0% found this document useful (0 votes)
14 views59 pages

Complex All

1) A complex number is of the form x + iy, where x is the real part and y is the imaginary part. 2) The limit of a complex function f(z) as z approaches z0 is defined as the number l such that the absolute value of f(z) - l can be made arbitrarily small by taking z sufficiently close to z0. 3) If the limit of a function f(z) as z approaches z0 exists, then the limit value is unique.

Uploaded by

mithun.swe
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
14 views59 pages

Complex All

1) A complex number is of the form x + iy, where x is the real part and y is the imaginary part. 2) The limit of a complex function f(z) as z approaches z0 is defined as the number l such that the absolute value of f(z) - l can be made arbitrarily small by taking z sufficiently close to z0. 3) If the limit of a function f(z) as z approaches z0 exists, then the limit value is unique.

Uploaded by

mithun.swe
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 59

Complex Variables

1
Limit, Continuity, Differentiation

Definition of Complex Number: Let 𝑥 and 𝑦 be two real numbers. Then a number of the form 𝑥 + 𝑖𝑦 = 𝑧
is called a complex number where 𝑖 = √−1.
And 𝑥 is called the real part of the complex number 𝑧 and 𝑦 is called the imaginary part of 𝑧 and these are
denoted by 𝑅𝑒 𝑧 and 𝐼𝑚 𝑧 respectively.
Note: The complex number 𝑥 + 𝑖𝑦 is also represented by the ordered pair of real numbers (𝑥, 𝑦), i.e.,
𝑥 + 𝑖𝑦 = (𝑥, 𝑦).

# Let 𝑧 = 𝑥 + 𝑖𝑦 and 𝑧 = 𝑥 + 𝑖𝑦 then 𝑧 ≤ 𝑧 if and only if 𝑥 ≤ 𝑥 and 𝑦 ≤ 𝑦 .


𝑧 ± 𝑧 = (𝑥 ± 𝑥 ) + 𝑖(𝑦 ± 𝑦 )
𝑧 𝑧 = (𝑥 𝑥 − 𝑦 𝑦 ) + 𝑖(𝑥 𝑥 + 𝑦 𝑦 )

Definition of Conjugate: Let the complex number 𝑧 = 𝑥 + 𝑖𝑦. Then the conjugate of 𝑧 is 𝑥 − 𝑖𝑦 and is
denoted by 𝑧̅ i.e., if 𝑧 = 𝑥 + 𝑖𝑦 then 𝑧̅ = 𝑥 – 𝑖𝑦.
# A complex number 𝑧 = 𝑥 + 𝑖𝑦 represents a point in the
complex plane i.e., 𝑅 plane (space) as shown in the figure 1.
# If 𝑧 = 𝑥 + 𝑖𝑦 is a complex number denoted by the point P
then r, the length of OP where O is the origin is called the
modulas of z and is denoted by | 𝑧 |.

i.e., 𝑟 = | 𝑧 | = 𝑥 +𝑦 .
And if 𝜃 is the angle between the line OP and the real axis then 𝜃
is called argument or amplitude of 𝑧 and is denoted by arg 𝑧 or Figure 1
amp 𝑧.

i.e., 𝜃 = arg 𝑧 = 𝑡𝑎𝑛

# From the figure


𝑥 = 𝑟 𝑐𝑜𝑠𝜃
𝑦 = 𝑟 𝑠𝑖𝑛 𝜃
Hence 𝑧 = 𝑟 𝑐𝑜𝑠 𝜃 + 𝑖𝑟 𝑠𝑖𝑛 𝜃
= 𝑟 (𝑐𝑜𝑠 𝜃 + 𝑖𝑠𝑖𝑛 𝜃)
= 𝑟𝑒 [polar form of 𝑧]
Similarly 𝑧̅ = r𝑒
Therefore any 𝑧̅ = − arg 𝑧
2
Limit, Continuity, Differentiation

And | 𝑧 | = |𝑧̅|

# If 𝑧 = 𝑟 𝑒 and 𝑧 = 𝑟 𝑒
Then
( )
𝑧 𝑧 =𝑟𝑟 𝑒
|𝑧 𝑧 |=|𝑟 𝑟 | = |𝑧 ||𝑧 |
And, arg (𝑧 𝑧 ) = 𝜃 + 𝜃 = arg 𝑧 + arg 𝑧
Similarly,
| |
=| |
provided |𝑧 | ≠ 0

arg = arg 𝑧 − arg 𝑧

# |𝑧| = 𝑥 + 𝑦 = (𝑥 + 𝑖𝑦)(𝑥 − 𝑖𝑦) = 𝑧𝑧̅

𝑧 ± 𝑧 = (𝑥 ± 𝑥 ) + 𝚤(𝑦 ± 𝑦 ) = (𝑥 ± 𝑥 ) − 𝑖(𝑦 ± 𝑦 )
= (𝑥 − 𝑖𝑦 ) ± (𝑥 − 𝑖𝑦 ) = 𝑧 ± 𝑧
And, 𝑧 𝑧 = (𝑥 + 𝚤𝑦 )(𝑥 + 𝚤𝑦 ) = (𝑥 𝑥 − 𝑦 𝑦 ) + 𝚤(𝑥 𝑦 + 𝑥 𝑦 )
= (𝑥 𝑥 − 𝑦 𝑦 ) − 𝑖(𝑥 𝑦 + 𝑥 𝑦 )
= (𝑥 − 𝑖𝑦 )(𝑥 − 𝑖𝑦 ) = 𝑧 𝑧
And 𝑧̅̅ = 𝑧

Definition of Complex Variable: A variable 𝑧 which takes any value from a set of complex numbers is
called complex variable.
Problem: For any two complex number 𝑧 and 𝑧 prove that
(i) |𝑧 + 𝑧 | ≤ |𝑧 | + |𝑧 |
(ii) |𝑧 − 𝑧 | ≥ |𝑧 | − |𝑧 |
Proof:
(i) We have,
|𝑧 + 𝑧 | = (𝑧 + 𝑧 )(𝑧 + 𝑧 ) ∵ 𝑧 𝑧 = |𝑧 |
= (𝑧 + 𝑧 )(𝑧 + 𝑧 )
= 𝑧 𝑧 +𝑧 𝑧 +𝑧 𝑧 +𝑧 𝑧
= |𝑧 | + |𝑧 | + 𝑧 𝑧 + 𝑧 𝑧 ∵ 𝑧 + 𝑧̅ = 2𝑥 = 2𝑅𝑒(𝑧)
= |𝑧 | + |𝑧 | + 2 𝑅𝑒(𝑧 𝑧 )z=x+iy
≤ |𝑧 | + |𝑧 | + 2|𝑧 𝑧 | Re z =x≤√(x2+y2)=|z|
3
Limit, Continuity, Differentiation

= |𝑧 | + |𝑧 | + 2|𝑧 ||𝑧 |
= {|𝑧 | + |𝑧 |}
So, |𝑧 + 𝑧 | ≤ |𝑧 | + |𝑧 |
(ii) |𝑧 | = |(𝑧 − 𝑧 ) + 𝑧 | ≤ |𝑧 − 𝑧 | + |𝑧 |
Hence, |𝑧 − 𝑧 | ≥ |𝑧 | − |𝑧 |

Example: Show that |𝒙| + |𝒚| ≤ √2|𝒙 + 𝒊𝒚|


Solution: We know,
2|𝑥 + 𝑖𝑦| = 2(𝑥 + 𝑦 ) = 𝑥 + 𝑦 + 𝑥 + 𝑦 = |𝑥| + |𝑦| + |𝑥| + |𝑦|
≥ |𝑥| + |𝑦| + 2|𝑥||𝑦| ∵ 𝑥 + 𝑦 ≥ 2𝑥𝑦
≥ (|𝑥| + |𝑦|) i.e., AM≥GM

Hence, |𝑥| + |𝑦| ≤ √2|𝑥 + 𝑖𝑦|

LIMIT
Definition of Complex Functions: Let 𝐴 and 𝐵 be two sets. A function 𝑓: 𝐴 → 𝐵 is a correspondence such
that by 𝑓 each elements of 𝐴 corresponds to a unique element of 𝐵.
Here we shall always consider the function 𝑓: 𝐶 → 𝐶, where 𝐶 is the set of complex numbers.

Definition of Limit: Let 𝑓(𝑧) be a function. A number 𝑙 is called the limit of 𝑓(𝑧) at 𝑧 = 𝑧 if for any 𝜀 > 0
There exists a 𝛿 > 0 (depending on 𝜀) such that |𝑓(𝑧) − 𝑙| < 𝜀 where 0 < |𝑧 − 𝑧 | < 𝛿.
i.e., 𝑓(𝑧) → 𝑙 as 𝑧 → 𝑧
𝑙𝑖𝑚 𝑓(𝑧) = 𝑙

Example: prove that 𝑙𝑖𝑚 𝑧 = 𝑧


Proof: Here 𝑓(𝑧) = 𝑧


Choose any 𝜀 > 0
lim f(z) = z is true if there exists a 𝛿 > 0 such that |𝑓(𝑧) − 𝑧 | < 𝜀 whenever 0 < |𝑧 − 𝑧 | < 𝛿

Consider 𝛿 ≤ 1 ∴ |𝑧 − 𝑧 | < 𝛿 < 1


Then |𝑓(𝑧) − 𝑧 | = |𝑧 − 𝑧 | = |(𝑧 + 𝑧 )(𝑧 − 𝑧 )|
= |(𝑧 − 𝑧 + 2𝑧 )(𝑧 − 𝑧 )|
< 𝛿|(𝑧 − 𝑧 ) + 2𝑧 | ∵ |𝑧 − 𝑧 | < 𝛿
≤ 𝛿(|𝑧 − 𝑧 | + 2|𝑧 |)
4
Limit, Continuity, Differentiation

< 𝛿(1 + 2|𝑧 |) ∵ |𝑧 − 𝑧 | < 𝛿 < 1

Suppose 𝛿 = min 1, | |

Then |𝑓(𝑧) − 𝑧 | < | |


(1 + 2|𝑧 |) = 𝜀 whenever 0 < |𝑧 − 𝑧 | < 𝛿

Hence, lim → 𝑓 (𝑧 ) = 𝑧
lim → 𝑧 =𝑧 [Proved]
Similarly we can show that lim 𝑧 = 𝑧

Theorem: If 𝐥𝐢𝐦𝒛→𝒛𝟎 𝒇(𝒛) exists then its value is unique.

Proof: Suppose lim → 𝑓(𝑧) = 𝑙 and lim → 𝑓 (𝑧 ) = 𝑙

We have to show that 𝑙 = 𝑙 .

Choose 𝜀 > 0. Then there exists 𝛿 > 0, 𝛿 > 0 such that|𝑓(𝑧) − 𝑙 | < , whenever 0 < |𝑧 − 𝑧 | < 𝛿

and|𝑓 (𝑧) − 𝑙 | < , whenever 0 < |𝑧 − 𝑧 | < 𝛿

Let 𝛿 = min(𝛿 , 𝛿 )
Hence, for 0 < |𝑧 − 𝑧 | < 𝛿
|𝑙 − 𝑙 | = |𝑙 − 𝑓(𝑧) + 𝑓 (𝑧) − 𝑙 |
≤ |𝑙 − 𝑓 (𝑧)| + |𝑓(𝑧) − 𝑙 |
= |𝑓(𝑧) − 𝑙 | + |𝑓(𝑧) − 𝑙 |

< + =𝜀

Or, |𝑙 − 𝑙 | < 𝜀
Since 𝜀 is arbitrary,
|𝑙 − 𝑙 | = 0 𝑜𝑟 𝑙 −𝑙 =0 𝑜𝑟 𝑙 =𝑙

Properties: Suppose 𝐥𝐢𝐦𝒛→𝒛𝟎 𝒇(𝒛) = 𝑨 and 𝐥𝐢𝐦𝒛→𝒛𝟎 𝒈(𝒛) = 𝑩. Then:

(i) 𝐥𝐢𝐦𝒛→𝒛𝟎 {𝒇(𝒛) ± 𝒈(𝒛)} = 𝑨 ± 𝑩


(ii) 𝐥𝐢𝐦𝒛→𝒛𝟎 𝒇(𝒛). 𝒈(𝒛) = 𝑨. 𝑩
𝟏 𝟏
(iii) 𝐥𝐢𝐦𝒛→𝒛𝟎 = provided 𝑩 ≠ 𝟎
𝒈(𝒛) 𝑩
𝒇(𝒛) 𝑨
(iv) 𝐥𝐢𝐦𝒛→𝒛𝟎 = provided 𝑩 ≠ 𝟎
𝒈(𝒛) 𝑩

Proof:
(i) Choose 𝜀 > 0
5
Limit, Continuity, Differentiation

Since lim 𝑓(𝑧) = 𝐴, so there exist 𝛿 > 0such that |𝑓(𝑧) − 𝐴| < whenever 0 < |𝑧 − 𝑧 | < 𝛿

Again since lim 𝑔(𝑧) = 𝐵, so there exist 𝛿 > 0 such that |𝑔(𝑧) − 𝐵| < whenever 0 < |𝑧 − 𝑧 | < 𝛿

Let, 𝛿 = min (𝛿 , 𝛿 )
So, for 0 < |𝑧 − 𝑧 | < 𝛿
|𝑓(𝑧) + 𝑔(𝑧) − (𝐴 + 𝐵 )| = |{𝑓(𝑧) − 𝐴} + {𝑔(𝑧) − 𝐵}|
≤ |𝑓(𝑧) − 𝐴| + |𝑔(𝑧) − 𝐵|

< + =𝜀

Hence, lim → {𝑓(𝑧) + 𝑔(𝑧)} = 𝐴 + 𝐵

Similarly, lim → {𝑓(𝑧) − 𝑔(𝑧)} = 𝐴 − 𝐵

(ii) We get,
|𝑓(𝑧)𝑔(𝑧) − 𝐴𝐵| = |𝑓(𝑧)𝑔(𝑧) − 𝑓(𝑧)𝐵 + 𝑓(𝑧)𝐵 − 𝐴𝐵|
= |𝑓 (𝑧){𝑔(𝑧) − 𝐵} + 𝐵{𝑓 (𝑧) − 𝐴}|
≤ |𝑓(𝑧)||𝑔(𝑧) − 𝐵| + |𝐵||𝑓(𝑧) − 𝐴|
Since, lim → 𝑓(𝑧) = 𝐴, So there exists 𝛿 > 0 such that |𝑓(𝑧) − 𝐴| < 1 whenever 0 < |𝑧 − 𝑧 | < 𝛿
But |𝑓(𝑧)| − |𝐴| ≤ |𝑓(𝑧) − 𝐴|
∴ |𝑓(𝑧)| − |𝐴| < 1
Or |𝑓(𝑧)| < 1 + |𝐴|

Since lim 𝑔(𝑧) = 𝐵, so there exists 𝛿 > 0 such that |𝑔(𝑧) − 𝐵| < ( | |)
whenever 0 < |𝑧 − 𝑧 | < 𝛿 .

Since lim 𝑓(𝑧) = 𝐴, so there exists 𝛿 > 0 such that |𝑓(𝑧) − 𝐴| < (| |)
whenever 0 < |𝑧 − 𝑧 | < 𝛿.

Let 𝛿 = min (𝛿 , 𝛿 , 𝛿 )
So for 0 < |𝑧 − 𝑧 | < 𝛿
𝜀 𝜀
|𝑓(𝑧)𝑔(𝑧) − 𝐴𝐵| < (1 + |𝐴|) + |𝐵|
2(1 + |𝐴|) 2|𝐵|
= + =𝜀
|𝑓(𝑧)𝑔(𝑧) − 𝐴𝐵| < 𝜀 whenever 0 < |𝑧 − 𝑧 | < 𝛿
∴ lim 𝑓(𝑧)𝑔(𝑧) = 𝐴𝐵

(iii) We get,
1 1 |𝑔(𝑧) − 𝐵|
− =
𝑔 (𝑧 ) 𝐵 |𝐵||𝑔(𝑧)|
| |
Since lim 𝑔(𝑧) = 𝐵 ≠ 0, so there exists 𝛿 > 0 such that |𝑔(𝑧) − 𝐵| < for 0 < |𝑧 − 𝑧 | < 𝛿

| |
Again since lim 𝑔(𝑧) = 𝐵 so there exists 𝛿 > 0 such that |𝑔(𝑧) − 𝐵| < whenever 0 < |𝑧 − 𝑧 | <

𝛿 .
6
Limit, Continuity, Differentiation

But 𝐵 = 𝐵 − 𝑔(𝑧) + 𝑔(𝑧)


∴ |𝐵| ≤ |𝑔(𝑧) − 𝐵| + |𝑔(𝑧)|
| |
< + |𝑔(𝑧)|
1
𝑜𝑟 |𝑔(𝑧)| > 𝐵
2
1 2
𝑜𝑟 <
|𝑔(𝑧)| |𝐵|
Let 𝛿 = min (𝛿 , 𝛿 )
So for 0 < |𝑧 − 𝑧 | < 𝛿 we get,
1 1 |𝑔(𝑧) − 𝐵| 1
− = .
𝑔 (𝑧 ) 𝐵 |𝐵| |𝑔(𝑧)|
| |
≤ | |
.2|𝐵| = 𝜀
Or ( )
− < 𝜀 whenever 0 < |𝑧 − 𝑧 | < 𝛿
Therefore lim ( )
=

( )
(iv) lim ( )
= lim 𝑓(𝑧). ( )
= lim 𝑓 (𝑧). lim ( )
→ → → →
= 𝐴. [By (ii) and (iii)]
=
[Proved]

𝒛
Example: Show that 𝐥𝐢𝐦 does not exist
𝒛→𝒛𝟎 𝒛

Solution: Let 𝑧 = 𝑥 + 𝑖𝑦
̅
Thenlim = lim
→ ( , )→( , )

Suppose z→0 along x axis (y=0, x→0)


̅
Then lim = lim = lim =1
→ → , →

Now let z→0 along y axis (x=0, y→0)


̅
Then lim = lim = lim = −1
→ , → →

̅
Since the two limits are unequal, so lim does not exist.

7
Limit, Continuity, Differentiation

CONTINUITY
Definition of Continuity: A function 𝑓(𝑧) is said to be continuous at 𝑧 if for any 𝜀 > 0 there exist
𝛿 > 0 (depending on 𝜀 and also on 𝑧 such that |𝑓(𝑧) − 𝑓(𝑧 )| < 𝜀 whenever |𝑧 − 𝑧 | < 𝛿 or if lim 𝑓(𝑧) =

𝑓(𝑧 ) .
Definition of Uniformly Continuous: A function 𝑓(𝑧) is said to be uniformly continuous in a region 𝑅 if
for any 𝜀 > 0 there exist 𝛿 > 0 (depending on 𝜀, not at any point in R) such that |𝑓(𝑧 ) − 𝑓(𝑧 )| < 𝜀
whenever |𝑧 −𝑧 | < 𝛿 in R where 𝑧 , 𝑧 ∈ 𝑅
Properties:
If f(z) and g(z) are continuous at 𝒛𝟎 then,
(i) 𝒇(𝒛) ± 𝒈(𝒛) is continuous at 𝒛𝟎
(ii) 𝒇(𝒛)𝒈(𝒛) is continuous at 𝒛𝟎
𝒇(𝒛)
(iii) is continuous at 𝒛𝟎 , provided 𝒈(𝒛) ≠ 𝟎
𝒈(𝒛)

Example: Show that 𝒇(𝒛) = 𝒛𝟐 is uniformly continuous on |z|<1.


Solution: We must show that for any 𝜀 > 0 there exists a 𝛿 > 0 (depending only on 𝜀) such that |𝑓(𝑧 ) −
𝑓 (𝑧 )| < 𝜀 whenever |𝑧 −𝑧 | < 𝛿.
Here 𝑧 and 𝑧 are any two point in |z|<1.
Now, |𝑓(𝑧 ) − 𝑓(𝑧 )| = |𝑧 −𝑧 |
= |𝑧 + 𝑧 ||𝑧 − 𝑧 |
≤ {|𝑧 | + |𝑧 |}|𝑧 − 𝑧 | [∵ |𝑧 | < 1, |𝑧 | < 1 ]
< 2|𝑧 − 𝑧 |
< 2𝛿 whenever |𝑧 − 𝑧 | < 𝛿

Choose 𝛿 = , we see that |𝑓(𝑧 ) − 𝑓(𝑧 )| < 𝜀 whenever |𝑧 − 𝑧 | < 𝛿

Hence 𝑓 (𝑧) = 𝑧 is uniformly continuous in |𝑧| < 1.

𝟏
Example: Show that 𝒇(𝒛) = is not uniformly continuous on the region |z|<1.
𝒛

Solution: If 𝑓(𝑧) is uniformly continuous in the region |𝑧| < 1 then for any 𝜀 > 0 there exist a δ>0
(depending on 𝜀) such that for any two points 𝑧 and 𝑧 in the region, |𝑓(𝑧 ) − 𝑓(𝑧 )| < 𝜀 whenever
|𝑧 − 𝑧 | < 𝛿.

Consider 𝛿 < 1. Choose two points 𝑧 = 𝛿 and 𝑧 = .


8
Limit, Continuity, Differentiation

Then |𝑧 − 𝑧 | = 𝛿 − = <𝛿

But, |𝑓(𝑧 ) − 𝑓(𝑧 )| = −

= −

= >𝜀 [∵ 𝛿 < 1]

This is contradiction. Hence 𝑓(𝑧) = is not uniformly continuous on the region |z|<1.

DIFFERENTIABILITY:
Definition of Differentiability: A function 𝑓(𝑧) is said to be differentiable at a point 𝑧 if
𝐿𝑖𝑚 ( ) ( ) 𝐿𝑖𝑚 ( ) ( )
exist and we write 𝑓 (𝑧 ) =
Δ𝑧 → 𝑧 Δ𝑧 → 𝑧
Here 𝑓′(𝑧 ) is called the derivative of 𝑓(𝑧) at 𝑧 .

If we put 𝑧 = 𝑧 + Δ𝑧

As 𝑧 → 𝑧 then Δ𝑧 → 0

𝐿𝑖𝑚 ( ) ( )
Then 𝑓′(𝑧 ) =
Δ𝑧 → 𝑧

Theorem: A function 𝒇(𝒛) which is differentiable at a point is also continuous there. But the converse
is not necessarily true.
Proof: Suppose 𝑓(𝑧) is differentiable at 𝑧 . Then,
𝐿𝑖𝑚 ( ) ( )
𝑓 (𝑧 ) = exist and a finite number.
Δ𝑧 → 𝑧
( ) ( )
Now, 𝑓(𝑧) − 𝑓(𝑧 ) = × (𝑧 − 𝑧 )

Taking limit on both sides as 𝑧 → 𝑧


𝑓(𝑧) − 𝑓(𝑧 )
lim {𝑓(𝑧) − 𝑓(𝑧 )} = lim × (𝑧 − 𝑧 )
→ → 𝑧−𝑧
𝐿𝑖𝑚 ( ) ( ) 𝐿𝑖𝑚
= × (𝑧 − 𝑧 )
𝑧→𝑧 𝑧→𝑧
= 𝑓 (𝑧 ) × 0 = 0
9
Limit, Continuity, Differentiation
𝐿𝑖𝑚
Hence, 𝑓(𝑧) = 𝑓(𝑧 )
𝑧→𝑧
Thus, 𝑓(𝑧) is continuous at 𝑧 .
To prove the converse, consider 𝑓(𝑧) = |𝑧|.
Here 𝑓(𝑧) is clearly continuous at 𝑧 = 0.
𝐿𝑖𝑚 ( ) ( ) 𝐿𝑖𝑚 ( ) ( )
But, =
Δ𝑧 → 0 Δ𝑧 → 0
𝐿𝑖𝑚 | | | |
=
Δ𝑧 → 0
𝐿𝑖𝑚
( ) ( )
= Δ𝑥 → 0
Δ𝑦 → 0
Suppose Δ𝑧 → 0 along Δ𝑥 = 0, Δ𝑦 → 0
Then,
𝐿𝑖𝑚
( ) ( ) 𝐿𝑖𝑚
limit = Δ𝑥 = 0 = = = −𝑖
Δ𝑦 → 0
Δ𝑦 → 0
Suppose Δ𝑧 → 0 along Δ𝑥 → 0, Δ𝑦 = 0
Then,
𝐿𝑖𝑚
( ) ( ) 𝐿𝑖𝑚
limit = Δ𝑥 → 0 = =1
Δ𝑦 = 0 Δ𝑥 → 0

𝐿𝑖𝑚 ( ) ( )
Since the two limits are unequal, so does not exist or 𝑓 (0) does not exist.
Δ𝑧 → 0

ANALYTIC FUNCTION
Definition of analytic function: Let 𝑓(𝑧) be a function define on region 𝑅. Then 𝑓(𝑧) is said to be analytic
in 𝑅 if it is differentiable at every point of 𝑅.
Definition: A function 𝑓(𝑧) is called analytic at a point 𝑧 if there exists a neighbourhood of 𝑧 such that
𝑓 (𝑧) exists at each point of the neighbourhood.

Example: Show that 𝒇(𝒛) = |𝒛|𝟐 is differentiable at z=0 but not analytic there.
Proof: Here 𝑓 (𝑧) = |𝑧|
( ) ( ) | |
Then, lim → = lim →

| |
= lim →
10
Limit, Continuity, Differentiation
̅
= lim →

= lim → 𝑧̅ = 0
Hence f(z) is differentiable at z=0
Choose any point 𝑧 ≠ 0, and consider the limit:
𝑓(𝑧 + Δ𝑧) − 𝑓 (𝑧 )
lim
→ Δ𝑧
| | | |
= lim

( ) ( )
= lim where, Δ𝑧 = Δ𝑥 + 𝑖Δ𝑦

𝑧 = 𝑥 + 𝑖𝑦
( ) ( )
= lim
→ , →

Suppose Δ𝑧 → 0 along Δ𝑥 = 0, Δ𝑦 → 0, then


( ) ( )
limit = lim
, →

( )
= lim

= lim = = −2𝑖𝑦

Again, let Δ𝑧 → 0 along Δ𝑥 → 0, Δ𝑦 = 0, then


( ) ( )
limit = lim
→ ,

( )
= lim

= lim (2𝑥 + Δ𝑥) = 2𝑥



( ) ( )
Since the two limits are unequal, so lim does not exist.

Therefore f(z) is not differentiable anywhere other than zero or origin.


Hence f(z) is not analytic at z=0.

Cauchy-Riemann equations (C-R equations):

If 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦) defined in a region R, then = and =− are called C-R equations.
11
Limit, Continuity, Differentiation

Necessary conditions for an analytic function:


Necessary condition for a function 𝒇(𝒛) = 𝒖(𝒙, 𝒚) + 𝒊𝒗(𝒙, 𝒚) to be analytic in a region R is that it
𝝏𝒖 𝝏𝒗 𝝏𝒖 𝝏𝒗
satisfies the C-R equations = and =− at each point of R.
𝝏𝒙 𝝏𝒚 𝝏𝒚 𝝏𝒙

Proof: Since f(z) is analytic in R, so 𝑓 (𝑧) exists for each point z in R.


( ) ( )
Hence lim →

[{ ( , ) ( , )} { ( , ) ( , )}]
= lim → , →

Must exist independent of manner in which Δ𝑧 approaches zero (or Δ𝑥, Δ𝑦 approaches zero).
Consider the path Δ𝑥 = 0, Δ𝑦 → 0, then
( , ) ( , ) { ( , ) ( , )}
limit = lim →

( , ) ( , ) ( , ) ( , )
= lim → + lim

= +

= -i +

= –i

Consider the path Δ𝑦 = 0, Δ𝑥 → 0, then


( , ) ( , ) ( , ) ( , )
limit = lim → + 𝑖 lim →

𝜕𝑢 𝜕𝑣
= +𝑖
𝜕𝑥 𝜕𝑥
Since 𝑓′(𝑧) exists, above two limits must be equal. Hence

−𝑖 = +𝑖

Equating the real and imaginary part from both sides we get:

= and =−

So C-R equations are satisfied.

Sufficient condition for an analytic function:


𝝏𝒖 𝝏𝒗 𝝏𝒖 𝝏𝒗
If w=f(z)=u(x,y)+i v(x,y) satisfy the C-R equations = and =− in a region R and partial
𝝏𝒙 𝝏𝒚 𝝏𝒚 𝝏𝒙
𝝏𝒖 𝝏𝒖 𝝏𝒗 𝝏𝒗
derivatives , , , are continuous in R, then f(z) is analytic in R.
𝝏𝒙 𝝏𝒚 𝝏𝒙 𝝏𝒚
12
Limit, Continuity, Differentiation

Proof: Since and are supposed continuous, so

Δ𝑢 = 𝑢(𝑥 + Δ𝑥, 𝑦 + Δ𝑦) − 𝑢(𝑥, 𝑦)


= {𝑢(𝑥 + Δ𝑥, 𝑦 + Δ𝑦) − 𝑢(𝑥, 𝑦 + Δ𝑦)} + {𝑢(𝑥, 𝑦 + Δ𝑦) − 𝑢 (𝑥, 𝑦)}
𝜕𝑢 𝜕𝑢
= +𝜀 Δ𝑥 + +𝜂 Δ𝑦
𝜕𝑥 𝜕𝑦
𝜕𝑢 𝜕
= Δ𝑥 + Δ𝑦 + 𝜀 Δ𝑥 + 𝜂 Δ𝑦
𝜕𝑥 𝜕𝑦
Where 𝜀 → 0 𝑎𝑠 Δ𝑥 → 0 and 𝜂 → 0 𝑎𝑠 Δ𝑦 → 0

Similarly, since and are supposed continuous, so

𝜕𝑣 𝜕𝑣
Δ𝑣 = +𝜀 Δ𝑥 + +𝜂 Δ𝑦
𝜕𝑥 𝜕𝑦
𝜕𝑣 𝜕𝑣
= Δ𝑥 + Δ𝑦 + 𝜀 Δ𝑥 + 𝜂 Δ𝑦
𝜕𝑥 𝜕𝑦
Where 𝜀 → 0 𝑎𝑠 Δ𝑥 → 0 𝑎𝑛𝑑 𝜂 → 0 𝑎𝑠 Δ𝑦 → 0
Hence, Δ𝑤 = Δ𝑢 + 𝑖Δ𝑣

= +𝑖 Δ𝑥 + +𝑖 Δ𝑦 + (𝜀 + 𝑖𝜀 )Δ𝑥 + (𝜂 + 𝑖𝜂 )Δ𝑦

= +𝑖 Δ𝑥 + +𝑖 Δ𝑦 + 𝜀Δ𝑥 + 𝜂Δ𝑦

Where 𝜀 = 𝜀 + 𝑖𝜀 → 0 𝑎𝑠 Δ𝑥 → 0
And 𝜂 = 𝜂 + 𝑖𝜂 → 0 𝑎𝑠 Δ𝑦 → 0

= +𝑖 Δ𝑥 + − +𝑖 Δ𝑦 + 𝜀Δ𝑥 + 𝜂Δ𝑦

= +𝑖 Δ𝑥 + 𝑖 +𝑖 Δ𝑦 + 𝜀Δ𝑥 + 𝜂Δ𝑦

𝜕𝑢 𝜕𝑣
∴ Δ𝑤 = +𝑖 (Δ𝑥 + 𝑖Δ𝑦) + 𝜀Δ𝑥 + 𝜂Δ𝑦
𝜕𝑥 𝜕𝑥

i.e., Δ𝑤 = +𝑖 Δ𝑧 + 𝜀Δ𝑥 + 𝜂Δ𝑦 where Δ𝑧 = Δ𝑥 + 𝑖Δ𝑦

Δ𝑤 𝜕𝑢 𝜕𝑣 𝜀Δ𝑥 + 𝜂Δ𝑦
∴ = +𝑖 +
Δ𝑧 𝜕𝑥 𝜕𝑥 Δ𝑧

Notice that, = |𝜀| ≤ |𝜀| as |Δ𝑥| ≤ |Δ𝑧|, which tends to zero as Δ𝑧 → 0

i.e., lim → =0

similarly, lim → =0
13
Limit, Continuity, Differentiation

So, lim → =0

Hence, lim = +𝑖

Therefore, = 𝑓′(𝑧) exist at each point of the region R, and hence f(z) is analytic in R.

Note: 𝑓 (𝑧) = +𝑖 = −𝑖

Note: If 𝑓(𝑧) = 𝑢 + 𝑖𝑣 is analytic then

= =−

= =−

Therefore for analytic function = and + = 0 and the equation + = 0 is called


Laplace’s equation.

Definition of Harmonic Function: A function which satisfy Laplace’s equation is called harmonic
function.
Note: So u is called harmonic function, similarly v is called harmonic function. u is called harmonic
conjugate of v and v is called harmonic conjugate of u.
𝒙𝟑 (𝟏 𝒊) 𝒚𝟑 (𝟏 𝒊)
, 𝒂𝒕 𝒛 ≠ 𝟎
Example: Show that𝒇(𝒛) = 𝒙𝟐 𝒚𝟐 is continuous everywhere. Also show that the C-R
𝟎 𝒂𝒕 𝒛 ≠ 𝟎
equations are satisfied at origin, but is not analytic there.
Proof:
Part1:
Let 𝑓(𝑧) = 𝑢 + 𝑖𝑣
Now for 𝑧 ≠ 0
𝒙𝟑 (𝟏 + 𝒊) − 𝒚𝟑 (𝟏 − 𝒊)
𝑓 (𝑧 ) =
𝒙𝟐 + 𝒚𝟐
𝑥 −𝑦 𝑥 +𝑦
= +𝑖
𝑥 +𝑦 𝑥 +𝑦

∴ 𝑢(𝑥, 𝑦) = where (𝑥, 𝑦) ≠ (0,0)

=0 at (0,0)

∴ 𝑣 (𝑥, 𝑦) = where (𝑥, 𝑦) ≠ (0,0)

=0 at (0,0)
14
Limit, Continuity, Differentiation

It is clear that 𝑥 − 𝑦 , 𝑥 + 𝑦 , 𝑥 + 𝑦 are continuous everywhere.


So for 𝑧 ≠ 0 both 𝑢 and 𝑣 are continuous.
Hence 𝑓(𝑧) is continuous for all 𝑧 ≠ 0.
Now for 𝑧 = 0, choose 𝜀 > 0, then
𝑥 −𝑦 |𝑥 − 𝑦||𝑥 + 𝑥𝑦 + 𝑦 | 𝑥 +𝑦
|𝑢(𝑥, 𝑦) − 0| = = ∵ ≥ 𝑥𝑦
𝑥 +𝑦 |𝑥 + 𝑦 | 2

| |
⇒ 𝑥 + 𝑦 ≥ 𝑥𝑦
≤ | |
∵ |𝑥| + |𝑦| ≤ √2|𝑥 + 𝑖𝑦|
= 2|𝑥 − 𝑦|
= √2 𝑥 + 𝑦
≤ 2(|𝑥| + |𝑦|) ≤ 4 𝑥 + 𝑦

Consider, 𝛿 = , then =2 𝑥 +𝑦

|𝑢(𝑥, 𝑦) − 0| < 4. whenever 𝑥 +𝑦 <𝛿

i.e., |𝑢(𝑥, 𝑦) − 0| < 𝜀 whenever |(𝑥, 𝑦) → (0,0)| < 𝛿


Hence, lim( , )→( , ) 𝑢 (𝑥, 𝑦) =0

i.e., 𝑢(𝑥, 𝑦) is continuous at (0,0)


Similarly 𝑣(𝑥, 𝑦) is continuous at (0, 0)
Hence 𝑓(𝑧) is continuous at 𝑧 = 0
There 𝑓(𝑧) is continuous everywhere.
Part 2:
Now we get,
( , ) ( , ) ( )
= lim → = lim → ( )
=1
,

( , ) ( , ) ( )
= lim → = lim → ( )
= −1
,

( , ) ( , ) ( )
= lim → = lim → ( )
=1
,

( , ) ( , ) ( )
= lim → = lim → ( )
= −1
,

Thus = and =−

Therefore C-R equations are satisfied at (0, 0)


15
Limit, Continuity, Differentiation
( ) ( )
Consider lim →

( ) ( )
= lim → , → ( )( )

Let 𝑧 → 0 along x=0, 𝑦 → 0. Then


−𝑦 (1 − 𝑖 ) 𝑖−1 𝑖−1
𝑙𝑖𝑚𝑖𝑡 = lim = lim = =1+𝑖
→ 𝑦 (𝑖𝑦) → 𝑖 𝑖
Again let 𝑧 → 0 along y=x. Then
𝑥 (1 + 𝑖 ) − 𝑥 (1 − 𝑖 ) 1+𝑖−1+𝑖 𝑖
𝑙𝑖𝑚𝑖𝑡 = lim = lim =
→ 2𝑥 (𝑥 + 𝑖𝑥) → 2(1 + 𝑖) 1+𝑖
Since the two limits are unequal so 𝑓 (0) does not exist.
So 𝑓(𝑧) is not analytic at 𝑧 = 0.
𝒙𝟐 𝒚𝟔 (𝒙 𝒊𝒚)
𝒘𝒉𝒆𝒏 𝒛 ≠ 𝟎
Example: Show that 𝒇(𝒛) = 𝒙𝟒 𝒚𝟏𝟐 is not analytic at origin (0, 0), but C-R equations
𝟎 𝒘𝒉𝒆𝒏 𝒛 = 𝟎
are satisfied there.
( ) ( )
Solution:Consider lim →

( )
= lim → , → ( )( )

Consider the path y=mx, then


𝑚 𝑥 𝑚 𝑥
𝑙𝑖𝑚𝑖𝑡 = lim = lim =0
→ 𝑥 +𝑚 𝑥 → 1+𝑚 𝑥
Again consider the path 𝑥 = 𝑦 . Then
𝑦 1
𝑙𝑖𝑚𝑖𝑡 = lim =
→ 𝑦 +𝑦 2
Since the two limits are unequal so 𝑓 (0)does not exist and hence 𝑓(𝑧) is not analytic at origin (0, 0).
Let 𝑓(𝑧) = 𝑢 + 𝑖𝑣. Then

𝑤ℎ𝑒𝑛 (𝑥, 𝑦) ≠ 0 𝑤ℎ𝑒𝑛 (𝑥, 𝑦) ≠ 0


𝑢= 𝑣=
0 𝑎𝑡 (𝑥, 𝑦) = 0 0 𝑎𝑡 (𝑥, 𝑦) = 0
Now
𝜕𝑢 𝑢(𝑥, 0) − 𝑢 (0,0) 0−0
= lim = lim =0
𝜕𝑥 ( , ) → Δ𝑥 → Δ𝑥

𝜕𝑢 𝑢(0, 𝑦) − 𝑢(0,0) 0−0


= lim = lim =0
𝜕𝑦 ( , )
→ Δ𝑦 → Δ𝑦
16
Limit, Continuity, Differentiation

Similarly = =0
, ,

Hence, = and =−

i.e., C-R equations are satisfied at (0, 0)

Example: Show that 𝒇(𝒛) = 𝟐𝒙 + 𝒊𝒙𝒚𝟐 is not analytic at anywhere.


Solution: Let 𝑓 (𝑧) = 𝑢 + 𝑖𝑣 , then
𝑢 = 2𝑥, 𝑣 = 𝑥𝑦 . So,
𝜕𝑢 𝜕𝑢 𝜕𝑣 𝜕𝑣
= 2, = 0, =𝑦 , = 2𝑥𝑦
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦

Hence ≠− when 𝑦 ≠ 0

i.e., C-R equations are not satisfied for any 𝑧 ≠ 0.

More over at 𝑧 = 0, ≠

∴ 𝑓(𝑧) is not analytic anywhere.

Example: Let 𝒖 = 𝟑𝒙𝟐 𝒚 + 𝟐𝒙𝟐 − 𝒚𝟑 − 𝟐𝒚𝟐. Show that u is harmonic and find its harmonic conjugate
v. Then find𝒇(𝒛) = 𝒖 + 𝒊𝒗 in terms of z.
Solution: Here 𝑢 = 3𝑥 𝑦 + 2𝑥 − 𝑦 − 2𝑦
𝜕𝑢
= 6𝑥𝑦 + 4𝑥
𝜕𝑥
𝜕𝑢
= 3𝑥 − 3𝑦 − 4𝑦
𝜕𝑦
𝜕 𝑢
= 6𝑦 + 4
𝜕𝑥
𝜕 𝑢
= −6𝑦 − 4
𝜕𝑦

Hence + =0

i.e., 𝑢 is a harmonic function.

Now by C-R equations = and =−

𝜕𝑣
∴ = 6𝑥𝑦 + 4𝑥
𝜕𝑦
17
Limit, Continuity, Differentiation

𝑣 = 3𝑥𝑦 + 4𝑥𝑦 + 𝐹(𝑥)


𝜕𝑣
= 3𝑦 + 4𝑦 + 𝐹 (𝑥)
𝜕𝑥
⇒ −3𝑥 + 3𝑦 + 4𝑦 = 3𝑦 + 4𝑦 + 𝐹 (𝑥)
⇒ 𝐹 (𝑥) = −3𝑥
⇒ 𝐹 (𝑥) = −𝑥 + 𝐶
∴ 𝑣 = 3𝑥𝑦 + 4𝑥𝑦 − 𝑥 + 𝐶, the harmonic conjugate of u.
Now, 𝑓 (𝑧) = 𝑢 + 𝑖𝑣
= 3𝑥𝑦 + 2𝑥 − 𝑦 − 2𝑦 + 𝑖[3𝑥𝑦 + 4𝑥𝑦 − 𝑥 + 𝐶]
= 2{𝑥 + 2𝑥𝑖𝑦 + (𝑖𝑦) } − 𝑖{𝑥 + 3𝑥(𝑖𝑦) + 3𝑥 𝑖𝑦 + (𝑖𝑦) }
= 2(𝑥 + 𝑖𝑦) − 𝑖(𝑥 + 𝑖𝑦) + 𝑖𝐶
= 2𝑧 − 𝑖𝑧 + 𝑖𝐶

Example: Show that 𝒖 = 𝒙𝟑 − 𝟑𝒙𝒚𝟐 + 𝟑𝒙𝟐 − 𝟑𝒚𝟐 + 𝟏 is harmonic. Find the analytic f(z)=u+ivin
terms of z.
Solution: Here 𝑢 = 𝑥 − 3𝑥𝑦 + 3𝑥 − 3𝑦 + 1

= 3𝑥 − 3𝑦 + 6𝑥 = φ (𝑥, 𝑦) 𝑠𝑎𝑦

𝜕 𝑢
= 6𝑥 + 6
𝜕𝑥

𝜕𝑢
= −6𝑥𝑦 − 6𝑦 = 𝜑 (𝑥, 𝑦) 𝑠𝑎𝑦
𝜕𝑦
𝜕 𝑢
= −6𝑥 − 6
𝜕𝑦

Hence + = 6𝑥 + 6 − 6𝑥 − 6 = 0

i.e., 𝑢 is a harmonic function.


Now by Milne’s method.
𝑓 (𝑧) = 𝜑 (𝑧, 0) − 𝑖𝜑 (𝑧, 0) = 3𝑧 + 6𝑧 − 𝑖(0) = 3𝑧 + 6𝑧
∴ 𝑓(𝑧) = 𝑧 + 3𝑧 + 𝐶
Example: Show that an analytic function of constant modulas is a constant.
18
Limit, Continuity, Differentiation

Proof:𝑓 (𝑧) = 𝑢 + 𝑖𝑣

∴ |𝑓(𝑧)| = √𝑢 + 𝑣 = √𝑘 ≠ 0(𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡) ______________________________ (1)


Differentiating (1) with respect to x we get,

2𝑢 + 2𝑣 =0⇒𝑢 +𝑣 = 0 ___________________________________ (2)

Differentiating (1) with respect to y we get,

𝑢 +𝑣 = 0 ⇒ −𝑢 +𝑣 = 0 ____________________________________ (3) [ By C-R equations]

Squaring (2) and (3) and adding we have,

𝜕𝑢 𝜕𝑣
(𝑢 + 𝑣 ) + (𝑢 + 𝑣 ) =0
𝜕𝑥 𝜕𝑥
𝜕𝑢 𝜕𝑣
⇒ (𝑢 + 𝑣 ) + =0
𝜕𝑥 𝜕𝑥

⇒ + =0 [∵ (𝑢 + 𝑣 ) ≠ 0]

⇒ |𝑓 (𝑧)| = 0
i.e., 𝑓 (𝑧) = 0
∴ 𝑓(𝑧) = 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡
𝝏𝟐 𝝏𝟐
Example: Show that 𝟐
+ |𝒇(𝒛)|𝟐 = 𝟒|𝒇 (𝒛)|𝟐 when f(z) is analytic functions.
𝝏𝒙 𝝏𝒚𝟐

Solution: Let 𝑓(𝑧) = 𝑢 + 𝑖𝑣. Then,


|𝑓(𝑧)| = 𝑢 + 𝑣
𝜕 𝜕𝑢 𝜕𝑣
∴ |𝑓(𝑧)| = 2𝑢 + 2𝑣
𝜕𝑥 𝜕𝑥 𝜕𝑥

|𝑓(𝑧)| = 2 +2 + 2𝑢 + 2𝑣 ___________________(1)

Similarly,

𝜕 𝜕𝑢 𝜕𝑣 𝜕 𝑢 𝜕 𝑣
|𝑓(𝑧)| = 2 +2 + 2𝑢 + 2𝑣
𝜕𝑦 𝜕𝑦 𝜕𝑦 𝜕𝑦 𝜕𝑦

=2 +2 + 2𝑢 + 2𝑣 ___________________(2) [ By C-R equations]

From (1) and (2) we get,

+ |𝑓(𝑧)| = 4 + + 2𝑢 + + 2𝑣 + ______________(3)

As both u and v are harmonic so,


19
Limit, Continuity, Differentiation

+ =0 and + =0

So from (3) we get,

𝜕 𝜕 𝜕𝑢 𝜕𝑣
+ |𝑓(𝑧)| = 4 + = 4|𝑓 (𝑧)|
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑥
𝝏𝟐 𝝏𝟐 𝝏𝟐 𝝏𝟐 𝝏𝟐
General Formula: Show that + =𝟒 𝐰𝐡𝐞𝐫𝐞 𝒛 = 𝒙 + 𝒊𝒚 provided = .
𝝏𝒙𝟐 𝝏𝒚𝟐 𝝏𝒛𝝏𝒛 𝝏𝒙𝝏𝒚 𝝏𝒚𝝏𝒙

Solution: We have 𝑧 = 𝑥 + 𝑖𝑦, 𝑧̅ = 𝑥 − 𝑖𝑦


1 1 𝑖
∴𝑥= (𝑧 + 𝑧̅) 𝑦= (𝑧 − 𝑧̅) = − (𝑧 − 𝑧̅)
2 2𝑖 2
𝜕𝑥 1 𝜕𝑥 1
∴ = , =
𝜕𝑧 2 𝜕𝑧̅ 2
𝜕𝑦 𝑖 𝜕𝑦 𝑖
=− , =
𝜕𝑧 2 𝜕𝑧̅ 2

Now = +

= −𝑖

And, = + = +𝑖
̅ ̅ ̅

𝝏𝟐 1 𝜕 𝜕 1 𝜕 𝜕
∴ = −𝑖 +𝑖
𝝏𝒛𝝏𝒛 2 𝜕𝑥 𝜕𝑦 2 𝜕𝑥 𝜕𝑦
1 𝝏𝟐 𝝏𝟐
= +
4 𝝏𝒙𝟐 𝝏𝒚𝟐

∴ + =4 [Proved]
̅

𝝏𝟐 𝝏𝟐
Example: Show that 𝟐 +
|𝒇(𝒛)|𝒌 = 𝒌𝟐 |𝒇(𝒛)|𝒌 𝟐 |𝒇(𝒛)|𝟐 where f(z) is analytic function.
𝝏𝒙 𝝏𝒚𝟐

Solution:We get,
𝜕 𝜕 𝜕
+ |𝑓(𝑧)| = 4 {|𝑓(𝑧)| }
𝜕𝑥 𝜕𝑦 𝜕𝑧𝜕𝑧̅

=4 ̅
𝑓 (𝑧)𝑓(𝑧)

=4 ̅
{𝑓 (𝑧) 𝑓 (𝑧) }

=4 {𝑓(𝑧) 𝑓(𝑧) 𝑓′(𝑧̅)}


20
Limit, Continuity, Differentiation

= 4[ 𝑓(𝑧) 𝑓 (𝑧) 𝑓 (𝑧̅) 𝑓 (𝑧̅)]

= 𝑘 𝑓 (𝑧 ) 𝑓 (𝑧̅) 𝑓′(𝑧)𝑓′(𝑧̅)

= 𝑘 𝑓 (𝑧)𝑓(𝑧) |𝑓 (𝑧)|

= 𝑘 |𝑓(𝑧)| |𝑓′(𝑧)| [Proved]

SINGULARITIES AND SINGULAR POINT


Definition of Singularity: A Point at which a function 𝑓(𝑧) is not analytic or fails to be analytic is called a
singular point or singularity of 𝑓(𝑧).
Definition of Pole: A singularity 𝑧 of 𝑓(𝑧) is called a pole of order n if lim → (𝑧 − 𝑧 ) 𝑓(𝑧) exist
and ≠ 0.
For example,

If 𝑓(𝑧) = ( ) ( )
then

𝑧 = 2 is a pole of order 3 of 𝑓(𝑧).


z=-3 is pole of order 1 of 𝑓(𝑧).
Isolated Singularities: A singularity 𝑧 of 𝑓(𝑧) is said to be isolated singularity if there exists a
neighborhood N of 𝑧 which contains no singularity other than 𝑧 .
Removable Singularity: A singularity 𝑧 of f(z) is called a removable singularity if lim → 𝑓(𝑧) exists.

For example,
sin 𝑧
𝑓 (𝑧 ) =
𝑧
𝑧 = 0 is a singular point.

But lim → =1

Therefore 𝑧 = 0 is a removable singularity.


Essential Singularity: A singularity 𝑧 of 𝑓(𝑧) which is neither a pole nor an isolated singularity is called
essential singularity.

For example, 𝑓(𝑧) = 𝑒 has an essential singularity at 𝑧 = 2.


Singularity at infinity: A point 𝑧 is called a singularity at infinity of 𝑓(𝑧) if 𝑧 is a pole of order m of
𝑓 .

Definition: A function f(z) has a singularity at infinity if 𝑓 has a singularity at 𝑤 = 0.

For example,
21
Limit, Continuity, Differentiation
1 1
𝑓(𝑧) = 𝑧 ∴ 𝑓 =
𝑤 𝑤

Here 𝑤 = 0 is a pole of order 3 of 𝑓 .

Hence 𝑧 = ∞ is a pole of order 3 of 𝑓(𝑧).


21
Complex Integration
Definition: let 𝑓(𝑧) be continuous on a curve C of finite length. let us divide the curve by the points,
𝑎 = 𝑧 , 𝑧 , 𝑧 , ......., 𝑧 = 𝑏

Now for the sum, Sn = ∑ 𝑓(𝜉 )∆𝑧

where ∆𝑧 = zk - zk-1 and 𝜉 is any point on the


curve between 𝑧 and 𝑧 .
Let us increase the number of sub division in such a
way that the largest of the chord length |∆𝑧 |
approaches to zero . Then Sn approaches to a value
which is known as complex line integral of
𝐿𝑖𝑚
𝑓(𝑧) along 𝐶 denoted by ∫ 𝑓(𝑧)𝑑 𝑧 or ∫ 𝑓(𝑧)𝑑𝑧 . I.e., ∫ 𝑓 (𝑧)𝑑 𝑧 = ∑𝑛 𝑓(𝜉𝑘 )∆𝑧𝑘
𝑛 →∝ 𝑘=1
property of complex integration:

1. ∫ (𝑓(𝑧) + 𝑔(𝑧))𝑑𝑧 = ∫ 𝑓(𝑧)𝑑𝑧


2. ∫ 𝑘𝑓(𝑧)𝑑𝑧 = 𝑘 ∫ (𝑧)𝑑𝑧
3. ∫ (𝑧)𝑑𝑧 = − ∫ 𝑓(𝑧)𝑑𝑧
4. ∫ 𝑓(𝑧)𝑑𝑧 = ∫ 𝑓(𝑧)𝑑𝑧 + ∫ 𝑓 (𝑧)𝑑𝑧
5. | ∫ 𝑓(𝑧)𝑑𝑧 | ≤ ML where k is the length of the curve 𝑐 and |𝑓(𝑧)| ≤ 𝑀 𝑧 𝜖 𝐶

𝐿𝑖𝑚
Proof of (5) : by definition ∫ 𝑓(𝑧)𝑑 𝑧 = ∑𝑛 𝑓(𝜉𝑘 )∆𝑧𝑘
𝑛 →∝ 𝑘=1
Now, ∑ 𝑓(𝜉 )∆𝑧 ≤ ∑𝑛𝑘=1 𝑓 𝜉 |∆𝑧𝑘 |
≤ 𝑀∑ |∆ 𝑧 |
≤ 𝑀𝐿

As ∑ |∆ 𝑍 | is the sum of all chord length which must be less than the length of the curve. Taking limit
on both side as n tends to infinity we have by definition

| 𝑓 (𝑧)𝑑𝑧| ≤ 𝑀𝐿

Definition: A closed curve which does not intersect itself is called a simple closed curve.
Definition: A region R is called simply connected region if every simple closed curve in R can be shrank to
a point without leaving R . Otherwise R is called multiply connected region.

Greens Theorem: If 𝑃(𝑥, 𝑦) and 𝑄(𝑥, 𝑦) are continuous and have continuous partial derivatives inside and
on a simple closed curve 𝐶 Then ∫ 𝑃𝑑𝑥 + 𝑄𝑑𝑦 = ∬ − 𝑑𝑥𝑑𝑦 where 𝑅 is the region bounded by
𝐶.
The theorem is valid for both simply and multiply-connected regions.
22
Complex Integration
Ex : Prove Green’s theorem in the plane if C is a simple closed curve which has the property that any
straight line parallel to the coordinate axes cuts C in at most two points

Proof:
Let the equation of the curve EGF and EHF be 𝑦 = 𝑌 (𝑥) and
𝑦 = 𝑌 (𝑥) respectively then,

∬ ( 𝑑𝑦)𝑑𝑥 = ∫ [∫ 𝑑𝑦]𝑑𝑥
( )
=∫ 𝑃(𝑥, 𝑦) | ( ) 𝑑𝑥

= ∫ [𝑃(𝑥, 𝑌 ) − 𝑃(𝑥, 𝑌 )]𝑑𝑥


=∫ 𝑃 (𝑥, 𝑌 )𝑑𝑥 − ∫ 𝑃(𝑥, 𝑌 )𝑑𝑥
=− ∫ 𝑃(𝑥, 𝑌 )𝑑𝑥 − ∫ 𝑃 (𝑥, 𝑌 )𝑑𝑥

=− 𝑃𝑑𝑥

Then, ∮ 𝑃𝑑𝑥 = − ∬ 𝑑𝑥𝑑𝑦 ------- (1)

Now let the equation of the curve GEH and GFH be 𝑥 = 𝑋 (𝑦) and 𝑥 = 𝑋 (𝑦) respectively
Thus, ∬ 𝑑𝑥𝑑𝑦 = ∫ ∫ 𝑑𝑥 𝑑𝑦
= ∫ [𝑄(𝑥 , 𝑦) − 𝑄(𝑥 , 𝑦)]𝑑𝑦
= ∫ 𝑄(𝑥 , 𝑦)𝑑𝑦 + ∫ 𝑄(𝑥 , 𝑦)𝑑𝑦

= ∮ 𝑄𝑑𝑦

Then ∮ 𝑄𝑑𝑦 = ∬ 𝑑𝑥𝑑𝑦 ------------------ (2)

Adding (1) ,(2) we get ,


𝛿𝑄 𝛿𝑝
𝑝𝑑𝑥 + 𝑄𝑑𝑦 = − 𝑑𝑥𝑑𝑦
𝛿𝑦 𝛿𝑦
Ex : Verify Greens theorem in the plane for

(2𝑥𝑦 − 𝑥 )𝑑𝑥 + (𝑥 + 𝑦 )𝑑𝑦

Where C is the closed curve in the region bounded by


𝑦 = 𝑥 and 𝑥 = 𝑦

Solution: Here 𝑃 = 2𝑥𝑦 − 𝑥 and 𝑄 = 𝑥 + 𝑦 . The plane


curves 𝑦 = 𝑥 and 𝑥 = 𝑦 in and intersect at (0,0) and
23
Complex Integration
(1,1). The positive direction of the curve 𝐶 is as shown in the figure.
Along 𝑦 = 𝑥 , the line integral
𝐼 =∫ {(2𝑥)(𝑥 ) − 𝑥 }𝑑𝑥 + {𝑥 + (𝑥 ) }𝑑(𝑥 )
=∫ (2𝑥 + 𝑥 + 52)𝑑𝑥 =
Along 𝑦 = 𝑥, the line integral
𝐼 =∫ {2(𝑦 )(𝑦) − (𝑦 ) }𝑑(𝑦 ) + {{𝑦 + 𝑦 }𝑑𝑦
=∫ (4𝑦 − 2𝑦 + 2𝑦 )𝑑𝑦=−
So the Line integral = − = . On the other hand

∬ ( − )𝑑𝑥𝑑𝑦=∬ ( (𝑥 + 𝑦 ) − (2𝑥𝑦 − 𝑥 )𝑑𝑥𝑑𝑦

=∬ ( (1 − 2𝑥)𝑑𝑥𝑑𝑦
𝑑𝑥𝑑𝑦

=∫ ∫ (1 − 2𝑥)𝑑𝑦𝑑𝑥

=∫ (𝑦 − 2𝑥𝑦)|√ dx

=∫ (𝑥 − 2𝑥 − 𝑥 + 2𝑥 )𝑑𝑥

=
Hence the Green’s theorem is verified.

Ex : Extend the proof Green’s theorem in the plane to the curve C for which lines parallel to the coordinate
axes may cut C in more than two points
Solution :
Consider a simple closed curve C such as shown
in the figure in which lines parallel to the axes
may meet C in more than two points. By
constructing line ST, the region is divided into
two regions R1 and R2 whose boundaries are
STUS and SVTS respectively which are of the
type that any straight line parallel to the
coordinate axes cuts C and C2 atmost
most two points
and for which Green’s theorem applies, i.e.
∮ 𝑝𝑑𝑥 + 𝑄𝑑𝑦 = ∬ − 𝑑𝑥𝑑𝑦 ……………………..………(1)

∮ 𝑝𝑑𝑥 + 𝑄𝑑𝑦 = ∬ − 𝑑𝑥𝑑𝑦 ……………………..………(2)


Adding the left-hand
hand sides of (1) and (2), we have, omitting the integrand 𝑃𝑑𝑥 + 𝑄𝑑𝑦 in each case
case.

+ = + + + = + =

using the fact that using the fact that ∫ = − ∫


Adding the right-hand
hand sides of (1) and (2), omitting the integrand,
24
Complex Integration

+ =

Then
𝛿𝑄 𝛿𝑝
𝑝𝑑𝑥 + 𝑄𝑑𝑦 = − 𝑑𝑥𝑑𝑦
𝛿𝑦 𝛿𝑦
and the theorem is proved.
NB: We have proved Green’s theorem for the simply-connected region bounded by the simple closed curve
C. For more complicated regions, it may be necessary to construct more lines, such as ST, to establish the
theorem.
Green’s theorem is also true for multiply-connected regions.

Theorem (Cauchy’s theorem) : If 𝑓(𝑥) is analytic inside and on a simple closed curve C and if 𝑓′(𝑧) is
continuous Then
∮ 𝑓(𝑧)𝑑𝑧 = 0

Proof. Let 𝑓 (𝑧) = 𝑢 + 𝑖𝑣


Then 𝑓 (𝑧) = +𝑖
= −𝑖 [by C-R equation]

Here 𝑢, 𝑣, , , , were all continuous

hence ∮ 𝑓(𝑧)𝑑𝑧 = ∮ (𝑢 + 𝑖𝑣 )(𝑑𝑥 + 𝑖𝑑𝑦)


= ∮ (𝑢𝑑𝑥 − 𝑣𝑑𝑦) + 𝑖 ∮ (𝑣𝑑𝑥 + 𝑢𝑑𝑦)
=∬ − − 𝑑𝑥𝑑𝑦 + 𝑖 ∬ − 𝑑𝑥𝑑𝑦
=0
Where 𝑅 is the region bounded by 𝐶

Theorem : (Cauchy’s fundamental Theorem) :

If 𝑓(𝑧) is analytic inside and on a simple closed curve 𝐶 then ∮ 𝑓 (𝑧)𝑑𝑧 = 0.


Or,
If 𝑓(𝑧)𝑑𝑧 is analytic in a region and on the boundary bounded by a simple closed curve C Then
∮ 𝑓(𝑧)𝑑𝑧 = 0.
Proof. To proof we need the followi ng lemm a

Lem ma 1: (i) ∮ 𝑑𝑧
(ii) ∮ 𝑧𝑑𝑧

( ) ( )
Lem ma 2: Given any ε > 0 , it is possi ble to d ivide the region in side 𝐶 into finite num bers of m ashes either complete or partial squares such that w ithin each m esh there exist a point z0 where − f (z ) < 𝜀 for all values of 𝑧 in the mesh.

Now proof of le mma 1

Let C be the curve with end poin ts a and b D ivided c by the poin ts a = z1,z2 ,z3… ..zn= b then,

(𝑖) 𝑑𝑧 = lim 𝑓(𝜗) ∆𝑧

= lim ∑ (𝑧 − 𝑧 )
= lim (𝑧 − 𝑧 )
= b-a
Since is the closed curve so a= b

∴ 𝑑𝑧 = 0

(ii) again by definition,

∫ 𝑧𝑑𝑧 = lim ∑ 𝑓 (𝜗 )∆𝑧 | where 𝜗 ∈ [𝑧 , 𝑧 ]


= ∴ ∫ 𝑧𝑑𝑧 = lim ∑ 𝜗 ( ∆𝑧 − 𝑧 )
= ∴ ∫ 𝑧𝑑𝑧 = lim ∑ 𝜗 ( ∆𝑧 − 𝑧 ) ------------ (1)
= ∴ ∫ 𝑧𝑑𝑧 = lim ∑ 𝜗 (∆𝑧 − 𝑧 ) --------------- (2)

Now from (1) and (2)

∴ ∫ 𝑧𝑑𝑧 = lim ∑ (∆𝑧 − 𝑧 ) + ( ∆𝑧 − 𝑧 )


= lim ∑ ( ∆𝑧 − 𝑧 )
= lim (𝑧 − 𝑧 + 𝑧 − 𝑧 + … … … … … + 𝑧 − 𝑧 )
= lim 𝑧 − 𝑧)
= (𝑏 − 𝑎 )

Since c is closed so a= b

∴ 𝑧𝑑𝑧 = 0

P roof of lemma 2:

Consi der the square 𝑆 is in closing 𝐶. Divides 𝑆 into fin ite numbers of small equal squares. Then the region in side 𝐶 is divided int o fini te num ber of m ashes ( either complete or partial squares).
( ) ( )
Suppo se − 𝑓 (𝑧 ) | <∈ fails at least on one mesh, then s ub div ide the m esh by joining it s midpoin t of si des of these st ill rem ain any parts which don ’t sat isfy (i) we again sub div ided it in the sam e way. The process may end after a finite number of steps when (i) is sat isfied for every sub divisi on or the process m ay go on indefinitely .
in the 2nd case we get a decreasing square of meshes inside here (i) is not satisfied. T hen there exist a po int 𝑧 inside each m esh of the sequence
()– ( )
such that − 𝑓 (𝑧) ≮ 𝜀 ; whenever |𝑧 − 𝑧 | is very small.
This im plies that 𝑓(𝑧 ) is not analy tic at 𝑧 which is a po int insi de 𝐶 or on 𝐶 . This g ives a contradiction. S o the lem ma is proved.

Proof of the theorem: By lem ma (2) ,


𝑓 (𝑧) = 𝑓 ( 𝑧 ) + 𝑓’ (𝑧 ) ( 𝑧 – 𝑧 ) + ɳ (𝑧 – 𝑧 )
w here ɳ → 0 as z → 𝑧
i.e, |ɳ| < 𝜀 whenever z – 𝑧 is very sm all .
Hence, ∫ 𝑓(𝑧) 𝑑𝑧 = 𝑓( 𝑧 ) ∫ 𝑑𝑧+ 𝑓′(𝑧 ) ∫ (𝑧 − 𝑧 )𝑑𝑧
=∮ ɳ( z – 𝑧 ) 𝑑𝑧 ; by lem ma(1)
Now , ∮ ɳ ( z – 𝑧 ) 𝑑𝑧 = ∑ ∫ ɳ( z – 𝑧 ) 𝑑𝑧 + ∑ ∫ ɳ( z – 𝑧 ) 𝑑𝑧
W here 𝑐 𝑎𝑛𝑑 𝑝 represent complete and partial sequence inside C.
Integration alon g each sides of each squares (complete or partial) happen to be ta ken twice in opposite direction . And so they cancel each other ,
Then ∫ ɳ z – 𝑧 dz ≤ ∑ . ∫ ɳ z – 𝑧 𝑑𝑧 + ∑ . ∫ ɳ z – 𝑧 𝑑𝑧
Now , ∫ ɳ z – 𝑧 𝑑𝑧 = 𝜀 𝑙 √2 4𝑙 = 𝜀 √2 4𝐴 ; where l is the length of each sides each small square , 𝐴 𝑖𝑠 𝑡ℎ𝑒 area of sm all square and z – 𝑧 is very small .
Hence , ∑ . ∫ ɳ z – 𝑧 𝑑𝑧 < 𝜀 4 √2 ∑ 𝐴
Again, ∫ ɳ z – 𝑧 𝑑𝑧 ≤ 𝜀 𝑙√2 (4l + 𝓈 )
= 𝜀 √ 2(4𝐴 + 𝑙𝓈 ) ; where 𝓈 is the lengt h of C in 𝑝 .
Hence ∑ . ∫ ɳ z – 𝑧 𝑑𝑧 ≤ 𝜀 √2( 4∑ 𝐴 + 𝑙∑ 𝓈 ) = 𝜀 √2( 4∑ 𝐴 + 𝑙.𝐿)
W here L is the length of C.

Hence ∮ 𝑓(𝑧) 𝑑𝑧 ≤ 4 𝜀√2 (∑ 𝐴 + ∑ 𝐴 ) + 𝜀 √2 𝑙. 𝐿


≤ 𝜀 √2 (4A + l.L) ; W here A is the area of square S since 𝜀 is arbitrate
So, ∮ 𝑓(𝑧) 𝑑𝑧 = 0 i,e. ∮ 𝑓 (𝑧)𝑑𝑧 = 0
25
Complex Integration
Theorem : Prove Cauchy theorem for a triangle .
Proof : We know that if 𝐶 is a simple closed curve then (i) ∮ 𝑑𝑧 = 0 and (ii) ∮ 𝑧𝑑𝑧 = 0
Let 𝑓(𝑧) is analytic inside and on triangle ABC denoted by ∆. We need to prove that, ∮∆ 𝑓(𝑧)𝑑𝑧 = 0. Let
D, E, and F are the middle points of BC , CA , AB Respectively. Join
DE , EF , and FD then we obtain four triangle ∆ , ∆ , ∆ and ∆ .
Now ,
∮∆ 𝑓(𝑧)𝑑𝑧 = ∮ 𝑓(𝑧)𝑑𝑧 + ∮ 𝑓(𝑧)𝑑𝑧 + ∮ 𝑓(𝑧)𝑑𝑧

=(∫ + ∫ )+(∫ + ∫ )+(∫ + ∫ )+


(∫ + ∫ +∫ )

[ ∵ ∫ =−∫ ]
= ∫ ∆ + ∫∆ + ∫ ∆ + ∫ ∆
Hence , ∫∆ 𝑓(𝑧)𝑑𝑧 ≤ ∫∆ 𝑓(𝑧)𝑑𝑧 + ∫∆ 𝑓(𝑧)𝑑𝑧 + ∫∆ 𝑓(𝑧)𝑑𝑧 + ∫∆ 𝑓(𝑧)𝑑𝑧
Let ∆ represents that triangle above four so that ∫∆ 𝑓(𝑧)𝑑𝑧 is the largest ( if these are more triangles
with these property then we can choose any one of them ∆ )
Thus , ∫∆ 𝑓(𝑧)𝑑𝑧 ≤ 4 ∫∆ 𝑓(𝑧)𝑑𝑧
Now joining the midpoints of the triangle ∆ 𝑎𝑛𝑑 proceeding as above we obtain a triangle ∆
So that ∫∆ 𝑓(𝑧)𝑑𝑧 ≤ 4 ∫∆ 𝑓(𝑧)𝑑𝑧
Proceeding similarly we obtain a sequence of triangle {∆ }
Such that , ∆> ∆ > ∆ >………..
And ∫∆ 𝑓(𝑧)𝑑𝑧 ≤ 4 ∫∆ 𝑓(𝑧)𝑑𝑧
Then there exist a point 𝑧 such that 𝑧 ∈ ∆ for each 𝑘 = 1,2, … . . 𝑛.

Since 𝑓(𝑧) is analytic at 𝑧 then for any 𝜀 > 0 there exist 𝛿 > 0 such that
𝑓(𝑧) = 𝑓(𝑧 ) + 𝑓 ‘(𝑧 ) (𝑧 − 𝑧 ) + ɳ(𝑧 − 𝑧 ) ; where |ɳ| < 𝜀 𝑤ℎ𝑒𝑛𝑒𝑣𝑒𝑟 |𝑧 − 𝑧 | < 𝛿
∫∆ 𝑓(𝑧)𝑑𝑧 = 𝑓(𝑧 ) ∫∆ 𝑑𝑧 + 𝑓′(𝑧 )∫∆ (𝑧 − 𝑧 )𝑑𝑧 + ∫∆ ɳɳ(𝑧 − 𝑧 )𝑑𝑧
= ∫∆ ɳɳ(𝑧 − 𝑧 )𝑑𝑧 ; by (i) and (ii)
Now for any 𝑧 on ∆ |𝑧 − 𝑧 | < ; where p is the perimeter of ∆.
Now |ɳ(𝑧 − 𝑧 )| = |ɳ||(𝑧 − 𝑧 )| < 𝜀 .
Choose 𝑛 sufficiently large so that < 𝛿.
Then ∫∆ ɳɳ(𝑧 − 𝑧 )𝑑𝑧 < 𝜀 ; when ever |𝑧 − 𝑧 | < < 𝛿

Thus ∫∆ 𝑓(𝑧)𝑑𝑧 ≤ ; when ever |𝑧 − 𝑧 | < < 𝛿


Therefore ∫∆ 𝑓(𝑧)𝑑𝑧 ≤ 4 ∫∆ 𝑓(𝑧)𝑑𝑧 ≤ 𝜀𝑝
Since 𝜀 𝑖𝑠 arbitrate so ∫∆ 𝑓(𝑧)𝑑𝑧 = 0 i.e. ∫∆ 𝑓(𝑧)𝑑𝑧 = 0
26
Complex Integration
Ex. Prove Cauchy’s theorem for a polygon.
Proof : Let ABCDEF be a polygon. Suppose 𝑓(𝑧) is analytic inside and on the polygon .we need to prove
that ∫ 𝑓(𝑧)𝑑𝑧 = 0. Join AC , AD , AE then
∫ 𝑓(𝑧)𝑑𝑧 = ∫ + ∫ + ∫ + ∫
= (∫ + ∫ ) + (∫ +∫ + ∫ ) + (∫ + ∫ + ∫ )
+(∫ + ∫ )
=∫ + ∫ + ∫ + ∫ = 0+ 0 + 0 + 0 = 0 .
[ As the Cauchy’s theorem holds for triangle]

Ex. If 𝑓(𝑧) is a analytic inside and on the region bounded by 𝐶 𝑎𝑛𝑑 𝐶 as shown by the figure. Then
∮ 𝑓(𝑧)𝑑𝑧 = ∮ 𝑓(𝑧)𝑑𝑧
Proof : Let ABCDA represent the curve 𝐶 𝑎𝑛𝑑 𝐸𝐹𝐺𝐻 𝑟𝑒𝑝𝑟𝑒𝑠𝑒𝑛𝑡 𝐶 .
Construct a cross cut AE then AEHGFEABCDA becomes a closed curve.
Moreover 𝑓(𝑧) is analytic inside and on the region. Then by Cauchy’s
theorem
∫ 𝑓(𝑧)𝑑𝑧 = 0
⇒∫ + ∫ + ∫ + ∫ =0
⇒ ∮ 𝑓(𝑧)𝑑𝑧 − ∮ 𝑓(𝑧)𝑑𝑧 = 0
∴ ∮ 𝑓(𝑧)𝑑𝑧 = ∮ 𝑓(𝑧)𝑑𝑧

Ex : Suppose 𝑓(𝑧) is analytic in a simply-connected region 𝑅. Prove that ∫ 𝑓(𝑧)𝑑𝑧 is independent of the
path in 𝑅 joining any two points 𝑎 and 𝑏 in 𝑅.

Proof: Consider the two paths 𝐶 = 𝐴𝐷𝐵 and 𝐶 = 𝐴𝐸𝐵 from 𝑎 to 𝑏 in 𝑅 as shown by the figure. Now we
have to show that
∫ 𝑓(𝑧)𝑑𝑧 = ∫ 𝑓(𝑧)𝑑𝑧
Clearly 𝐴𝐸𝐵𝐷𝐴 is a simple closed curve in 𝑅 Then by Cauchy’s
theorem
∫ 𝑓(𝑧)𝑑𝑧 = 0
or, ∫ 𝑓(𝑧)𝑑𝑧 + ∫ 𝑓(𝑧)𝑑𝑧 = 0
or, ∫ 𝑓(𝑧)𝑑𝑧 − ∫ 𝑓(𝑧)𝑑𝑧 = 0
or, ∫ 𝑓(𝑧)𝑑𝑧 − ∫ 𝑓(𝑧)𝑑𝑧 = 0
or, 𝑓(𝑧)𝑑𝑧 = ∫ 𝑓(𝑧)𝑑𝑧
27
Cauchy’s Integral Formulas and Related Theorems
Cauchys Integral Formula: If 𝑓(𝑧) is analytic inside and on a simple closed curve 𝐶 and 𝑎 is any point inside
𝐶. Then,

1 𝑓(𝑧)
𝑓(𝑎) = 𝑑𝑧
2𝜋𝑖 𝑧−𝑎

( )
Proof: Construct a circle Γ inside 𝐶 with centre at 𝑎 and radius 𝑟. Then is analytic inside the region
bounded by Γ and 𝐶 and also on the boundary. Then by cauchys theorem for simple connected region we get

( ) ( )
∮ 𝑑𝑧 = ∮ 𝑑𝑧 ……………………………………(𝑖)

Now on Γ

|𝑧 − 𝑎| = 𝑟

i.e.,

𝑧 − 𝑎 = 𝑟𝑒

i.e.,

𝑧 = 𝑎 + 𝑟𝑒 ⇒ 𝑑𝑧 = 𝑖 𝑟𝑒 𝑑𝜃 where 0 ≤ 𝜃 ≤ 2𝜋

Hence

𝑓(𝑧) 𝑓 𝑎 + 𝑟𝑒
𝑑𝑧 = 𝑖𝑟𝑒 𝑑𝜃
𝑧−𝑎 𝑟𝑒

i.e.,

𝑓(𝑧)
𝑑𝑧 = 𝑖 𝑓(𝑎 + 𝑟𝑒 )𝑑𝜃 [𝑏𝑦(𝑖)]
𝑧−𝑎

Taking limit on both sides as 𝑟 → 0 we have


𝑓(𝑧)
𝑑𝑧 = 𝑖 lim 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃
𝑧−𝑎 →

= 𝑖∫ lim → 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃 [as 𝑓(𝑧) is analytic]

= 𝑖∫ lim → 𝑓(𝑎)𝑑𝜃 [as 𝑓(𝑧) continous at 𝑎]

= 2𝜋𝑖 𝑓(𝑎)
28
Cauchy’s Integral Formulas and Related Theorems
Hence

1 𝑓(𝑧)
𝑓(𝑎) = 𝑑𝑧
2𝜋𝑖 𝑧−𝑎

0 if 𝑎 is outside 𝐶
Ex. Show that ∮ =
2𝜋𝑖 if 𝑎 is inside 𝐶

Solution: If a is outside 𝐶 then is analytic inside and on 𝐶. So by cauchys theorem

𝑑𝑧
=0 [𝑝𝑟𝑜𝑣𝑒𝑑]
𝑧−𝑎

If 𝑎 is inside C, then put 𝑓(𝑧) = 1

Then by cauchys integral formula

1 𝑓(𝑧)
𝑓(𝑎) = 𝑑𝑧
2𝜋𝑖 𝑧−𝑎

i. e.,

1 𝑑𝑧
1=
2𝜋𝑖 𝑧−𝑎

i. e.,

𝑑𝑧
= 2𝜋𝑖
𝑧−𝑎

Theorem: If 𝑓(𝑧) is analytic inside and on a simple closed curve 𝐶 and 𝑎 is any pint inside 𝐶, Then

1 𝑓(𝑧)
𝑓′(𝑎) = 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎)²

Proof: Choose a point 𝑎 + ℎ inside 𝐶. Then any cauchys integral formula

1 𝑓(𝑧)
𝑓(𝑎 + ℎ) = 𝑑𝑧
2𝜋𝑖 𝑧−𝑎−ℎ

And

1 𝑓(𝑧)
𝑓(𝑎) = 𝑑𝑧
2𝜋𝑖 𝑧−𝑎
29
Cauchy’s Integral Formulas and Related Theorems
Now

𝑓(𝑎 + ℎ) − 𝑓(𝑎) 1 1 1
= − 𝑓(𝑧)𝑑𝑧
ℎ ℎ2𝜋𝑖 𝑧−𝑎−ℎ 𝑧−𝑎

1 ℎ
= 𝑓(𝑧) 𝑑𝑧
ℎ2𝜋𝑖 (𝑧 − 𝑎 − ℎ)(𝑧 − 𝑎)

1 (𝑧 − 𝑎)
= 𝑓(𝑧)𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎 − ℎ)(𝑧 − 𝑎)

1 (𝑧 − 𝑎 − ℎ) + ℎ
= 𝑓(𝑧)𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎 − ℎ)(𝑧 − 𝑎)

( ) ( )
= ∮ 𝑑𝑧 + ∮ 𝑑𝑧
( )² ( )( )²

Construct a circle Γ is side C with centre at 𝑎 and radius 𝑟. Choose |ℎ| small enough so that 𝑎 + ℎ lies inside Γ
and |ℎ| < . Since 𝑓(𝑧) is analytic inside and on 𝐶. So |𝑓(𝑧)| < 𝑀 for some +ve integer M. Now on Γ,

|𝑧 − 𝑎| = 𝑟

𝑟 𝑟
|𝑧 − 𝑎 − ℎ| ≥ |𝑧 − 𝑎| − |ℎ| > 𝑟 − =
2 2

Hence

1 ℎ 𝑓(𝑧) 𝑑𝑧 1 ℎ 𝑓(𝑧) 𝑑𝑧 1 |ℎ|𝑀


= < 2𝜋𝑟
2𝜋𝑖 (𝑧 − 𝑎 − ℎ)(𝑧 − 𝑎)² 2𝜋𝑖 (𝑧 − 𝑎 − ℎ)(𝑧 − 𝑎)² 2𝜋 𝑟²

2𝑀|ℎ|
= → 0 𝑎𝑠 ℎ → 0
𝑟²

Hence

𝑓(𝑎 + ℎ) − 𝑓(𝑎) 1 𝑓(𝑧) 𝑑𝑧


lim = +0
→ ℎ 2𝜋𝑖 (𝑧 − 𝑎)²

i.e.,

1 𝑓(𝑧) 𝑑𝑧
𝑓 (𝑎) =
2𝜋𝑖 (𝑧 − 𝑎)²
30
Cauchy’s Integral Formulas and Related Theorems
(Proved)

Cauchys general integral formula: If 𝑓(𝑧) is analytic inside and on a simple closed curve 𝐶 and 𝑎 is any point
inside C. Then

𝑛! 𝑓(𝑧)
𝑓 (𝑎) = 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎)

Proof: We use the method of induction. The formula holds for 𝑛 = 0 [𝑐𝑎𝑢𝑐ℎ𝑦𝑠 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑙 𝑓𝑜𝑟𝑚𝑢𝑙𝑎]

By the preceding theorem this formula holds for 𝑛 = 1 ( একটা অ ত দখােত হেব)

Suppose the formula holds for formula holds for 𝑛 = 𝑚 − 1. Then

(𝑚 − 𝑎)! 𝑓(𝑧)
𝑓 (𝑎) = 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎)

Choose any 𝑎 + ℎ inside 𝐶 then

𝑓 (𝑎 + ℎ) − 𝑓 (𝑎) (𝑚 − 1)! 1 1
= − 𝑓(𝑧) 𝑑𝑧
ℎ ℎ2𝜋𝑖 (𝑧 − 𝑎 − ℎ) (𝑧 − 𝑎)

(𝑚 − 1)! 1 1
= − 𝑓(𝑧) 𝑑𝑧
ℎ2𝜋𝑖 (𝑧 − 𝑎) (1 − ) (𝑧 − 𝑎)

(𝑚 − 1)! 1 ℎ
= 1− − 1 𝑓(𝑧) 𝑑𝑧
ℎ2𝜋𝑖 (𝑧 − 𝑎) 𝑧−𝑎

(𝑚 − 1)! 1 𝑚ℎ 𝑚(𝑚 + 1)ℎ


= 1+ + + ⋯ + ℎ𝑖𝑔ℎ𝑒𝑟 𝑝𝑜𝑤𝑒𝑟 𝑜𝑓 ℎ − 1 𝑓(𝑧)𝑑𝑧
ℎ2𝜋𝑖 (𝑧 − 𝑎) 𝑧−𝑎 2! (𝑧 − 𝑎)

(𝑚 − 1)! 𝑚 𝑚(𝑚 + 1) ℎ
= + + ⋯ + ℎ𝑖𝑔ℎ𝑒𝑟 𝑝𝑜𝑤𝑒𝑟 𝑜𝑓 ℎ 𝑓(𝑧)𝑑𝑧
ℎ2𝜋𝑖 (𝑧 − 𝑎) 2! (𝑧 − 𝑎)

Taking limit on the both side as h→0 we get

𝑓 (𝑎 + ℎ) − 𝑓 (𝑎)
lim
→ ℎ
(𝑚 − 1)! 𝑚 𝑚(𝑚 + 1)ℎ
= lim + + ⋯ ℎ𝑖𝑔ℎ𝑒𝑟 𝑝𝑜𝑤𝑒𝑟 𝑜𝑓 ℎ 𝑓(𝑧)𝑑𝑧
→ 2π𝑖 (𝑧 − 𝑎) 2! (𝑧 − 𝑎)

i.e.,
31
Cauchy’s Integral Formulas and Related Theorems

(𝑚 − 1)! 𝑓(𝑧)
𝑓 (𝑎) = 𝑓(𝑧)𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎)

i.e.,

𝑚! 𝑓(𝑧)
𝑓 (𝑎) = 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎)

i.e., the formula holds for n=m. Hence by the method of inductions the formula holds for all +ve integer n.

Cauchys inequality: If 𝑓(𝑧) is analytic inside and on a circle 𝐶 with centre at 𝑎 and radius 𝑟 and if
|𝑓(𝑧)| < 𝑀 on C. then

𝑀𝑛!
|𝑓 | ≤
𝑟

Proof: By Cauchys general integral formula

𝑛! 𝑓(𝑧)
𝑓 (𝑎) = 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑎)

Hence

𝑛! 𝑓(𝑧) 𝑛! 𝑀
|𝑓 (𝑎)| = 𝑑𝑧 ≤ 2𝜋𝑟
2𝜋𝑖 (𝑧 − 𝑎) 2𝜋 𝑟

[∴ length of 𝐶 = 2𝜋𝑟 and on 𝐶, (𝑧 − 𝑎) = 𝑟, |𝑓(𝑧)| < 𝑀]

i.e.,

!
|𝑓 (𝑎)| ≤ [𝑃𝑟𝑜𝑣𝑒𝑑]

Gauss mean value theorem: If 𝑓(𝑧) is analytic inside and on a circle 𝐶 with centre at 𝑎 and radius 𝑟, then

𝑓(𝑎) = ∫ 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃

Proof: By Cauchys integral formula

1 𝑓(𝑧)
𝑓(𝑎) = 𝑑𝑧
2𝜋𝑖 𝑧−𝑎

On 𝐶,

𝑧 − 𝑎 = 𝑟𝑒
32
Cauchy’s Integral Formulas and Related Theorems
i. e.,

𝑑𝑧 = 𝑖𝑟𝑒 𝑑𝜃

1 𝑓 𝑎 + 𝑟𝑒 𝑖𝑟𝑒 𝑑𝜃
𝑓(𝑎) =
2𝜋𝑖 𝑟𝑒

1
= 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃
2𝜋

Ex. If in a region R, ∮ 𝑓(𝑧)𝑑𝑧 = 0 for every simple closed curve 𝐶.Then for any two points 𝑎, 𝑏 ∈ 𝑅 .
∫ 𝑓(𝑧)𝑑𝑧 is independent of path in the region.

Proof: Let 𝐴 = 𝑎 and 𝐵 = 𝑏 two points in the region and let ACD and ABD be two paths on the region from A
to B. Then ABCDA is a simple closed curve in the region. By the given
condition

𝑓(𝑧)𝑑𝑧 = 0

Or, ∫ +∫ =0

=− =

Hence ∫ 𝑓(𝑧)𝑑𝑧 is independent of the path in 𝑅.

Moreras Theorem (convert to Cauchys theorem): If 𝑓(𝑧) is continuous on a simple connected region 𝑅 and
if ∮ 𝑓(𝑧) 𝑑𝑧 = 0 for every simple closed curve 𝐶 in 𝑅. Then 𝑓(𝑧) analytic in 𝑅.

Proof: Fix a point 𝑎 in 𝑅.

For any 𝑧 ∈ 𝑅, we define

𝐹(𝑧) = 𝑓(𝑢) 𝑑𝑢

Choose 𝑧 + ∆𝑧 in 𝑅. Then
33
Cauchy’s Integral Formulas and Related Theorems

𝐹(𝑧 + ∆𝑧) = 𝑓(𝑢) 𝑑𝑢

Now

( ∆ ) ( ) ∆
− 𝑓(𝑧)= ∫ 𝑓 (𝑢) 𝑑𝑢 − ∫ 𝑓 (𝑢) 𝑑𝑢 − 𝑓(𝑧)
∆ ∆


= ∫ 𝑓 (𝑢) 𝑑𝑢 − 𝑓(𝑧)


= ∫ 𝑓 (𝑢) − 𝑓 (𝑧) 𝑑𝑢 [Here z is constant and u is

variable]


Since ∮ 𝑓(𝑢)𝑑𝑢 = 0 for all simple closed curve 𝐶 in 𝑅. So ∮ {𝑓(𝑢) − 𝑓(𝑧)} is independent of path in 𝑅
so long as the path inside 𝑅.

So we can choose a path as a straight line. We also choose |∆𝑧| sufficiently small so that the line z to 𝑧 + ∆𝑧
lies in R. Since 𝑓(𝑢) is continuous at 𝑧, So for any 𝑢 on this straight line and any 𝜀 > 0 There exist a 𝛿 > 0
such that

|𝑓(𝑢) − 𝑓(𝑧)| < 𝜀 whenever |𝑢 − 𝑧| < 𝛿

Which will certainly hold if |∆𝑧| < 𝛿. i.e., |𝑓(𝑢) − 𝑓(𝑧)| < 𝜀 whenever |Δ𝑧| < 𝛿

Hence


𝐹(𝑧 + ∆𝑧) − 𝐹(𝑧) 1 𝜀
− 𝑓(𝑧) = {𝑓(𝑢) − 𝑓(𝑧)}𝑑𝑢 < |∆𝑧|
∆𝑧 ∆𝑧 |∆𝑧|

=𝜀 whenever |∆𝑧| < 𝛿

𝐹(𝑧 + ∆𝑧) − 𝐹(𝑧)


lim = 𝑓(𝑧)
∆ → ∆𝑧

i.e.,

𝐹′(𝑧) = 𝑓(𝑧)

∴ 𝐹(𝑧) is differentiable at every point 𝑧 in 𝑅. Hence 𝐹(𝑧) is analytic in 𝑅 and so 𝐹′(𝑧) = 𝑓(𝑧) is analytic in 𝑅.

Maximum modulas theorem: If 𝑓(𝑧) is analytic inside and on a simple closed curve 𝐶.Then the maximum
value of |f(z)| ocuurs on C unless 𝑓(𝑧) is a constant.
34
Cauchy’s Integral Formulas and Related Theorems
Proof: Since 𝑓(𝑧) is analytic and hence continuous inside and on 𝐶, it
follows that |𝑓(𝑧)| has a maximum value M for at least one value of z inside
or on 𝐶. Suppose this maximum value s not attained on the boundary of
𝐶 but it is attained at an interior point 𝑎, is |𝑓(𝑎)| = 𝑀. Let 𝐶 be a circle
inside 𝐶 with centre at 𝑎 enclosing the point 𝑏 such that |𝑓(𝑏)| < 𝑀.

Suppose |𝑓(𝑏)| = 𝑀 − 𝜀 where 𝜀 > 0

Since |𝑓(𝑧)| is continuous at 𝑏.Then for 𝜀 > 0, there exist 𝛿 > 0


such that

𝜀 𝜀 𝜀
|𝑓(𝑧)| − |𝑓(𝑏)| < =𝑀−𝜀+ =𝑀− whenever |𝑧 − 𝑏| < 𝛿
2 2 2

Construct a circle 𝐶 inside 𝐶 with centre at 𝑏 and radius 𝛿. then for all 𝑧 inside 𝐶

𝜀
|𝑓(𝑧)| < 𝑀 −
2

Now construct a circle 𝐶 with centre at a and radius = |𝑎 − 𝑏|. Then by gauss mean value theorem

1
𝑓(𝑎) = 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃
2𝜋

1 1
= 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃 + 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃
2𝜋 2𝜋

i.e.,

|𝑓(𝑎)| ≤ ∫ 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃 + ∫ 𝑓 𝑎 + 𝑟𝑒 𝑑𝜃

i.e.,

|𝑓(𝑎)| ≤ 𝑀− 𝛼+ (2𝜋 − 𝛼)

𝜀𝛼
=𝑀−
4𝜋

i.e.,

|𝑓(𝑎)| = 𝑀 ≤ 𝑀 −

which is a contradiction. Hence the maximum value of |𝑓(𝑧)| must occur on 𝐶.


35
Cauchy’s Integral Formulas and Related Theorems
Liouvilles Theorem: If 𝑓(𝑧) is analytic and bounded in the entire comple plan then 𝑓(𝑧𝑧) must be a constant.

Proof: Let 𝑎 and 𝑏 be any two points in the complex plane. Construct a circle C with centre at 𝑎 and radius 𝑟
enclosing 𝑏. Hence 𝑓(𝑧) is analytic inside and on 𝐶. So by Cauchys integral
formula

1 1 1
𝑓(𝑏) − 𝑓(𝑎) = − 𝑓(𝑧) 𝑑𝑧
2𝜋𝑖 𝑧−𝑏 𝑧−𝑎

1 𝑏−𝑎
= 𝑓(𝑧) 𝑑𝑧
2𝜋𝑖 (𝑧
( − 𝑎)(𝑧 − 𝑏)

Choose 𝑟 large enough so that |𝑎 − 𝑏| < .

Then on 𝐶, |𝑧 − 𝑎| = 𝑟 and

𝑟 𝑟
|𝑧 − 𝑏| = |(𝑧 − 𝑎) − (𝑏 − 𝑎)| ≥ |𝑧 − 𝑎
𝑎| − |𝑏 − 𝑎| > 𝑟 − =
2 2

Also 𝑓(𝑧) is bounded, so |𝑓(𝑧)| < 𝑀 for some +ve number M.

Therefore,

1 (𝑏 − 𝑎) 1 |𝑏 − 𝑎|𝑀 2𝑀|𝑏 − 𝑎|
|𝑓(𝑏) − 𝑓(𝑎)| = 𝑓(𝑧) 𝑑𝑧 ≤ 2𝜋𝑟 =
2𝜋𝑖 (𝑧 − 𝑎)(𝑧 − 𝑏) 2𝜋 𝑟. 𝑟

i.e.,

| |
|𝑓(𝑏) − 𝑓(𝑎)| ≤ → 0 𝑎𝑠 𝑟 → ∞

Hence

𝑓(𝑏
𝑏) = 𝑓(𝑎)

i.e., 𝑓(𝑧) is constant.

Definition: A number 𝛼 is called a zero of a function 𝑓(𝑧) if 𝑓(𝛼) = 0

Example: If 𝑓(𝑧) = 𝑧 + 5𝑧 + 6 then 𝑧 = 2,3 are zero of 𝑓(𝑧).

Definition: 𝛼 is called a root of a polynomial equation 𝑓(𝑧) = 0 if 𝑓(𝛼) = 0

Fundamental theorem of algebra: Every polynomial equation 𝑃(𝑧) = 𝑎 𝑧 + 𝑎 𝑧 + ⋯ + 𝑎 = 0 here


𝑛 ≥ 1 has at least one root.
36
Cauchy’s Integral Formulas and Related Theorems

Proof: Suppose the equation has no root. Then 𝑃(𝑧) ≠ 0 for all complex number. Then 𝑓(𝑧) = is analytic
( )
on the enire complex plane. more over 𝑓(𝑧) is bounded for all 𝑧. In fact |𝑓(𝑧)| tends to zero as |𝑧| →∝.
Hence by Liouvilles theorem 𝑓(𝑧) is must be constant i.e., is constant and so P(z) is constant. But this is a
( )
contradiction as 𝑛 ≥ 1

Therefore the equation has at least one root.

Argument Theorem: If 𝑓(𝑧) is analytic inside and on a simple closed curve 𝐶 except for a pole at 𝛼 of order 𝑝
and only one zero at 𝛽 of other n inside C. Then

1 𝑓 (𝑧)
𝑑𝑧 = 𝑛 − 𝑝
2𝜋𝑖 𝑓(𝑧)

Proof: Construct two circles 𝐶 and 𝐶 with centre at 𝛼 and 𝛽 respectively so that they lie insde 𝐶 and they do
not intersect each other.

( )
Then is analytic in the multiply connected region bounded by 𝐶 ,
( )
𝐶 and 𝐶. So by cauchys Theorem for multiply connected region

1 𝑓′(𝑧) 1 𝑓′(𝑧) 1 𝑓′(𝑧)


𝑑𝑧 − 𝑑𝑧 − 𝑑𝑧 = 0
2𝜋𝑖 𝑓(𝑧) 2𝜋𝑖 𝑓(𝑧) 2𝜋𝑖 𝑓(𝑧)

( ) ( ) ( )
⇒ ∮ 𝑑𝑧 = ∮ 𝑑𝑧 + ∮ 𝑑𝑧
( ) ( ) ( )
…………………………………..….…..(i)

Now for 𝐶 , 𝑓(𝑧) has a pole at 𝑧 = 𝛼 of order 𝑝. Then

( )
𝑓(𝑧) = ……………………………(ii)
( )

where 𝐹(𝑧) is analytic inside and on 𝐶 and contain no zero inside and on 𝐶 .

Taking logarithms (ii) on both sides we get

log 𝑓(𝑧) = log 𝐹(𝑧) − 𝑝 log(𝑧 − 𝛼)

Differentiating both sides we get

𝑓′(𝑧) 𝐹′(𝑧) 𝑝
= −
𝑓(𝑧) 𝐹(𝑧) 𝑧 − 𝛼
37
Cauchy’s Integral Formulas and Related Theorems
Hence We know
( )
( ) ( ) 𝑓(𝑎) = ∮ 𝑑𝑧
∮ 𝑑𝑧 = ∮ 𝑑𝑧 − ∮ ( )
( ) ( )
Put 𝑓(𝑧) = 1, then
𝑃
= 0− . 2𝜋𝑖 1 𝑑𝑧
2𝜋𝑖 1=
2𝜋𝑖 (𝑧 − 𝑎)
= −𝑃

For 𝐶 , since 𝑓(𝑧) has a zero at 𝛽 order n. so

𝑓(𝑧) = (𝑧 − 𝑃) 𝐺(𝑧) … … … … … … … (𝑖𝑖𝑖)

where 𝐺(𝑧) is analytic, continuous, no zero inside and on 𝐶 .

Taking logarithm on both side of (iii) we get

log 𝑓(𝑧) = log 𝐺(𝑧) + 𝑛 log(𝑧 − 𝛽)

Differentiating both sides we get

𝑓′(𝑧) 𝐺′(𝑧) 𝑛
= +
𝑓(𝑧) 𝐺(𝑧) 𝑧 − 𝛽

Hence

1 f′(z) 1 G′(z) n dz
dz = dz +
2πi f(z) 2πi G(z) 2πi z−β

𝑛
=0+ 2𝜋𝑖
2𝜋𝑖

=𝑛

Therefore form (i)

1 𝑓′(𝑧)
=𝑛−𝑃
2𝜋𝑖 𝑓(𝑧)

N.B.: Similarly we can show that if 𝑓(𝑧) is analytic inside on a simple closed curve 𝐶 except a finite number of
poles at 𝛼 1,𝛼 2, . . . ,𝛼 r inside 𝐶 of order 𝑝 , 𝑝 , … 𝑝 respectively and if 𝑓(𝑧) has zero at 𝛽 1,𝛽 2, . . . ,𝛽 s inside 𝐶 of
order 𝑛 , 𝑛 , . . . 𝑛 respectively then

1 f′(z)
dz = 𝑛 − 𝑃
2πi f(z)
38
Cauchy’s Integral Formulas and Related Theorems
Corollary: If 𝑓(𝑧) is analytic inside and on a simple closed curve 𝐶 except a pole at 𝛼 inside 𝐶 of order 𝑝 and if
𝑓(𝑧) has zero at 𝛽 inside 𝐶 of order 𝑛 then for any analytic function 𝑔(𝑧) inside and on 𝐶

1 f (z)
g(z) dz = n g − p g(α)
2πi f(z)

Proof: Constant two circles 𝐶 & 𝐶 inside C with centre at 𝛼 & 𝛽 repectively so that they do not intersect each
other.
( )
Then 𝑔(𝑧) is analytic in the multiple connected region bounded by C, C 1&C2 . Then by Cauchys theorem
( )
for multiple connected region we have

( ) ( ) ( )
∮ g(z) dz = ∮ g(z) dz + ∮ g(z) dz ………………(i)
( ) ( ) ( )

Since 𝑓(𝑧) has a pole at 𝑧 = 𝛼 of order 𝑝 then

( )
𝑓(𝑧) = where 𝐹(𝑧) is analytic inside and on 𝐶
( )

Taking log on both sides we have

log 𝑓(𝑧) = log 𝑓(𝑧) − 𝑝 log(𝑧 − 𝛼)

Differentiating we get

𝑓′(𝑧) 𝐹′(𝑧)
= − 𝑝⁄(𝑧 − 𝛼)
𝑓(𝑧) 𝐹(𝑧)

( ) ( ) ( )
⇒ 𝑔(𝑧) ( )
= 𝑔(𝑧) ( )

𝑓 (𝑧) 𝐹 (𝑧) 𝑔(𝑧)𝑝


𝑔(𝑧) 𝑑𝑧 = 𝑔(𝑧) 𝑑𝑧 − 𝑑𝑧
𝑓(𝑧) 𝐹(𝑧) (𝑧 − 𝛼)

𝑔(𝑧)
= 0−𝑝 𝑑𝑧 = −𝑝 2𝜋𝑖 𝑔( 𝛼)
(𝑧 − 𝛼)

= 2𝜋𝑖(−𝑔(𝛼)𝑃)

( )
Similarly we can show that ∮ 𝑔(𝑧) ( )
𝑑𝑧 = 2𝜋𝑖(−𝑔(𝛽)𝑛)

∴ From (i)
39
Cauchy’s Integral Formulas and Related Theorems

f (z)
g(z) dz = 2πi{(g(β)n + p(−g(α)p)}
f(z)

1 f (z)
g(z) dz = ng(β) − pg(α)
2πi f(z)

[Proved]

Rouchis Theorem: If 𝑓(𝑧) and 𝑔(𝑧) are analytic inside and on a simple closed curve 𝐶 and if |𝑔(𝑧)| < |𝑓(𝑧)|
on 𝐶. Then 𝑓(𝑧) and 𝑓(𝑧) + 𝑔(𝑧) haave same members of zeros inside 𝐶.

( )
Proof: Let 𝐹(𝑧) =
( )

So 𝑔 = 𝐹𝑓 ⇒ 𝑔′ = 𝐹𝑓′ + 𝐹′𝑓

( )
Here |𝐹(𝑧)| = <1 on 𝐶 𝑎𝑠 |𝑔(𝑧)| < |𝑓(𝑧)| 𝑜𝑛 𝐶
( )

Now we get |𝑓(𝑧)| ≠ 0 𝑜𝑛 𝐶

Hence 𝐹(𝑧) is analytic on 𝐶

Let 𝑁 and 𝑁 be the number of zeros of 𝑓 and 𝑓 + 𝑔 respectively inside 𝐶

Then by argument theorem

1 f + g′ 1 f′
𝑁 −𝑁 = dz − dz
2πi f+g 2πi f

1 f + Ff + F′f 1 f′
= dz − dz
2πi f + fF 2πi f

1 f (1 + f) + F′ f′
= − dz
2πi f(1 + F) f

1 f F′ f′
= + − dz
2πi f 1+F f

1 F′
= dz
2πi 1+F
40
Cauchy’s Integral Formulas and Related Theorems

Since 𝐹(𝑧) is analytic on 𝐶 and |𝐹(𝑧)| < 1 so 1 + 𝐹 ≠ 0 on 𝐶. Hence is also analytic on 𝐶. Therefore by
cauchys theorem

F′
dz = 0
1+F

Hence 𝑁 − 𝑁 = 0 i. e., 𝑁 = 𝑁

Therefore 𝑓(𝑧) and 𝑓(𝑧) + 𝑔(𝑧) have some same number of zeros inside 𝐶.

Ex. A polynomial equation 𝑎 𝑧 + 𝑎 𝑧 + ⋯ . +𝑎 𝑧 + 𝑎 = 0 of degree 𝑛 ≥ 1 has exactly 𝑛 roots

Proof: Let 𝑓(𝑧) = 𝑎 𝑧 and 𝑔(𝑧) = 𝑎 𝑧 + ⋯ . +𝑎 𝑧 + 𝑎 = 0

Consider a circle 𝐶 with centre at origin and radius 𝑟 > 1

Here both 𝑓(𝑧) and 𝑔(𝑧) are analytic inside and on 𝐶

Now on 𝐶

|g(z)| |a z +a z + ⋯ + a z + a|
=
|f(z)| |a z |

| | ⋯ | || | | |
≤ | || |

|a ||r | + ⋯ + |a ||r| + a
=
|a ||r |

|𝑎 |𝑟 + ⋯ + |𝑎 |𝑟 + |𝑎 |𝑟
≤ |∵𝑟<1
|𝑎 |𝑟

|𝑎 | + |𝑎 | + ⋯ + |𝑎 |
=
|𝑎 |𝑟

| ( )|
choosing 𝑟 large enough we can make | <1
( )|

Then by Rouchis theorem 𝑓(𝑧) and 𝑓(𝑧) + 𝑔(𝑧) have some number of zeros inside 𝐶. But 𝑓(𝑧) has 𝑛 mumber
of zero inside 𝐶 all located at origin. Therefore 𝑓(𝑧) + 𝑔(𝑧) has 𝑛 number of zero inside 𝐶.

i.e., 𝑎 𝑧 + 𝑎 𝑧 + ⋯ . +𝑎 𝑧 + 𝑎 = 0 has exact𝑙𝑦 n number of roots.


41
Cauchy’s Integral Formulas and Related Theorems

Ex. Evaluate ∮ 𝑑𝑧 where 𝐶 is


²

(a) The circle |𝑧| = 2


(b) The rectangle with vertices – 𝑖, 2 − 𝑖, 2 + 𝑖, 𝑖

Solution: We have =( )(
= −
² ) ( ) ( )

So ∮ 𝑑𝑧 = ∮ 𝑑𝑧 − ∮ 𝑑𝑧
²

(a) Let 𝑓(𝑧) = 𝑐𝑜𝑠 𝜋𝑧

Therefore

𝑓(1) = 𝑐𝑜𝑠 𝜋 = −1

And

𝑓(−1) = 𝑐𝑜𝑠(−𝜋) = −1

Hence by Cauchys integral formula

1 𝑓(𝑧) 1 𝑐𝑜𝑠𝜋𝑧
(1) =
−1 = 𝑓(1 𝑑𝑧 = 𝑑𝑧
2𝜋𝑖 𝑧−1 2𝜋𝑖 𝑧−1

1 𝑐𝑜𝑠𝜋𝑧
⇒ 𝑑𝑧 = −𝜋𝑖
2 𝑧−1

Similarly

((−1) = 𝑓(−1) = ∮ 𝑑𝑧

1 𝑐𝑜𝑠𝜋
⇒ 𝑑𝑧 = −𝜋𝑖
2 𝑧

Hence

𝑐𝑜𝑠𝜋𝑧
𝑑𝑧 = −𝜋𝑖 − (𝜋𝑖) = 0
𝑧 −1
42
Cauchy’s Integral Formulas and Related Theorems
(b) Since 𝑧 = −1 lies outside 𝐶, So
1 𝑐𝑜𝑠𝜋𝑧
𝑑𝑧 = 0
2 𝑧+1

But 𝑧 = 1 lies inside C


Let 𝑓(𝑧) = 𝑐𝑜𝑠𝜋𝑧, Since 𝑓(𝑧) is analytic inside and on 𝐶

So by Cauchys integral formula

1 𝑐𝑜𝑠𝜋𝑧
−1 = 𝑓(1) = 𝑑𝑧
2𝜋𝑖 𝑧−1

1 𝑐𝑜𝑠𝜋𝑧
⇒ 𝑑𝑧 = −𝜋𝑖
2 𝑧−1

Hence ∮ 𝑑𝑧 = 𝜋𝑖 − 0 = −𝜋𝑖

Ex. Find the value of ∮ 𝑑𝑧 where 𝐶 is the circle |𝑧| = 1.


( )

Solution : Let 𝑓(𝑧) = 𝑠𝑖𝑛 𝑧

Here 𝑓(𝑧) is analytic inside and on 𝐶. So by C


Cauchys integral formula

𝜋 2! 𝑠𝑖𝑛 𝑧
𝑓 ՜՜( ) = 𝑑𝑧
6 2𝜋𝑖 (𝑧 − )

Now,

𝑓(𝑧) = 𝑠𝑖𝑛 𝑧

𝑓՜(𝑧) = 6𝑠𝑖𝑛 𝑧 𝑐𝑜𝑠 𝑧

𝑓՜՜(𝑧) = 30 𝑠𝑖𝑛 𝑧𝑐𝑜𝑠 𝑧 − 6𝑠𝑖𝑛 𝑧

𝜋 1 √3 1
∴𝑓 = 30. . −6
6 2 2 2

= 30. . −
43
Cauchy’s Integral Formulas and Related Theorems

= − =

Hence

1 𝑠𝑖𝑛 𝑧 21
𝑑𝑧 =
𝜋𝑖 (𝑧 − ) 16

i.e., ∮ 𝑑𝑧 =
( )

Ex. Evaluate ∮ 𝑑𝑧 where 𝜖 > 0 and 𝐶 is the circle |𝑧|=3.


( )

Solution: We have ∮ 𝑑𝑧
( )

= ∮ ( ) (
𝑑𝑧
)

Here 𝑖 & − 𝑖 lie inside 𝐶. Construct two circle C1 & C2 inside C with centre at 𝑖 & − 𝑖 respectively so that they
do not overlap each other. Then by Cauchys
auchys theorem

1 𝑒
2𝜋𝑖 (𝑧 + 𝑖) (𝑧 − 𝑖)
1 𝑒 𝑑𝑧 1 𝑒
= + 𝑑𝑧
2𝜋𝑖 (𝑧 + 𝑖) (𝑧 − 𝑖)) 2𝜋𝑖 (𝑧 + 𝑖) (𝑧 − 𝑖)

=I1+I2 Say

For I, let 𝑓(𝑧) =


( )

Here 𝑓(𝑧) is analytic inside and on C1 and 𝑧 = 𝑖 is inside C.

So by Cauchys integral formula,

1 𝑒
𝑓((𝑖) = 𝑑𝑧
2𝜋𝑖 (𝑧 + 𝑖) (𝑧 − 𝑖)

Here

(𝑧 + 𝑖) + 𝑒 + 𝑒 2(𝑧 + 𝑖)
(𝑧) =
𝑓(𝑧
(𝑧 + 𝑖)

−4𝑡𝑒 − 𝑒 4𝑖 −𝑡𝑒 − 𝑖𝑒
∴ 𝑓((𝑖) = =
16 4
44
Cauchy’s Integral Formulas and Related Theorems

For I2, Let 𝑔(𝑧) =


( )

Here 𝑔(𝑧) is analytic inside and on C2.

Also 𝑧 = −𝑖 lies inside 𝐶2, So by Cauchys integral formula

1 𝑒
𝑔(−𝑖) = 𝑑𝑧
2𝜋𝑖 (𝑧 + 𝑖) (𝑧 − 𝑖)

Here

(𝑧 − 𝑖) + 𝑒 − 𝑒 (𝑧 − 𝑖)
𝑔(𝑧) =
(𝑧 + 𝑖)

Hence 𝑔(−𝑖) = =

Therefore

∮ =I1+I2
( )

= 𝑓(𝑖) + 𝑔(−𝑖)

−𝑡 𝑒 + 𝑒 − 𝑖(𝑒 − 𝑒 )
=
4
−2𝑡𝑐𝑜𝑠𝑡 + 2𝑠𝑖𝑛𝑡
=
4
1
= (𝑠𝑖𝑛𝑡 − 𝑡 𝑐𝑜𝑠𝑡)
2

( )( )
Ex. Let 𝑓(𝑧) = ( ) ( )

Evaluate

( ) ( )
(i) ∮ 𝑑𝑧 and (ii) ∮ 𝑑𝑧
( ) ( )

Where 𝐶 is the circle |𝑧| = 4.

Solution:

(1) put 𝑧 + 1 = 0 the 𝑧 = ±𝑖


45
Cauchy’s Integral Formulas and Related Theorems
𝑧 − 3 = 0 then 𝑧 = 3

𝑧 + 5 = 0 then 𝑧 = −5

So f(z) has zero at 𝑧 = 𝑖 of order 2

at 𝑧 = −𝑖 of order 2

at 𝑧 = 3 of order 1

at 𝑧 = −5 of order 3

But only i,-i,3 lie inside C

Hence the number of zero inside 𝐶 = 2 + 2 + 1 = 5

Again put 𝑧 + 2𝑧 + 2 = 0

Then 𝑧 = −1 ± 𝑖

Put 𝑧 − 6 = 0 then 𝑧 = 6

So 𝑓(𝑧) has pole at 𝑧 = −1 + 𝑖 of order 3

at 𝑧 = −1 − 𝑖 order 3

at 𝑧 = 6 of order 1

But −1 + 𝑖 and −1 − 𝑖 lie inside 𝐶

Therefore the member of poles inside 𝐶 = 3 + 3 = 6

Hence by argument theorem

՜( )
∮ 𝑑𝑧 = 5 − 6 = −1
( )

(Ans)

(ii) Put 𝑧 + 1 = 0 then z=±𝑖

𝑧 − 3 = 0 then 𝑧 = 3

𝑧 + 5 = 0 then 𝑧 = −5

So 𝑓(𝑧) has zero at 𝑧 = 𝑖 of order 2

at 𝑧 = −𝑖 of order 2
46
Cauchy’s Integral Formulas and Related Theorems
at 𝑧 = 3 of order 1

at 𝑧 = −5 of order 3

But only 𝑖, −𝑖, 3 lie inside 𝐶

Again put 𝑧 + 2𝑧 + 2 = 0 ⇒ 𝑧 = −1 ± 𝑖

𝑧−6=0 ⇒𝑧 =6

So 𝑓(𝑧) has zero at 𝑧 = −1 + 𝑖 of order 3

at 𝑧 = −1 − 𝑖 of order 3

at 𝑧 = 6 of order 1

But only −1 + 𝑖 and −1 − 𝑖 lie inside 𝐶

Now put 𝑔(𝑧) = 𝑧 + 𝑧 Here 𝑔(𝑧) is analytic inside and on 𝐶. therefore by argument theorem

1 𝑓՜(𝑧)
(𝑧 + 2) 𝑑𝑧 = 2𝑔(𝑖) + 2𝑔(−𝑖) + 1. 𝑔(2) − 3𝑔(−1 + 𝑖) − 3𝑔(−1 − 𝑖)
2𝜋𝑖 𝑓(𝑧)

= 2(−1 + 𝑖) + 2(−1 − 𝑖) + 12 − 3[−𝑖 − 1) − 3(𝑖 − 1)

= −4 + 12 + 3 + 3

=14

Ex. Prove that all roots of 𝑧 − 5𝑧 + 12 = 0 lie between the circle |𝑧| = 1 and |𝑧| = 2.

Proof: Let 𝑓(𝑧) = 𝑧7 and 𝑔(𝑧) = −5𝑧 + 12

Let C1 represent the circle |𝑧| = 1

And C2 represent the circle |𝑧| = 2

Now on C2 |𝑔(𝑧)| = |−5 − 𝑧 + 12| ≤ 5|𝑧| + 12 = 5. 2 + 12

= 52 < 2 = |𝑓(𝑧)|

Since both 𝑓(𝑧) and 𝑔(𝑧) are analytic inside and on 𝐶 2. So by Rouchis theorem 𝑓(𝑧) and 𝑓(𝑧) + 𝑔(𝑧) have
some number of inside C2.Since 𝑓(𝑧) has & zeroes inside C2 (all located origin)

So 𝑓(𝑧) + 𝑔(𝑧) = 𝑧7 − 5𝑧3 + 12 has 7 zeroes inside 𝐶 2

For C1 let 𝑓(𝑧) = 12 and 𝑔(𝑧) = 𝑧 − 5𝑧


47
Cauchy’s Integral Formulas and Related Theorems
Now |𝑔(𝑧)| ≤ |𝑧| + 5|𝑧| = 1 + 5 = 6 < 12 = |𝑓(𝑧)|

Also both 𝑓(𝑧) and 𝑔(𝑧) are analytic inside and on C1 So by Rouchis theorem

𝑓(𝑧) + 𝑔(𝑧) have some number of zeroes inside C1.But 𝑓(𝑧) has no zero inside 𝐶 1. Therefore 𝑓(𝑧) + 𝑔(𝑧) =
𝑧 = 5𝑧 + 12 has no zero onside C1. i.e., all zero of 𝑧 − 5𝑧 + 12 is outside of C1

∴All zeroes of 𝑧 − 5𝑧 + 12 lie between the circle |𝑧| = 1 & |𝑧| = 2.


48
Taylor’s Theorem
Taylor Theorem: If 𝑓(𝑧) is analytic inside a circle C with centre at a then for any z inside C

(𝑧 − 𝑎)
𝑓 (𝑧 ) = 𝑓 (𝑎 ) + (𝑧 − 𝑎 ) 𝑓 (𝑎 ) + 𝑓 (𝑎 ) + ⋯ … … … … … … … … … …
2!
Proof: Construct a circle C, inside C with Centre at a inclosing 𝑧. Then 𝑓 (𝑧) is analytic inside and on 𝐶.
Then by Cauchy’s integral formula,

1 𝑓(𝑤)
𝑓(z) = ∮ 𝑑𝑤
2𝜋𝑖 𝑤 − 𝑧
Now

=( ) (
= = (1 − )-1
) ( ){ }

Since on 𝐶, | |<1

So

= {1 + + + ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ } ……………………………………….……….. (1)

= [1 + + + ⋯⋯+ + 1+ + +⋯⋯ ]

( ) ( )
= + + + ⋯⋯⋯+ ( )
+ ⋯⋯⋯+ [By equation (1)]
( ) ( )

So

( ) ( ) ( ) ( ) ( ) ( ) ( ) ( )
= + + + ⋯⋯⋯+ ( )
𝑓(𝑤) +
( ) ( )

Or
( )
, ∮ 𝑑𝑤

( ) ( ) ( ) ( ) ( ) ( ) ( )
= ∮ 𝑑𝑤 + ∮ 𝑑𝑤 + ∮ 𝑑𝑤 + ⋯ ⋯ + ∮ 𝑑𝑤 + 𝑢
( ) ( ) ( )

( ) ( )
Where 𝑢 = ∮ 𝑑𝑤
( )

Now using Cauchy’s integral formula we get,


( ) ( )
𝑓(𝑧) = 𝑓(𝑎) + (𝑧 − 𝑎)𝑓′(𝑎) + 𝑓 (𝑎) + ⋯ ⋯ + ( )!
𝑓n-1(a)+𝑢
!

! ( )
(Since, f n (a)= ∮ 𝑑𝑤 )
( )

We need to show that,

lim → 𝑢 =0
49
Taylor’s Theorem

On𝑐 let =𝛽<1

|𝑤 − 𝑧| = |(𝑤 − 𝑎) − (𝑧 − 𝑎)| > |𝑤 − 𝑎| − |𝑧 − 𝑎| = 𝑟 − |𝑧 − 𝑎|

Where 𝑟 is the radius of𝑐

∴ < | |

Since 𝑓(𝑤) is analytic inside and on𝑐 . So|𝑓(𝑤)| ≤ 𝑀 for some positive number M.
( )
Hence |𝑢 | = ∮ ( ) 𝑑𝑤 < 2𝜋𝑟
| |

→0 as n→ ∞ | since β<1

( )
Therefore 𝑓(𝑧) = 𝑓(𝑎) + (𝑧 − 𝑎)𝑓′(𝑎) + 𝑓 (𝑎 ) + ⋯ ⋯ ⋯ ⋯
!

Laurent’s Theorem: If f (z) is analytic inside and on the boundary of the ring shaped region R bounded by
two concentric circle 𝑐 and 𝑐 with Centre at a and radius 𝑟 and 𝑟 respectively. Then for all z in R,

𝑎
𝑓 (𝑧) = 𝑎 ( ) +
(𝑧 − 𝑎)

( )
Where 𝑎 = ∮ ( )
𝑑𝑤

( )
And 𝑎 = ∮ 𝑑𝑤
( )

Proof: Since z is inside R, so by Cauchy’s


integral formula for multiple connected region

( ) ( )
𝑓 (𝑧) = ∮ 𝑑𝑤 − ∮ 𝑑𝑤 − − − − − − − (1)

For the first integral w is on 𝑐 .

Then <1

So =( ) (
= (1 − )
)

∴ = 1+ + + ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯(1)

= [1 + +⋯⋯+ ) + 1+ + ⋯⋯
50
Taylor’s Theorem
( ) ( )
= + + + ⋯⋯+ ( )
+( ) [ by (1) ]
( ) ( )

( ) ( ) ( ) ( ) ( )
Hence ∮ 𝑑𝑤 = ∮ 𝑑𝑤 + ∮ 𝑑𝑤 + ⋯ ⋯ + ∮ 𝑑𝑤 + 𝑢
( ) ( )

( )
Where 𝑢 = ∮ ( ) 𝑑𝑤

( )
Thus ∮ 𝑑𝑤 = 𝑎 + 𝑎 (𝑧 − 𝑎) + ⋯ ⋯ + 𝑎 (𝑧 − 𝑎 ) +𝑢

( )
Where 𝑎 = ∮ 𝑑𝑤
( )

For the second integral when w is on 𝑐 .

Then <1

Hence − = =( ) ( )
= (1 − )

= [1 + +( ) + ⋯⋯⋯]

= [1 + +( ) +⋯⋯+( ) +( ) {1 + + ) +⋯⋯ ]

( )
Or,− = +( )
+⋯⋯+ ( )
+( )

( )
So, − ∮ 𝑑𝑤 = ∮ 𝑓(𝑤)𝑑𝑤 + ∮ 𝑓(𝑤)(𝑤 − 𝑎)𝑑𝑤 + ⋯
( ) ( )

⋯+
( )
∮ (𝑤 − 𝑎 ) 𝑓 (𝑤)𝑑𝑤 + 𝑉

( )
Where, 𝑉 = ∮ ( ) 𝑑𝑤

( )
Thus − ∮ 𝑑𝑤 = + + ⋯⋯+ ( )
+𝑉
( )

( )
Where 𝑎 = ∮ 𝑑𝑤 𝑛 = 1,2,3, … … …
( )

Hence by (1),
𝑎 𝑎 𝑎
𝑓(𝑧) = 𝑎 + 𝑎 (𝑧 − 𝑎) + ⋯ ⋯ + 𝑎 (𝑧 − 𝑎) + + + ⋯⋯+ + 𝑢 + 𝑉 ⋯ (2)
𝑧 − 𝑎 (𝑧 − 𝑎) (𝑧 − 𝑎)

Where w is on 𝑐 , let =𝛽<1

Since 𝑓(𝑧) is analytic inside and on the region bounded by 𝑐 and 𝑐 so |𝑓(𝑧)| <M for some positive
number.

Also on 𝑐 |𝑤 − 𝑧| = |(𝑤 − 𝑎) − (𝑧 − 𝑎)|


51
Taylor’s Theorem
≥|𝑤 − 𝑎| − |𝑧 − 𝑎|

= 𝑟 − |𝑧 − 𝑎|

So ≤
| |

( ) ( )
Therefore | ∮ ( )
𝑑𝑤| ≤ | |
2𝜋𝑟

→0 𝑎𝑠 𝑛→∞ |𝑠𝑖𝑛𝑐𝑒 𝛽 < 1

islim → 𝑢 =0

When w is on 𝑐 , let =𝛾<1

on 𝑐 , |𝑧 − 𝑤| = |(𝑧 − 𝑎) − (𝑤 − 𝑎)|

≥ |𝑧 − 𝑎| − |𝑤 − 𝑎|

= |𝑧 − 𝑎| − 𝑟

is| |≤| |

( )
Hence | ∮ ( ) 𝑑𝑤| ≤ | |
2𝜋𝑟

→0 𝑎𝑠 𝑛→∞ |since 𝛾 < 1

Is lim → 𝑉 =0

Hence from (2)


𝑎 𝑎
𝑓 (𝑧) = 𝑎 + 𝑎 (𝑧 − 𝑎) + 𝑎 (𝑧 − 𝑎) + ⋯ ⋯ + + + ⋯⋯⋯
𝑧 − 𝑎 (𝑧 − 𝑎)

Is 𝑓 (𝑧) = ∑ 𝑎 (𝑧 − 𝑎) + ∑
( )

Note: we can also write the Laurent’s expression as

𝑓 (𝑧 ) = 𝑎 (𝑧 − 𝑎)

( )
Where 𝑎 = ∮ 𝑑𝑤
( )

And c is any simple closed curve, the region enclosing the point 𝑧 = 𝑎

֍ If Laurent’s expansion ∑ 𝑎 (𝑧 − 𝑎) is called analytic part and ∑ is called principal part.


( )

֍If 𝑓(𝑧) has no singularity inside and on 𝑐 (is 𝑓(𝑧) is analytic) then the series doesn’t possess any principal
part. Thus the Laurent’s explanation concide with Taylor’s explanation,is the principal part occurs only
because of singularity of function.
52
Taylor’s Theorem
Note: In Laurent’s expansion of a function 𝑓(𝑧) , if number of terms in the principal part is finite say n, then
the singularity is pole of order n.

If the number of terms is infinite then the singularity is an essential singularity.

If the principal part vanishes then the singularity is called a removable singularity.

Ex. Find Laurent’s series about the indicated singularity. For each function name the singularity in each case
and give the region of convergence of each series.

(a) 𝑓(𝑧) = ; 𝑧=2


( )

(b) 𝑓(𝑧) = (𝑧 − 4)𝑠𝑖𝑛 ; 𝑧 = −3


(c) 𝑓(𝑧) = ; 𝑧=0

Solution: (a) put 𝑧 − 2 = 𝑢, then 𝑧 = 𝑢 + 2

𝑒 𝑒 ( )
𝑒
∴ 𝑓 (𝑧 ) = = = 𝑒
(𝑧 − 2) 𝑢 𝑢

𝑒 (3𝑢 ) (3𝑢)
= [1 + 3𝑢 + + + ⋯⋯]
𝑢 2! 3!
1 3 9 1 9 1 27 81
=𝑒 [ + + + + + 𝑢 + ⋯⋯]
𝑢 𝑢 2𝑢 2𝑢 8 40
1 3 9 1 9 1 27 81
=𝑒 [ + + + + + (𝑧 − 2) + ⋯ ⋯ ⋯ ]
(𝑧 − 2) (𝑧 − 2) 2 (𝑧 − 2 ) 2 𝑧 − 2 8 40

Thus is the Laurent’s series of the function around 𝑧 = 2

Here the principal part consists of only 4 terms. So 𝑧 = 2 is a pole of order 4.

The series is convergent only if


( )
1 + 3𝑢 + + ⋯ ⋯ ⋯ is convergent, provided 𝑢 ≠ 0
!

( )
Here 𝑎 = ( )!

(3𝑢)
𝑎 =
𝑛!
𝑎 (3𝑢) (𝑛 − 1)! 3
∴ = × = |𝑢|
𝑎 𝑛! ( 3𝑢 ) 𝑛
3
Now lim → | | = lim 𝑛 |𝑢| = 0 < 1

Hence by ratio test the series is convergent for all 𝑢 ≠ 0

That is the series is convergent for all 𝑧 ≠ 2


53
Taylor’s Theorem

(b) Here 𝑓 (𝑧) = (𝑧 − 4)𝑠𝑖𝑛 ; 𝑧 = −3

Put 𝑧 ≠ 3 = 𝑢 Then 𝑧=𝑢−3

1
∴ 𝑓(𝑧) = (𝑢 − 7)𝑠𝑖𝑛
𝑢
1 1 1
= (𝑢 − 7 ){ − + −⋯⋯⋯}
𝑢 3! 𝑢 5! 𝑢
7 1 7 1 7
=1− − + + − − ⋯⋯⋯
𝑢 3! 𝑢 3! 𝑢 5! 𝑢 5! 𝑢
7 1 7 1 7
=1− − + + − ⋯⋯⋯
𝑧 + 3 3! (𝑧 + 3) 3! (𝑧 + 3) 5! (𝑧 + 3) 5! (𝑧 + 3)

Which is the required Laurent’s series around 𝑧 = −3

Since the number of terms of the principal part in the expansion is infinite, so 𝑧 = −3 is an essential
singularity.

This series is convergent only if

− + − ⋯⋯⋯ is convergent
! !

(−1)
𝑎 =
(2𝑛 − 1)! 𝑢

(−1)
𝑎 =
(2𝑛 + 1)! 𝑢

𝑎 (−1) (2𝑛 − 1)! 𝑢 ( )


1
∴ = × =
𝑎 (2𝑛 + 1)! 𝑢 (−1) 2𝑛(2𝑛 + 1)|𝑢|

Now lim → = lim ( )| |


= 0 < 1 provided 𝑢 ≠ 0

Hence by ratio test the series is convergent for all 𝑢 ≠ 0.

Is the Laurent series is convergent for all 𝑧 ≠ −3.

(c)Given that 𝑓(𝑧) = ; 𝑧=0

We get = {𝑧 − 𝑧 − + − ⋯⋯⋯ }
! !

= ( − + − ⋯⋯⋯)
! ! !

= − + − ⋯⋯⋯
! ! !

Which is the Laurent series for the function around 𝑧 = 0.


54
Taylor’s Theorem
Since the series has no principal part so 𝑧 = 0 is a removable singularity.

Here 𝑎 = (−1) ( )!

𝑧
𝑎 = (−1)
(2𝑛 + 3)!

𝑎 |𝑧|
∴ =
𝑎 (2𝑛 + 3)(2𝑛 + 2)
𝑎
lim =0<1
→ 𝑎

So by ratio test the series converges for all z.

Ex. Expand 𝑓 (𝑧) = in a Laurent series valid for

(a) 2 < |𝑧| < 3


(b) |𝑧| > 3
(c) 1 < |𝑧 + 1| < 2
(d) |𝑧| < 2

Solution: Here 𝑓 (𝑧) = = = −


( )( )

(a) Given condition is 2 < |𝑧| < 3

When 2 < |𝑧|

2 2 2 2
= = (1 + )
𝑧 + 2 𝑧(1 + ) 𝑧 𝑧

2 2 4 8
= (1 − + − + ⋯ ⋯ ⋯ )
𝑧 𝑧 𝑧 𝑧
2 4 8 16
= − + − + ⋯⋯⋯
𝑧 𝑧 𝑧 𝑧
When |𝑧| < 3

3 3 𝑧
= = (1 + )
𝑧 + 3 3(1 + ) 3

𝑧 𝑧 𝑧 𝑧
= 1− + − + − ⋯⋯⋯
3 9 27 81
Hence the Laurent series is

𝑧 𝑧 𝑧 2 4 8 16
1− + − +⋯⋯⋯ − − + − + ⋯⋯⋯
3 9 27 𝑧 𝑧 𝑧 𝑧
55
Taylor’s Theorem

=⋯ ⋯ ⋯ + − + − +1− + − + ⋯⋯⋯

(b) Given region is |𝑧| > 3

Now = = (1 + )-1
( )

2 2 4 8
= 1 − + − + ⋯⋯⋯
𝑧 𝑧 𝑧 𝑧
2 4 8 16
= − + − + ⋯⋯⋯
𝑧 𝑧 𝑧 𝑧
3 3 3 3
= = (1 + )
𝑧 + 3 𝑧(1 + ) 𝑧 𝑧

3 3 9 27
= 1− + − ⋯⋯
𝑧 𝑧 𝑧 𝑧

= − + − + ⋯⋯⋯

So the Laurent series is

3 9 27 81 2 4 8 16
− + − + ⋯⋯⋯− − + − + ⋯⋯⋯
𝑧 𝑧 𝑧 𝑧 𝑧 𝑧 𝑧 𝑧
1 5 19 65
= − + − + ⋯⋯⋯
𝑧 𝑧 𝑧 𝑧
(c) Given region is 1 < |𝑧 + 1| < 2

Put 𝑧 + 1 = 𝑢 Then 1 < |𝑢| < 2

3 2 3 2
∴ 𝑓 (𝑧 ) = − = −
𝑧+3 𝑧+2 𝑢+2 𝑢+1
When 1 < |𝑢|

Then = 1+

= 1− + − +⋯⋯⋯

= − + − + ⋯⋯⋯

When |𝑢| < 2; Then

3 3 𝑢
= 1+
𝑢+2 2 2

= 1− + − + ⋯⋯⋯
56
Taylor’s Theorem

= − + − + ⋯⋯⋯

Hence the series is

3 3𝑢 3𝑢 3𝑢 2 2 2 2
− + − + ⋯⋯⋯ − − + − + ⋯⋯⋯
2 𝑢 8 16 𝑢 𝑢 𝑢 𝑢

2 2 2 2 3 3𝑢 3𝑢 3𝑢
= ⋯⋯⋯+ − + − + − + − + ⋯⋯⋯
𝑢 𝑢 𝑢 𝑢 2 4 8 16
2 2 2 2 3 3 (𝑧 + 1 ) 3 (𝑧 + 1 ) 3(𝑧 + 1)
= ⋯⋯⋯+ − + − + − + − +⋯⋯⋯
(𝑧 + 1) (𝑧 + 1 ) (𝑧 + 1) (𝑧 + 1) 2 4 8 16

(d) Given region is |𝑧| < 2

Now = =1− + − + ⋯⋯⋯

2 2 𝑧 𝑧 𝑧
= =1− + − + ⋯⋯⋯
𝑧+2 2 1+ 2 4 8

Hence the series is

𝑧 𝑧 𝑧 𝑧 𝑧 𝑧
1− + − + ⋯⋯⋯− 1 − + − + ⋯⋯⋯
3 9 27 2 4 8

𝑧 5 19
= − 𝑧 + 𝑧 −⋯⋯⋯⋯
6 36 196

Ex. Show that cosh 𝑧 + =𝑎 +∑ 𝑎 𝑧 +

Where 𝑎 = ∫ cosh(2𝑐𝑜𝑠𝜃) 𝑐𝑜𝑠𝑛𝜃 𝑑𝜃

Solution: Here cosh 𝑧 + is

analytic for every finite value of z except at 𝑧 = 0

So it is analytic in the ring shaped region 𝑟 < |𝑧| < 𝑅 where r is


sufficiently

small and R is sufficiently large. So by Laurent’s theorem cosh 𝑧 + =∑ 𝑎 𝑧 +∑

Where 𝑎 = ∮ 𝑑𝑧 , 𝑎 = ∮ 𝑑𝑧

Where c is a simple closed curve in the annulus (ring shaped region) enclosing the region.

Now choose c as the unit circle with centre at origin.

Then on c 𝑧=𝑒
57
Taylor’s Theorem
𝑑𝑧 = 𝑖𝑒 𝑑𝜃

𝑧+ =𝑒 +𝑒 = 2 cos 𝜃

( )
So 𝑎 = ∫ ( ) 𝑖𝑒 𝑑𝜃

1
= cos h (2𝑐𝑜𝑠𝜃)𝑒 𝑑𝜃
2𝜋

1 𝑖
= cos h(2𝑐𝑜𝑠𝜃) 𝑐𝑜𝑠𝑛𝜃 𝑑𝜃 − cos h(2𝑐𝑜𝑠𝜃) 𝑠𝑖𝑛 𝑛𝜃 𝑑𝜃
2𝜋 2𝜋

Now ∫ cos h(2𝑐𝑜𝑠𝜃) 𝑠𝑖𝑛 𝑛𝜃 𝑑𝜃

=∫ cos h(2 cos(2𝜋 − 𝜃 )) 𝑠𝑖𝑛 𝑛(2𝜋 − 𝜃) 𝑑𝜃

=∫ cos h (2𝑐𝑜𝑠𝜃)(− sin 𝑛𝜃) 𝑑𝜃

= −∫ cos ℎ(2 cos 𝜃) sin 𝑛𝜃 𝑑𝜃

∴ cos ℎ(2 cos 𝜃) sin 𝑛𝜃 𝑑𝜃 = 0

So 𝑎 = ∫ cos ℎ(2 cos 𝜃) cos 𝑛𝜃 𝑑𝜃

Also 𝑎 = ∫ cos ℎ(2 cos 𝜃) cos(−𝑛𝜃) 𝑑𝜃

= ∫ cos ℎ(2 cos 𝜃) cos 𝑛𝜃 𝑑𝜃

=𝑎 for 𝑛 = 1,2,3, ⋯ ⋯ ⋯

Hence cos ℎ 𝑧 + =𝑎 +∑ 𝑎 𝑧 +∑

=𝑎 +∑ 𝑎 𝑧 +

(Proved)

You might also like