You are on page 1of 91

Dr Janet Semester 2, 2020/21

CALCULUS III
Chapter 1:
COMPLEX NUMBERS &
FUNCTIONS
1.1 Introduction
1.2 The Argand diagram
1.3 The arithmetic of complex numbers
1.4 Polar form of a complex number
1.5 Euler’s formula
1.6 Powers and roots of complex numbers
1.7 Circular and hyperbolic functions
1.8 Logarithm of a complex number
1.9 Complex functions and mappings
1.10 Complex differentiation. Cauchy-Riemann
equations
1.11 Conjugate and harmonic functions
1.12 Power series. Taylor series
1.13 Laurent series
2
1.1 Introduction
We are familiar with “ordinary” numbers.
E.g. 1,  2, 3 /10, 2.634,  , e, 2, etc.
Mathematicians call these real numbers.
They can be plotted on a ‘real axis’ or ‘number line’.

We know - how to add, subtract, multiply, divide them.


- they can be positive, negative or zero.
- for any real number x, x2  0.
In this chapter we consider a new set of numbers
whose square is not necessarily positive!
These are called complex numbers. 3
The Imaginary Number
Fundamental to complex numbers is the symbol j which
has the property j 2  1.
j is called the imaginary number*.
• Don’t worry about the fact that j doesn’t exist!
• We just define j then use it (in a similar way to real
numbers) … and it is found to be extremely useful in
a wide range of practical applications! e.g. a.c.
circuits, fluid flow, heat transfer, signal processing.
We assume we can write j  1
Hence 4 
7 
By the usual rules of algebra, j 
3

j 
4

*Note: Sometimes i is used instead of j. 4


Definitions
A complex number is an ordered pair of real numbers,
usually denoted by z (or w, etc.).
For example, if x and y are real numbers then we can
have a complex number z  x  jy where j 2  1.

Examples:

For a complex number z  x  jy,


x is the real part of z, written x  Re z
y is the imaginary part of z, written y  Im z

E.g. for z1  3  2 j, Re z1  3, Im z1  2
If x = 0 then z = yj is said to be pure imaginary. 5
1.2 The Argand Diagram
A complex number z  x  jy involves a pair of real
numbers x,y, so is naturally represented by a point in a
plane. We take the usual Oxy plane but call it the complex
plane. Ox is called the real axis, Oy the imaginary axis.
A plot of the complex plane is called an Argand diagram.

E.g. z = 2 + 3j is represented by the point (2, 3).


w = -1+ j by (-1, 1).

6
1.3 Complex Arithmetic
Consider two complex numbers: z1  x1  jy1
z2  x2  jy2
Equality
z1 = z2 if and only x1 = x2 and y1 = y2.
I.e. Two complex numbers are equal if and only if their real
parts are the same and their imaginary parts are the same.

Addition & Subtraction


z1  z2  ( x1  x2 )  j ( y1  y2 )
z1  z2  ( x1  x2 )  j ( y1  y2 )
Note: this is just like for vectors!
7
Example 2 Given z  3  2 j and w  2  j,
a) Find z  w, b) Find z  w, c) Find 2w,
d) Plot z, w, 2w, z + w and z – w on an Argand diagram

8
Multiplication
Multiplication is done in the obvious way: multiply all the
terms together and replace j2 with -1 wherever it occurs.
I.e. for z1  x1  jy1 and z2  x2  jy2
z1 z2  ( x1  jy1 )( x2  jy2 )
 x1 x2  jx1 y2  jx2 y1  j 2 y1 y2
 ( x1 x2  y1 y2 )  j ( x1 y2  x2 y1 )
E.g. for z  3  2 j, w  2  j we have
zw =

9
Example 3 Given z  2  3 j and w  1  j , find
a) z  2w

b) zw

2
c) z

10
Conjugate
The conjugate of a complex number z  x  jy is
z  z*  x  jy
E.g. for z1  2  3 j, z1  2  3 j

On an Argand diagram,
where is z* in relation to z?

11
Example 4 For z  x  y j , find
a) 1 ( z  z )
2
1 (z  z )
b) 2 j

c) zz
2
d) z

12
Note that z2 may be negative or complex.
But zz is always real and non-negative:
zz  ( x  yj )( x  yj )  x 2  jxy  jxy  j 2 y 2
x y
2 2

Also for any z and w,


zz (z  z )
Re( z )  and Im( z ) 
2 2j

 w  w
z z
z  w  z  w, zw  z w,

13
Modulus
The modulus of a complex number z  x yj is
z  x 2  y 2  zz .
|z| is always a non-negative real number.
E.g. for z1  2  3 j, | z1 | 22  32  13

On an Argand diagram, what


is the significance of |z|?

14
Division
Consider z1  x1  y1 j
z2 x2  y2 j
We can simplify this by rationalizing: multiply top
and bottom by z2* (to obtain a real denominator):
z1 x1  y1 j x2  y2 j x1 x2  y1 y2  j ( x2 y1  x1 y2 )
  
z2 x2  y2 j x2  y2 j x22  y22
x1 x2  y1 y2 ( x2 y1  x1 y2 )
 j
x2  y2
2 2
x22  y22

Note: Remember the method not the formula!


15
Example 5 Given z  3  2 j and w  2  j, find
a) 1
z

b) z
w

16
Example 5, cont. Given z  3  2 j , w  2  j , find

c) 1
w

d) z  w
zw

17
Quadratic Equations and Complex Roots
We know az  bz  c  0 has solutions z 
2 b  b 2
 4ac
.
2a
If b2  4ac  0 , no real solutions exist … but complex
solutions always exist!
Example 6 Find the roots of the equations
a) z  16  0
2

b) z2  2z  3  0

18
Example 7
Find the roots of the equation z  z  1  0
2

and plot them on an Argand diagram.

19
Complex Numbers and Polynomials
In Examples 6 & 7 we see that the roots of the quadratic
are complex conjugates of each other.
In fact this is always the case, provided the coefficients of
the quadratic equation are real.
It can also be shown to be true for higher order
polynomials

Theorem
For a polynomial equation with real coefficients,
if any complex roots occur, the complex roots
always occur in complex conjugate pairs.

20
1.4 Polar Form of a Complex
Number
Consider again the Argand Diagram.
A point P can be specified by
P Cartesian coordinates (x, y)
or by polar coordinates (r, q)
where
x  r cos q, y  r sin q
y
r  x  y , tan q 
2 2 2

x
This gives an alternative form for representing a
complex number: z  x  y j  r (cosq  j sin q )
21
Notes
1) r is again called the modulus of z, since r = |z|.
2 q is called the argument of z, q = arg(z).
3) For a given x,y, the equation tan q  y / x has two
solutions in the range -π ≤ θ ≤ π. We must choose q
so that the point lies in the correct quadrant. (Finding
arctan( y / x) on a calculator gives the solution in the
range –π/2 ≤ θ ≤ π/2 which may or may not be correct!)
 When converting from Cartesian to polar form
you are strongly advised to plot the point!
4) The value of q in the range -π < θ ≤ π is called the
principal argument, Arg(z).
5) Any value (q + 2k), for integer k, will be equivalent.
I.e. the polar form is not unique.
6) Generally q may be given in radians or degrees. But
in calculus, radians must be used. 22
Example 8 Complete the following table

Cartesian form Polar form


a) 1+j
b) 3j
c) –2
d) – 3 + 4j
e) 2(cos /3 + j sin /3)
f) cos(–5/4) + j sin(–5/4)

23
Example 10
On an Argand diagram, sketch the curve or region defined:
a) Re( z )  1 e) | z  3 j |  2


b) Arg( z ) 
3

c) | z |  1

d) z  x  2 j

24
Complex Arithmetic Revisited
The polar form makes multiplication and division is easier.
Multiplication
Let z1  r1 (cos q  j sin q ) z2  r2 (cos   j sin  )
Then z1 z2  r1r2 (cos q  j sin q )(cos   j sin  )
 r1r2 [(cos q cos   sin q sin  )
 j (cos q sin   sin q cos  )]
Using the identities
cos(q   )  cosq cos   sin q sin 
sin(q   )  sin q cos   cosq sin 
gives z1 z2  r1r2 [cos(q   )  j sin(q   )]
i.e. we multiply the moduli and add the arguments.
25
Division
Similarly, it can be shown that for
z1  r1 (cos q  j sin q ) z2  r2 (cos   j sin  )
z1 r1
 [cos(q   )  j sin(q   )]
z2 r2
i.e. divide the moduli and subtract the arguments.

E.g. for w  2[cos( / 3)  j sin( / 3)]


z  cos( / 6)  j sin( / 6)
wz 
w
z 26
Multiplication and Rotations
Consider w  r (cos   j sin  ), z  cosq  j sin q .
Then wz  r[cos(q   )  j sin(q   )]
Argand Diagram:

We see the effect of multiplying by z is to rotate w


anticlockwise through the angle q.
Similarly, dividing by z rotates w clockwise by angle q.
27
Example 9 Let z1  2  2 j, z 2  3 j.
a) Write z1 and z2 in polar form and use the polar forms to
find z1z2 and z1/z2. b) Confirm your answers using the
Cartesian forms.

28
1.5 Euler’s Formula
(& Exponential Form of a Complex Number)
A third way of representing a complex number relies on
the important result known as Euler’s Formula:
e jq  cos q  j sin q
Using this we can write z  r (cos q  j sin q )  re jq .
Thus a complex number can be represented in 3 ways:
• Cartesian form: z  x  jy
• Polar form: z  r (cosq  j sin q )
jq
• Exponential form: z  re
You should be able to move confidently between these forms!
29
Proof of Euler’s Formula (1):
2 3 4
x x x
From Calculus II, e x  1  x     ...
2! 3! 4!
So ( jq ) 2
( jq ) 3
( jq ) 4
( jq ) 5
e jq  1  ( jq )      ...
2! 3! 4! 5!
q2 q3 q4 q5
 1  jq  j  j  ...
2! 3! 4! 5!
q2 q4 q3 q5
 [1    ...]  j [q    ...]
2! 4! 3! 5!
 cosq  j sin q
Since
2 4 6 3 5 7
x x x x x x
cos x  1     ..., sin x  x     ...
2! 4! 6! 3! 5! 7! 30
Proof of Euler’s Formula (2):

Let f (q )  cosq  j sin q


Then df
  sin q  j cos q  j (cos q  j sin q )  f j
dq
d
Also (e jq )  je jq
dq
jq
So f (q )  cos q  j sin q  e

Note: The exponential form of a complex number is very


easy to differentiate, integrate and combine with other
exponentials. Hence it is the form most commonly used in
physics and engineering.
31
Example 8 extended Complete the following table
Cartesian Polar form Exponential
form form
a) 1+j
b) 3j
c) –2
d) – 3 + 4j
e) 2(cos /3 + j sin /3)
f) cos(–5/4) + j sin(–5/4)
g) e j/6
h) 3e4j
32
Example 10
Given z  re jq and w  t e j , find the following.
(Are the results what you expect? Explain.)
a) z
b) zz

c) zw

1
d)
z

33
1.6 Powers & Roots of
Complex Numbers
jq jpq
By the laws of indices, (e )  e p
for any p.
Hence the exponential form of a complex number is
very convenient for finding roots and powers.
By Euler’s formula this can be rewritten as:
(cos q  j sin q )  cos pq  j sin pq
p

This is known as De Moivre’s Theorem.


De Moivre’s theorem provides a quick method for
proving trigonometrical multiple angle formulae.
34
Example 11
By considering e3jq, show that cos 3q  4cos3 q  3 cos q .

3 jq jq 3
We have e  (e )
So cos3q  j sin 3q  (cosq  j sin q )3
 cos3 q  3 cos2 q . j sin q  3 cosq ( j sin q )2  ( j sin q )3
 cos3 q  3 cosq sin2 q  j[3 cos2 q .sin q  sin3 q ]

Equating the real parts gives


cos3q  cos3 q  3 cosq (1  cos2 q )

Hence cos 3q  4cos q  3 cos q .


3

35
Powers
To find integer powers of complex numbers, we can
simply use jq p p jpq
z  (re )  r e
p

Example 12
Consider z = 1 + j. Find z5 and z9. (Give your
answers in the form a + jb.)

36
jq 1/ p
Roots What is 8 , (2  3 j )
1/3 1/4
or in general ( re ) ?
• These are the solutions of the equations
jq
z  8  0,
3
z 4
 2  3 j  0 or in general z p
 re  0.
• We expect an equation in zp to have p roots.
jq j (q  2 k )
• To find all the roots, remember that e  e
where k is an integer.
To find a pth root,
jq j (q  2 k )
take z  re  re with k = 0, 1, 2, …, p -1.
1/ p j (q 2k )/ p
Then z
1/ p
r e for all these values of k.

37
Example 13 Find the square roots of 1  3 j.

Note: you should find 2 distinct values for a square root, with
equal magnitude but opposite sign! 38
Example 14 Find the cube roots of 1. Plot your
answers on an Argand diagram.

Note that all the pth roots of a complex number lie on a circle in
the complex plane, separated by angles 2/p. 39
Example 15 Find the cube roots of 8j. Plot your
answers on an Argand diagram.

40
1.7 Circular and Hyperbolic
Functions
We have jq
e  cos q  j sin q
 jq
e  cos q  j sin q
So we can write 1 jq
cos q  (e e  jq
)
2
1
sin q  (e jq  e  jq )
2i

Trigonometric functions are related to the geometry of a


circle so are also called circular functions.
41
Hyperbolic Functions
In a similar way, certain combinations of ex and e-x
occur so frequently that they have been given
special names. We have the following definitions:
• The hyperbolic cosine function, cosh x,
(pronounced 'cosh') 1 x x
cosh x  (e  e )
2
• The hyperbolic sine function, sinh x,
(pronounced 'shine'), 1 x x
sinh x  (e  e )
2
So also
e x  cosh x  sinh x, e x  cosh x  sinh x
These functions are called hyperbolic functions
because they are related to the geometry of hyperbolae.42
Example 16
From the definitions, find
a) sinh(0)

b) cosh(0)

d
c) (sinh x)
dx

d
d) (cosh x)
dx

43
From these, further functions are defined:

Graphs:

Are these functions odd or even? 44


There are also hyperbolic identities such as:

Osborne’s Rule:
Hyperbolic function identities are obtained from
trigonometric identities by replacing sinq with sinhq
and cosq with coshq, except that every occurrence
of sin2q is replaced by –sinh2q. 45
Relationship of Hyperbolic & Circular Functions
Compare
1 jq
cos q  (e  e )
 jq
sin q
1 jq  jq
 (e  e )
2 2j
1 x x
cosh x  (e  e ) sinh x 
1 x
(e x
e )
2 2
We can see that
cosh( jz)  cos z sinh( jz )  j sin z
Also cos( jz)  cosh z sin( jz)  j sinh z

46
Example 16 Verify the identity cosh 2 x  sinh 2 x  1
a) From the definitions on slide 42.

b) From slide 46, given sin2z + cos2z = 1 for all z

47
1.8 Logarithm of a Complex
Number
Suppose we have a complex number z = re jq and
want to find its natural logarithm, say w = ln z.
The natural log is defined as the inverse of the
exponential function, that is by ew = z.
In general w will also be complex. Let w = u + vj.
Then u  jv jq
e  e e  re .
u jv

We see eu  r , v  q

Hence ln z  ln r  jq where r  z  0,
q  arg z. 48
Note
1. A real number has a real log: ln z = ln r.
2. The complex natural logarithm is not unique, it is
infinitely many valued! This is because arg(z) is not
unique: arg(z) = Arg(z) + 2k for any integer k.
• The principal value of ln z is Ln z  ln r  j Arg( z)
• The other values of ln z are ln z  Ln z  2k j
j ( /4 2 k )
E.g. for z  1  j  2 e
Ln z  ln 2  j /4  12 ln 2  j /4
ln z  12 ln 2  j ( /4  2k )

3. So also eln z  z but ln(e z )  z  2k j


49
Example 18 Complete the table below
Cartesian exponential
Ln z ln z
form form

a) 2

b) j

c) –4

d) –1 – j

e) 3 – 4j

50
1.9 Complex Functions & Mappings
Functions – Review from Calculus 1
A function f is a rule that assigns to each element in
some set D(f) exactly one element f(x) in a set R(f),
called the value of f at x.

It is like a machine:

And can be visualized as an arrow diagram:

We can say the function f maps the set D to the set R. 51


Complex Functions
Previously we studied real functions, so the independent
variable x and dependent variable y = f(x) were both real.
If the independent variable is complex, say z = x + yj,
then the function w = f(z) will generally be complex. It is
called a complex function.
Example
Consider w = f(z) = z2 + 2z
For z = x + yj then w  ( x  yj )  2( x  yj )
2

 ( x  2 x  y )  (2 xy  2 y) j
2 2

Note that w = u + jv = u(x, y) + jv(x, y).


Thus a complex function is equivalent to a pair of real
functions. 52
Complex Mappings
A complex function can be visualized as a mapping
between regions on two Argand diagrams:

[From James]

53
Example 20
Let w  f ( z )  2 jz  6 z .
a) Find u and v. b) Find the value of f at z = 1+ 2j

54
Example 21 Find the image of the line y = 2x + 4
(in the z-plane, z = x + jy) in the w-plane
under the mapping w = f(z) = 2z + 6.

55
Example 22
Find the image of the line z = x + j in the w plane
under the mapping w = f(z) = z2 + z.

56
Limits
Suppose a function f(z) is defined in a neighbourhood of
z0 (except possibly at z0). We say lim f ( z )  L if
z z0
f ( z )  L as z  z0 where z may approach z0 from
any direction in the complex plane.

Continuity
A function f(z) is said to be continuous at the point z = z0
if f(z0) exists, Lim f ( z ) exists and lim f ( z )  f ( z0 ).
z  z0 z z0

A function f(z) is said to be continuous in a domain if it is


continuous at every point in the domain.
57
1.10 Complex Differentiation.
Cauchy-Riemann Equations
The derivative of the complex function f(z) at the point z0
is defined by f ( z )  f ( z0 )
f '( z0 )  lim
z  z0 z  z0

or equivalently f '( z )  lim f ( z0  z )  f ( z0 ) where


z 0 z z  x  jy.
0

• The definition is very similar to that in real calculus so


the rules of differentiation are the same too.
• A point at which the derivative does not exist is called
a singular point of the function.
• If a function is differentiable throughout some region R of
the complex plane, it is said to be analytic in R. 58
For example, consider the function f ( z ) 
z
.
z j
• The function is singular at z = – j.

• Elsewhere, by the usual rules of differentiation


(quotient rule), we have
z jz j
f ( z )   .
z  j z  j
2 2

• So, for example,


j j 1
f (2 j )   2  j
2 j  j
2
9j 9
59
Cauchy-Riemann Equations
Note that the definition of f´(z0) involves a limit as z → z0
where z may approach z0 from any direction. This may
seem a very stringent requirement. Perhaps surprisingly,
there are many analytic functions.

We have the following important theorem:


Let z  x  yj and f ( z)  u( x, y)  v( x, y) j.
f(z) is analytic in a region if and only if in that region
u v u v
 and  .
x y y x
These are called the Cauchy-Riemann Equations.
60
Cauchy-Riemann Equations: Proof
Consider f '( z )  lim f ( z0  z )  f ( z0 ) where
0
z 0 z z  x  jy.

• Suppose we choose z  x.


Letting z  x  yj and f ( z)  u( x, y)  v( x, y) j,
u ( x0  x, y0 )  jv( x0  x, y0 )  u ( x0 , y0 )  jv( x0 , y0 )
f '( z0 )  lim
x0 x
u ( x0  x, y0 )  u ( x0 , y0 ) v( x0  x, y0 )  v( x0 , y0 )
 lim j
x0 x x
We see that these are the partial derivatives:
 u v 
f '( z0 )    j 
 x x  x x0 , y  y0 61
• Similarly choosing z  jy gives
u( x0 , y0  y)  jv( x0 , y0  y)  u( x0 , y0 )  jv( x0 , y0 )
f '( z0 )  lim
y 0 jy
1 u( x0 , y0  y)  u( x0 , y0 ) v( x0 , y0  y)  v( x0 , y0 )
 lim 
y 0 j y y
 v u 
  j 
 y y  x x0 ,y  y0

• f is analytic if and only if the limit f´(z0) is the same no


matter what path is followed. Hence we require
u v u v
 and  .
x y y x
62
Example 24
Determine whether the function f ( z )  z is analytic.

63
Example 25
Consider the function f ( z )  x  jy , z  x  jy. Find
2 2

the set of points z = x + jy at which f is differentiable.

64
Example 26
Show that the function f(z) = z2 is analytic everywhere
and find its derivative.

65
1.11 Conjugate & Harmonic
Functions
• A pair of functions u(x, y) and v(x, y), where x and y
are real variables, that satisfy the Cauchy-Riemann
equations are called conjugate functions.
• A function that satisfies Laplace’s equation in 2D is
called a harmonic function. That is, u(x, y) is
harmonic if  2u  2u
 2  0.
x 2
y
It is easily shown that if f ( z)  u( x, y)  v( x, y) j is analytic
then u and v are both harmonic. Therefore u and v are
harmonic conjugate functions. Such functions occur
frequently in electrostatics, fluid flow and other areas. 66
Proof
Suppose f ( z)  u( x, y)  v( x, y) j is analytic.

By the Cauchy-Riemann equations


u v u v
 and  .
x y y x
Differentiating again,
u v
2 2  2u  2v
 and  .
x 2
xy y 2
yx

So  2u  2u
 2 0
x 2
y

The result for v can be shown in a similar way.


67
Example 28
a) Verify that u ( x, y )  x 2  y 2  y is harmonic
everywhere.

68
Example 29
b) For u ( x , y )  x 2
 y 2
 y, find a harmonic conjugate
function v(x, y) and an analytic function f(z) = u + vj.

69
Example 30
Determine whether or not u ( x, y )  x  3 xy is harmonic.
3 2

If it is, find a conjugate function v(x, y) and an analytic


function f(z) = u + vj.

70
One special property of conjugate functions is that
curves u = constant and v = constant always intersect at
right angles, that is they are orthogonal curves.

Proof
u u
On the curve u(x, y) = constant, du  dx  dy  0
dy u u x y
So 
dx u x y
v v
On the curve v(x, y) = constant, dv  dx  dy  0
dy v  v x y
So 
dx v x y
By the Cauchy-Riemann equations
dy dy u v u v
  1 so the curves are
dx u dx v x x y y orthogonal. 71
Practical example: 2D electrostatics.
Equipotential curves (solid curves) and electric field
lines (dotted curves) are conjugate functions.

Equipotentials and field lines for an electric dipole 72


1.12 Power Series. Taylor Series
Calculus 2 introduced power series of real variables


n0
a n x n  a 0  a 1 x  a 2 x 2  a 3 x 3  ...  a n x n  ...

Or more generally


n0
a n ( x  c ) n  a 0  a 1 ( x  c )  ...  a n ( x  c ) n  ...

Remember that a power series is valid for | x  c |  R


where R is called the radius of convergence.
R is usually found using the ratio test: a series converges
n 1
for L < 1 where an 1 ( x  c )
L  lim .
n  an ( x  c ) n
73
Examples of Real Series

1
1) f ( x )    x  1  x  x  ...
n 2
for x 1
1  x n0

2) Binomial Series:
For any real number k and |x| < 1,

k n k ( k  1) 2 k ( k  1)( k  2 ) 3
(1  x )     x  1  kx 
k
x  x  ...
n0  n  2! 3!

where  k   k  k ( k  1)( k  2)...( k  n  1)


   1,   n 1
0 n n!

74
We can use a known series to find other (related) series.

E.g. 1
f ( x)  n
  x  1  x  x  ...
2
for x 1
1 x n0

Hence we can find



1 1
  f (  x 2
)  (  x 2 n
)
1  x 1  ( x )
2 2
n 0

 1  x 2  x 4  x6  ... for  x 2  1, that is x  1


And
1 1 1 x 1  x n  xn
  f ( 2 )   ( 2 )   n 1
2  x 2 1  x2  2 2 n 0 n 0 2
2 3
1 x x x x
     ... for  1, that is x  2
2 4 8 16 2 75
Note that by rearranging in different ways we get series
valid on different intervals.

 f ( 2 )   ( x2 )n
x x x x x
E.g.

2  x 2 1  x2  2 2 n 0
x
for  1, that is x  2
2
Meanwhile


 n
  2  
x 1 2 2
for  1, that is x  2
2 x 1 x n 0 x x

 x  x  1
x x n
 for x  1  1,
2  x 1  ( x  1) n 0
that is, 0  | x |  2 76
Complex Series
• A complex power series has the general form


n0
a n ( z  z 0 ) n  a 0  a 1 ( z  z 0 )  ...  a n ( z  z 0 ) n  ...

where z is a complex variable, the coefficients an are


constants (real or complex) and the centre of the series,
z0, is a fixed point in the complex plane.
• Again the power series is valid for | z  z0 |  R where
R is the radius of convergence. R can be found using
the ratio test but note that we must take the ratio of the
moduli of the complex terms. That is, a series
converges for L < 1 where an 1 ( z  z0 ) n1
L  lim .
n  an ( z  z0 ) n

77
Two particularly useful results, from which many
other series can be derived, are:

1
1.   z =1  z  z  z  ... for z  1
n 2 3

1  z n 0
y
[Note: This is also a Taylor series |z|=1
– see later.]
O x
2. Binomial series
k ( k  1) 2 k ( k  1)( k  2 )...( k  n  1) n
(1  z )  1  k z 
k
z  ...  z  ...
2! n!
for z  1
78
Example 32
1
Find a power series represention of f ( z ) 
1 z
valid for |z| < 1.

79
Example 33
1
Find a power series represention of f ( z ) 
z 3
in the given form in the given region.

a) |z| < 3;  n
a
n0
z n

80

an
b) |z| > 3; 
n 0 z n
[Note: this is a power series about z0 =  !
See James.]


c) |z – 2| < 1;  n
a
n0
( z  2) n

81
Example 34
Use the binomial series to find a power series
represention of the function f ( z )  1 centred at
z0 = 1 . ( z  2) 2

82
Taylor Series
For a real function f(x), the Taylor series expansion
of the function f at c is:

f (x)  
n0
an ( x  c)n where

I.e.

The special case c = 0 is called the Maclaurin Series:



f (n)
(0 ) n f  (0 ) f  (0 ) 2 f  (0 ) 3
f (x)  
n0 n!
x  f (0 ) 
1!
x
2!
x 
3!
x  ...

83
Complex Taylor Series
Similarly, for a complex function f(z), analytic inside
and on a simple closed curve C, the Taylor series of
the function f at a point z0 inside C is

f (z)  n0
an ( z  z0 )n where

The series representation is valid for |z – z0| < R


where R and C are y
related as shown: C
R
z0
O x
Again the case z0 = 0 is called the Maclaurin Series:

f (n)
(0 ) n f  (0 ) f  (0 ) 2 f (0 ) 3
f (z)  
n0 n!
z  f (0 ) 
1!
z
2!
z 
3!
z  ...
84
Example 35 Find the Maclaurin series for f ( z )  e z
and its radius of convergence.

85
Example 36
Find the Taylor series for f (z)  sinh z about 
z0  j
2

86
The results of Example 35 mirrors the result for real
series. Similarly, the following can be shown:
 n 2 3
e    1  z    ...
z z z z
n 0
n! 2! 3!
 2 n 1 3 5 7
sin z   (1)
z z z z
n
 z     ...
n 0
(2n  1)! 3! 5! 7!
 2n 2 4 6
cos z   (1)
z z z z
n
 1     ...
n 0
(2n)! 2! 4! 6!
all valid everywhere in the complex plane.

87
1.13 Laurent Series
Any Taylor series is only valid within a circle, centre z0,
up to the nearest singularity. No Taylor expansion is
possible about a singularity. Near singularities we can
instead use the Laurent Series.

Suppose a function f(z) is


analytic on and between two
concentric circles C1 and C2,
circles centred on z0 with radii
r1 and r2, r1 > r2 .
This is called an annular region.
(Note that f(z) is not
necessarily analytic at z0.)
88
In this case, at every point z within the annulus, f(z) may
be represented by the Laurent series:

f (z)  
n  
an ( z  z0 )n
a n a  n 1 a1
 ...  n  n 1
 ...   a0
( z  z0 ) ( z  z0 ) ( z  z0 )
 a1 ( z  z 0 )  ...  a n ( z  z 0 ) n  ...
1 

 an ( z  z0 )n   an ( z  z0 )n
We can write
f (z) 
n   n0
Principal part of Taylor Series
the Laurent Series

• If f(z) is analytic at z0 we find all the coefficients in the


principal part are zero so the series is a Taylor series. 89
Example 37
1
Find the Laurent series expansion of f ( z )  2
about a) z = 0, b) z = –1. z ( z  1)
0 0
State the region of validity of each series.

90
Example 38
1
Find the Laurent series expansion of f ( z ) 
( z  1)( z  3)
valid for 1  z  3.

91

You might also like