician Cardano. The term imaginary number was introduced later by
Descartes who did not think highly of the concept. His opinion notwithstanding,
complex numbers have important real applications, as we will see.
b. Introducing a new number as the solution of an equation that didnt already
have a solution among the accepted set of numbers was not new, even in the 16th
century. The irrational number 2 was introduced around 2500 years ago, when
it became apparent that the equation x2 = 2 had no rational solutions.
Exercises
1.1. Find the solutions of the equation 2x2 + 4x + 5 = 0.
1.2. Find the real and imaginary parts of the solutions of the equation x2 +3x+5 = 0.
1.3. Show that z = z for any complex number z.
2.
Complex arithmetic.
Complex numbers may be added and multiplied, just like real numbers, and the
usual properties (commutativity, associativity and distributivity) continue to hold.
2.1 Addition and subtraction.
To add two complex numbers, we add their real and imaginary components separately and use the distributive rule bi + di = (b + d)i for the imaginary parts. In
other words,
(a + bi) + (c + di) = (a + c) + (b + d)i.
Subtraction is defined similarly:
(a + bi) (c + di) = (a c) + (b d)i.
Example 2.1.
and
(u + v) + w = u + (v + w).
Likewise,
z z = (a + bi) (a bi) = 2bi = 2Im(z)i.
2.2 Multiplication.
To multiply the numbers a + bi and c + di, we use the familiar foil rule from
elementary algebra while keeping in mind that i 2 = 1, and collecting real and
imaginary terms:
(2.1)
Example 2.2.
and
z3 z1 = (3 + 5i)(3 + 2i)
= 9 6i + 15i + 10i 2
= (9 + (10)) + (15 + 10)i
= 19 + 25i.
When one of the two factors is real, then multiplication is particularly simple,
since real numbers have no imaginary parts. E.g., 7 z2 = 7(4 3i) = 28 21i.
As with addition, the usual properties of commutativity and associativity hold
for multiplication of complex numbers, as you can verify for yourself (see exercises).
And, multiplication distributes over addition, as usual:
(a + bi) (c + di) + (e + f i) = (a + bi) (c + e) + (d + f )i
= a(c + e) b(d + f ) + a(d + f )) + b(c + e) i
= (ac + ae bd bf ) + (ad + af + bc + be)i
= (ac bd) + (ad + bc)i + (ae bf ) + (af + be)i
= (a + bi)(c + di) + (a + bi)(e + f i).
2.3 Division.
Division is properly thought of as multiplication by the inverse, where the inverse
of a complex number z is the complex number z 1 satisfying z z 1 = 1. As with
the real numbers, a complex number has a multiplicative inverse if and only if it is
not zero. Finding the multiplicative inverse is simply a matter of solving a pair of
linear equations in two unknowns.
Specifically, given a (nonzero) complex number a + bi, we want to find a complex
number x + yi such that (a + bi)(x + yi) = 1. Using the rule for multiplication of
complex numbers this gives
(ax by) + (ay + bx)i = 1 = 1 + 0i,
which reduces to a pair of linear equations in the variables x and y:
ax by = 1
bx + ay = 0.
(2.2)
(2.3)
Multiplying equation (2.2) by a and equation (2.3) by b and adding the results
together gives
(a2 x aby) + (b2 x + aby) = a + 0 = (a2 + b2 )x = a =
x=
a2
a
.
+ b2
Note that division by (a2 + b2 ) is allowed since a + bi 6= 0, which means that either
a 6= 0, b 6= 0 or both, so a2 + b2 > 0.
In similar fashion, multiplying equation (2.2) by b and equation (2.3) by a and
adding the results together gives
(abx + b2 y) + (abx + a2 y) = b + 0 = (a2 + b2 )y = b =
y=
a2
b
.
+ b2
a2
b
a
2
i.
2
+b
a + b2
and
Compute
z1
z1
and , for the numbers z1 , z2 and z3 from Example
z2
z3
z1
3 + 2i
12 6 9 8
6
17
=
=
+
i=
i,
z2
4 3i
25
25
25 25
z1
3 + 2i
9 + 10 15 6
1
21
=
=
+
i=
i.
z3
3 + 5i
34
34
34 34
Exercises
2.1. Let u = 3 + 2i and v = 5 2i. Compute u + v, u v, u/v and v/u.
2.2. Compute the following products, quotients and powers:
(a) (1 + 4i) (2 3i) =
(b) 2/(3 + i) =
(c) (1 + i)3 =
(d) (2 i)2 =
2.3. Compute (1 + i)1 and (3 4i)1 .
2.4. Let =
1
2
1 i.
2
n
3.
(copy of)
u+v
(copy of)
There is a coordinate-free description of addition as well. The sum u+v is the complex number
such that the triangle with vertices 0, u and v in the complex plane is congruent to the triangle
with vertices u, v and u + v.
z=a+bi
bi
|z|=(a2+b2)1/2
since i 2 = 1.
Using the modulus and the conjugate, we can also shorten the expression for z 1 .
Namely, if z 6= 0, then
z 1 =
(3.1)
z
,
|z|2
as you can check by comparing to Proposition 2.2. It is also important to note that
the modulus is a multiplicative function. I.e., the modulus of a product is equal to
the product of the moduli.
Proposition 3.2. For any complex numbers z and w, |z w| = |z| |w|.
Proof: It follows from Propositions 3.1 and 2.3 and the commutative property of
multiplication that
|z w|2 = (z w) (z w) = z w z w = (z z) (w w) = |z|2 |w|2 .
Taking square roots of both sides then gives
p
p
p
p
|z w| = |z w|2 = |z|2 |w|2 = |z|2 |w|2 = |z| |w|,
as claimed.
Proposition 3.3. If z =
6 0 then z 1 = (|z|)1 .
Proof:
See exercises.
10
z=a+bi
bi
+4
+2
-a
a
+
-bi
-z=-a-bi
b. If a + bi is in the second quadrant, then /2 < arctan(b/a) < 0, and the correct
value of the argument is + arctan(b/a). In Figure 4, this corresponds to the
case of z2 , in which case arctan(b/ a) = (in red), but arg(z2 ) = . I.e.,
A2.
c. If a + bi is in the third quadrant, then 0 < arctan(b/a) < /2, and the correct
value of the argument is once again given by + arctan(b/a). In Figure 4,
this corresponds to the case of z3 , in which case arctan(b/ a) = , but
arg(z3 ) = + . I.e.,
By specifying that the angle fall in this range, we are technically defining what is called the
principal branch of the arctangent function. These are the values produced by (most) calculators.
11
z2
z1
bi
-a
a
-
+
2-
-bi
z3
z4
d. If a + bi is in the fourth quadrant, then /2 < arctan(b/a) < 0, and the correct
value of the argument is given by 2 + arctan(b/a). In Figure 4, this corresponds
to the case of z4 , in which case arctan(b/a) = , but arg(z4 ) = 2 . I.e.,
A4.
12
arg(u)0.9273 rad.s
-3
-2
-1
10
11
12
13
14
15
-1
arg(v)5.8884 rad.s
-2
-3
-4
-5
-6
13
rsin
rcos
Figure 6. Proof of Proposition 3.4.
Proof: Let z1 = r(cos + sin i) and z2 = (cos + sin i), then it follows from
equation (2.1) and the rules for sine and cosine of a sum of angles that
z1 z2 = (r cos )( cos ) (r sin )( sin ) + (r cos )( sin ) + (r sin )( cos ) i
= r (cos cos sin sin ) + (cos sin + sin cos )i
= r cos( + ) + sin( + )i ,
Which implies that + is equal to (one of the possible values of) arg(z1 z2 ).
Comments:
a. If arg(z1 ) and arg(z2 ) are the principal values of z1 and z2 and if their sum is
greater than 2, then we need to subtract 2 from arg(z1 ) + arg(z2 ) to obtain
the principal value of arg(z1 z2 ).
b. Some authors use the shorthand cis() = cos + sin i, but we shall not. Instead,
well use exponential notation.
Combining Propositions 3.2 and 3.5, we obtain a simple geometric characterization of multiplication by a (nonzero) complex number.
Proposition 3.6. If z 6= 0 and w is any complex number, then multiplying w by z
rotates w by the angle arg(z) and scales (stretches or shrinks) the modulus of w by
the factor |z|.
Example 3.2. Returning to the numbers u and v from Example 3.1, and using
formula (2.1), we have
uv = (36 + 20) + (15 + 48)i = 56 + 33i.
14
It follows that
|uv| =
and
arg(uv) = arctan(33/56) 0.5325.
Now, |u||v| = 5 13 = 65, as it should, but
arg(u) + arg(v) 0.9273 + 5.8884 = 6.8157 6= 0.5325.
This discrepancy is immediately cleared up, however, when we subtract 2 (
6.2832) from the sum of the arguments, as you should verify.
3.5 Exponential notation.
The polar coordinate representation of a complex number, equation (3.2), can be
simplified by using Eulers formula:
(3.3)
cos + sin i = ei ,
where is real and ez is the familiar exponential function defined by the Taylor
series,k
X
zn
.
(3.4)
ez = 1 +
n!
n=1
In most calculus classes, the variable in the series is assumed to be a real number.
But the operations of addition and multiplication make perfectly good sense for
complex numbers too and it is not hard to show that the series above converges for
every complex number z. Furthermore, the resulting function has all of its familiar
properties, most notably
(3.5)
and
ez1 z2 = (ez1 )z2
for all complex numbers z1 and z2 .
Using Eulers formula (3.3), we can restate Proposition 3.4 as
Proposition 3.7. If |z| = r > 0 and = arg(z), then
(3.6)
z = r ei .
Proposition 3.5 (the fact that the argument of a product is equal to the sum of the
arguments) can now be justified by using property (3.5) of the exponential function:
if z1 = r1 e1 i and z2 = r2 e2 i , then
z1 z2 = r1 ei 1 r2 ei 2 = r1 r2 e1 i+2 i = r1 r2 e(1 +2 )i .
Leonhard Euler (pronounced oiler), 1707 1783, is one of the most prolific and important
mathematicians of all time. His work has influenced (almost) every field in mathematics and
mathematical physics.
k
If Taylor series are unfamiliar or they fill you with a vague sense of dread, then you may skip
the explanation that follows and accept formula (3.3) and Proposition(3.6) on faith.
15
Note that cos(2) = cos(0) = 1 and sin(2) = sin(0) = 0, and this together with
Eulers formula (3.3) implies that
e2i = 1.
Furthermore, since ek = (e )k the previous identity generalizes to
k
(3.7)
e2ki = e2i = 1k = 1
for any integer k.
Combining the identities (3.7) and (3.4) proves the following useful fact.
Proposition 3.8. For any number and any integer k,
e(+2k)i = ei .
The rest of this subsection explains Eulers formula. You may skip it without
loss of continuity, but I recommend that you at least skim it.
To understand Eulers formula, we set z = i in the series on the right-hand side
of (3.4) (with a real number ), and manipulate the series a little as follows.
To begin, it helps to recognize that the powers of i follow a simple pattern, namely
i 0 = 1, i 1 = i, i 2 = 1, i 3 = i, i 4 = 1, i 5 = i, i 6 = 1, i 7 = i, etc. In other
words, for any integer n,
(3.8)
and if n = 2k + 1, we have
(3.9)
Next, we substitute z = i in the Taylor series for ez in (3.4), split the series into
two sub-series based on the parity (evenness or oddness) of the power n, and use
the formulas (3.8) and (3.9) to simplify.
16
ei = 1 +
X
(i)n
n=1
X
(i)2k
k=0
(3.10)
n!
X
(i)2k+1
+
(2k)!
(2k + 1)!
k=0
X
(1)k 2k
k=0
(2k)!
X
(1)k 2k+1
k=0
(2k + 1)!
i.
At this point, we are done because the series on the left-hand side of line (3.10) is
the Taylor series for cos , and the series on the right-hand side is the Taylor series
for sin , showing that
ei = cos + sin i,
as claimed.
Exercises
3.1. Compute the moduli and arguments of the numbers
z1 = 1 i,
z2 = 3 4i,
z3 = 8 + 6i,
z4 =
3/2 + 0.5i.
3.2. Compute the arguments and moduli of (z1 z2 ), (z1 z4 ) and (z2 z3 ), with
z1 , z2 , z3 and z4 as above.
3.3. Express the four numbers in exercise 3.1 using exponential notation.
3.4. Show that multiplication by i corresponds to counterclockwise rotation by an
angle of /2.
3.5. Show that if z 6= 0, then z 1 = |z|1 .
3.6. Show that ei + 1 = 0.
3.7. Show that if z 6= 0, and 0 arg(z) < , then arg(z) = arg(z) + . How
does this claim need to be adjusted if arg(z) < 2, assuming that we want the
principal value of arg(z)?
3.8. Show that ei = 1 for any real number .
4.
Roots of polynomials.
P (z) = 0,
17
P (z) = cn (z 1 ) (z 2 ) (z n ),
This claim can be proved using long division of polynomials. See the exercises.
18
In the last few sections below, well make a couple of general observations about
polynomials with real coefficients and then look at some simple types of polynomials
whose roots are relatively easy to find.
4.2 Polynomials with real coefficients.
Since the real numbers are a subset of the complex numbers, it follows that any
polynomial with real coefficients also has a root, and such polynomials may be
factored as in equation (4.2). On the other hand, even though the coefficients of
the polynomial may be real, there is no guarantee that its roots will be real, as in
Example 1.1.
The complex roots of a real polynomial do have a special property however: they
occur in complex conjugate pairs, as proved below.
Proposition 4.3. If a0 , a1 , . . . , an are real numbers and is a root of the polynomial
P (z) = an z n + + a1 z + a0 , then is also a root of P (z).
Proof: is a root of P (z) if and only if an n + + a1 + a0 = 0. Taking
conjugates of both sides of this equation gives
(4.3)
an n + + a1 + a0 = 0 = 0,
since 0 is real. Now, applying Proposition 2.3 (and exercise 2.7) to the left-hand
side of equation (4.3) gives
an n + + a1 + a0 = an n + + a1 + a0
= an ()n + + a1 () + a0
= an ()n + + a1 () + a0 ,
(4.4)
since the coefficients a0 , . . . , an are real. Finally, replacing the left-hand side of (4.3)
by the right-hand side of (4.4) shows that
P () = an ()n + + a1 () + a0 = 0,
so that is also a root of P (z), as claimed.
P (z)
= z 2 1,
z 2 2z + 2
so that
P (z) = (z (1 + i)) (z (1 i)) (z 2 1) = (z (1 + i)) (z (1 i)) (z 1)(z + 1),
which means that the roots of P (z) are 1 = 1 + i, 2 = 1 i, 3 = 1 and 4 = 1.
19
Exercise.
Exercise.
and
arg(2 ) = arg(1 ) + .
First, we have
22 = (1 )2 = 12 =
p 2
|u| ei(/2+/2) = |u|ei = u,
20
1
1
i = ei/4 = cos(/4) + sin(/4)i = + i.
2
2
Next, we have | 2i| = 2 and arg 2i = 3/2, so
2i = 2e3i/4 = 2 cos(3/4) + sin(3/4)i = 1 + i.
3 + 4i = 5e1.10715i = 5 cos(1.10715) + sin(1.10715)i = 1 + 2i.
To solve a quadratic equation with complex coefficients, we use the familiar quadratic formula.
Proposition 4.7. The solutions of the equation az 2 + bz + c = 0 are given by
b b2 4ac
b + b2 4ac
and z2 =
,
(4.7)
z1 =
2a
2a
where a, b and c may be any complex numbers, as long as a 6= 0.
The proof is identical to the case in which the coefficients are real (i.e., completing
the square), because none of the steps make any use of the nature of the coefficients
(real or complex), and I leave it to you as an exercise.
Example 4.3. Find the solutions of the equation z 2 + (1 2i)z 2i = 0.
Applying the quadratic formula (with a = 1, b = 1 2i and c = 2i) we find
that the two solutions are
p
p
(1 2i)2 (1 2i)2 + 8i
(1 2i) + (1 2i)2 + 8i
and 2 =
.
1 =
2
2
Next, note that (1 2i)2 + 8i = 3 4i + 8i = 3 + 4i, so the two solutions simplify
to
1 + 2i + 1 + 2i
1 + 2i 1 2i
1 =
= 2i and 2 =
= 1,
2
2
as you can check.
4.4 The nth roots of a complex number.
If n is a positive integer and > 0 is a real number then the nth root of is
defined
to be the positive real number satisfying n = , which we denote by
21
(4.8)
Now, for a general polynomial of degree n, there is no guarantee that the s that
appear in the factorization (4.2) are all different from one another. But in this case,
i.e., in (4.8), the roots are distinct.
Proposition 4.8. If u 6= 0, = arg(u) and n is a positive integer, then the solutions
of the equation z n = u are given by
p
(4.9)
k = n |u|e(+2k)i/n ,
for k = 0, 1, 2, . . . , n 1. Furthermore, these n numbers are all distinct from one
another.
Proof: To see that kn = u for each integer k, between 0 and n 1, we simply
evaluate:
p
n
kn = n |u|e(+2k)i/n
p n
= n |u| en(+2k)i/n
= |u| e(+2k)i
= |u|ei
= u,
where the transition from line 3 to line 4 is justified by Proposition 3.8.
Now suppose that 0 j < k n 1, then
p
n
kj
|u|e(+2k)i/n
k
e(+2k)i/n
= p
=
= e( n )2i ,
(+2j)i/n
n
j
e
|u|e(+2j)i/n
as you should verify (using the fact that e /e = e ). It follows that
arg(k /j ) =
kj
2.
n
1 = e2i/4 = i,
2 = e4i/4 = 1,
3 = e6i/4 = i;
22
1
3
1
3
4i/6
0 = e
= 1, 1 = e
= +
i, 2 = e
= +
i, 3 = e6i/6 = 1
2 2
2 2
3
3
1
1
8i/6
10i/6
4 = e
=
i and 5 = e
=
i.
2
2
2
2
The arguments of the nth roots of unity increase from 0 to 2(n 1)/n in increments of 2/n, and the roots of unity themselves are equally spaced points on the
unit circle in the complex plane. This is illustrated below, where the 10th roots of
unity are marked with blue dots on the unit circle.
0i/6
2i/6
e3i/5
e2i/5
ei/5
e4i/5
/5
1=e0i/5
e5i/5=-1
e9i/5
e6i/5
e7i/5
-i
e8i/5
k = n || e2ki/n = n nk ,
for k = 0, . . . , n 1, since || = in this case.
k
n is called a primitive nth root of unity and not the primitive nth root of unity because for
any n, there are other nth roots of unity with the same property. E.g., it is possible to represent
the 5th roots of unity as powers of e4i /5 .
23
Exercises
4.1. Find the solutions of the quadratic equations below. Express your answers in
rectangular coordinates (i.e., in the form a + bi).
(a) z 2 + iz + 1 = 0.
(b) 2z 2 3iz + 2 = 0.
(c) iz 2 + 2z 1 = 0.
4.2. Write down the 8th roots
of unity using rectangular coordinates. Do not round
your answers, instead use 2 as needed. Hint: first express your answer in polar
coordinates.
4.3. Write down the 12th roots
of unity using rectangular coordinates. Do not round
your answers, instead use 3 as needed. Hint: first express your answer in polar
coordinates.
4.4. Prove Proposition 4.4.
4.5. Prove Proposition 4.5.
4.6. Use Proposition 4.5 to show that a polynomial with real coefficients and odd
degree always has at least one real root.
4.7. Show that if n is even then 1 is an nth root of unity and if n is divisible by 4,
then i is an nth roots of unity.
4.8. Show that if i is an nth root of unity, then n is divisible by 4.
4.9. Show that if is an nth root of unity, then so is k for any integer k.
4.10. Show that if is a root of the polynomial P (z) = cn z n + + c1 z + c0 , then
P (z) = (z ) P1 (z),
where P1 (z) is a polynomial of degree n 1.
Hints: If you use long division to divide (z ) into P (z) you obtain an identity
of the form
P (z) = (z ) P1 (z) + R(z),
where R(z) is a polynomial whose degree is less than the degree of (z ). What
is the degree of (z ) and what is the nature of a polynomial whose degree is less
than this? Finally, use the fact that P () = 0.