Thanks to visit codestin.com
Credit goes to www.scribd.com

0% found this document useful (0 votes)
60 views5 pages

Complex Analysis Spring 2023 HW 8

The document provides a proof regarding contour integration of a complex function. It defines the function, contour, and parametrizations. It then evaluates the integrals over each contour segment and shows their limits vanish as the contours expand and shrink. The conclusion is the integral equals the residue at the pole times 2πi.

Uploaded by

asampath
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)
60 views5 pages

Complex Analysis Spring 2023 HW 8

The document provides a proof regarding contour integration of a complex function. It defines the function, contour, and parametrizations. It then evaluates the integrals over each contour segment and shows their limits vanish as the contours expand and shrink. The conclusion is the integral equals the residue at the pole times 2πi.

Uploaded by

asampath
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/ 5

COMPLEX ANALYSIS HOMEWORK 7

ATHARV SAMPATH

Exercise 10
Proof. First, let us find the residue at 𝑖𝑎 and then integrate over the given contour. Let
𝑓 (𝑧) = 𝑧log 𝑧
2 +𝑎2 , where we take the branch cut of the logarithm along −𝑖𝑏 for all 𝑏 ∈ [0, ∞).

Furthermore, 𝑖𝑎 is a pole of order 1. Finding the residue at 𝑖𝑎, we have


log 𝑧 log 𝑧 log 𝑖𝑎 log 𝑎 𝜋
res𝑖𝑎 𝑓 = lim (𝑧 − 𝑖𝑎) = lim = = +
𝑧→𝑖𝑎 2
𝑧 +𝑎 2 𝑧→𝑖𝑎 ( 𝑧 + 𝑖𝑎 ) 2𝑖𝑎 2𝑖𝑎 2𝑎
Label the contours from the portion on the positive real axis 𝛾1 , the larger arc 𝛾2 , the
portion on the negative real axis 𝛾3 , and the smaller arc 𝛾4 . Choose 𝜖 < min{𝑎, 1}, 𝑅 >
max{𝑎, 1}. Parametrize 𝛾1 with 𝑧(𝑡) = 𝑡 from 𝜖 to 𝑅, 𝛾2 with 𝑧(𝑡) = 𝑅𝑒 𝑖𝑡 from 0 to 𝜋, 𝛾3 with
𝑧(𝑡) = 𝑡 from −𝑅 to −𝜖, and 𝛾4 with 𝑧(𝑡) = 𝜖𝑒 𝑖𝑡 from 𝜋 to 0 . Integrating over the 𝛾2 and
taking absolute values, we have
| log 𝑧 | | 𝜋 log 𝑅𝑒 𝑖𝑡 |
| 𝑑𝑧 |=| 𝑅𝑖𝑒 𝑖𝑡 |
𝑑𝑡
|∫ 𝑧 2 + 𝑎2 | |∫ (𝑅𝑒 𝑖𝑡 ) + 𝑎2 2 |
| 𝛾2 | | 0 |
𝜋| 𝑖𝑡 |
log 𝑅𝑒
≤ ∫ || 2 2𝑖𝑡 2
𝑅𝑖𝑒 𝑖𝑡 || 𝑑𝑡
0 |𝑅 𝑒 + 𝑎 |
𝜋| 𝑖𝑡 |
log 𝑅𝑒
= ∫ || 2 2𝑖𝑡 2|
| 𝑅𝑑𝑡
0 | 𝑅 𝑒 + 𝑎 |
𝜋| |
log 𝑅 + 𝑖𝑡
= ∫ || 2 2𝑖𝑡 2
| 𝑅𝑑𝑡
|
0 | 𝑅 𝑒 + 𝑎 |
𝜋
log 𝑅 + |𝑖𝑡|
≤∫ 𝑅𝑑𝑡
0 |𝑅2 𝑒 2𝑖𝑡 | + |𝑎2 |
𝜋 𝜋
log 𝑅 + 𝑡 log 𝑅 + 𝑡 log 𝑅 + 𝜋
=∫ |𝑎2 |
𝑑𝑡 ≤ ∫ |𝑎2 | 𝑑𝑡 ≤ 𝜋 2
0 |𝑅𝑒 2𝑖𝑡 | + 𝑅 0 𝑅+ 𝑅 𝑅 + |𝑎𝑅 |
|𝑎2 |
since 𝑡, log 𝑅 > 0. Since 𝑅 → ∞, log 𝑅 + 𝜋, 𝑅 + 𝑅
→ ∞,
log 𝑅 + 𝜋 1/𝑅 1
lim |𝑎2 |
= lim |𝑎2 |
= lim |𝑎2 |
= 0.
𝑅→∞ 𝑅+ 𝑅
𝑅→∞ 1− 𝑅2
𝑅→∞ 𝑅− 𝑅

Date: April 2023.


1
2 ATHARV SAMPATH

Thus, as 𝑅 → ∞, the contribution along 𝛾2 vanishes to 0 . Similarly, for 𝛾4 , we have

| log 𝑧 | | 0 log 𝜖𝑒 𝑖𝑡 |
| 𝑑𝑧 |=| 𝜖𝑒 𝑖𝑡 |
𝑑𝑡
|∫ 𝑧 2 + 𝑎2 | |∫ (𝜖𝑒 𝑖𝑡 )2 + 𝑎2 |
| 𝛾4 | | 𝜋 |
0| 𝑖𝑡 |
log 𝜖𝑒
≤ ∫ || 2 2𝑖𝑡 2
𝜖𝑒 𝑖𝑡 || 𝑑𝑡
𝜋 | 𝜖 𝑒 + 𝑎 |
0| 𝑖𝑡 |
log 𝜖𝑒 |
= ∫ || 2 2𝑖𝑡 2|
𝜖𝑑𝑡
𝜋 |𝜖 𝑒 + 𝑎 |
0|
− log 𝜖 + 𝑖𝑡 ||
= ∫ || 2 2𝑖𝑡 2 |
𝜖𝑑𝑡
𝜋 | 𝜖 𝑒 +𝑎 |
0
− log 𝜖 + |𝑖𝑡|
≤∫ 2 2𝑖𝑡 2
𝜖𝑑𝑡
𝜋 |𝜖 𝑒 | + |𝑎 |
0 0
− log 𝜖 + 𝑡 − log 𝜖 + 𝑡 − log 𝜖 + 𝜋
≤∫ |𝑎2| 𝑑𝑡 ≤ ∫ |𝑎2| 𝑑𝑡 ≤ 𝜋 2 𝑑𝑡
𝜋 |𝜖𝑒 2𝑖𝑡 | + 𝜖 𝜋 𝜖+ 𝜖 𝜖 + |𝑎𝜖 |

|𝑎2 |
since 𝑡, − log 𝜖 > 0. Since 𝜖 → 0, − log 𝜖 + 𝜋, 𝜖 + 𝜖
→ ∞,

− log 𝜖 + 𝜋 −1/𝜖 −1
lim |𝑎2 |
= lim |𝑎2 |
= lim |𝑎2 |
= 0.
𝜖→0 𝜖+ 𝜖
𝜖→0 1− 𝜖2
𝜖→0 𝜖− 𝜖

Thus, as 𝜖 → 0, the contribution along 𝛾4 also vanishes to 0 . For the integral over 𝛾1 , 𝛾3 ,
we have
𝑅 −𝜖
log 𝑧 log 𝑡 log 𝑠
∫ 𝑑𝑧 = ∫ 𝑑𝑡 + ∫ 𝑑𝑠.
𝛾1 +𝛾3 𝑧 2 + 𝑎2 𝜖
2
𝑡 +𝑎 2 2
−𝑅 𝑠 + 𝑎
2

Letting 𝑠 = −𝑡, we have

𝑅 𝜖
log 𝑧 log 𝑡 log −𝑡
∫ 𝑑𝑧 = ∫ 𝑑𝑡 + ∫ (−1)𝑑𝑡
𝛾1 +𝛾3 𝑧 2 + 𝑎2 𝜖 𝑡 2 + 𝑎2 2
𝑅 (−𝑡) + 𝑎
2

𝑅 𝑅
log 𝑡 log −𝑡
=∫ 2 2
𝑑𝑡 + ∫ 𝑑𝑡
𝜖 𝑡 +𝑎 𝜖 𝑡 2 + 𝑎2
𝑅 𝑅
log 𝑡 log 𝑡 + 𝑖𝜋
=∫ 2 2
𝑑𝑡 + ∫ 𝑑𝑡
𝜖 𝑡 +𝑎 𝜖 𝑡 2 + 𝑎2
𝑅
log 𝑡 𝑅
1 𝑅
log 𝑡 𝑖𝜋 𝑡 ||𝑅
= 2∫ 2 2
𝑑𝑡 + 𝑖𝜋 ∫ 𝑑𝑡 = 2 ∫ 𝑑𝑡 + arctan |
𝜖 𝑡 +𝑎 𝜖 𝑡 + 𝑎2
2
𝜖 𝑡 2 + 𝑎2 𝑎 𝑎 |𝜖
COMPLEX ANALYSIS HOMEWORK 7 3

log 𝑎 𝜋
Thus we have, as 𝑅 → ∞, 𝜖 → 0 and as res𝑖𝑎 𝑓 = 2𝑖𝑎
+ 2𝑎
, we have
𝑅 | 𝑅
log 𝑡 𝑖𝜋 𝑡 log 𝑎 𝜋
lim 2∫ 𝑑𝑡 + arctan || = 2𝜋𝑖 +
𝑅→∞,𝜖→0 ( 𝜖 𝑡2
+𝑎 2 𝑎 𝑎 |𝜖 ) ( 2𝑖𝑎 2𝑎 )
𝑅
log 𝑡 𝑖𝜋 𝑡 ||𝑅 𝜋 log 𝑎 𝑖𝜋 2
lim 2∫ 𝑑𝑡 + lim arctan = +
𝑅→∞,𝜖→0 ( 𝜖 𝑡 2 + 𝑎2 ) 𝑅→∞,𝜖→0 ( 𝑎 𝑎 ||𝜖 ) 𝑎 𝑎
𝑅
log 𝑡 𝑖𝜋 2 𝜋 log 𝑎 𝑖𝜋 2
lim 2∫ 𝑑𝑡 + = +
𝑅→∞,𝜖→0 ( 𝜖 𝑡 2 + 𝑎2 ) 𝑎 𝑎 𝑎
𝑅
log 𝑡 𝜋 log 𝑎
lim 2∫ 𝑑𝑡 =
𝑅→∞,𝜖→0 ( 𝜖 𝑡 +𝑎 )
2 2 𝑎

log 𝑡 𝜋 log 𝑎
∫ 𝑡 2 + 𝑎2 𝑑𝑡 = 2𝑎
0
as desired. ■
Exercise 11
Proof. First, note that log |1 − 𝑎𝑒 𝑖𝜃 | is the real part of log(1 − 𝑎𝑒 𝑖𝜃 ). Let 𝑓 (𝑧) = log(1 − 𝑎𝑧).
Since 𝑓 is holomorphic on the unit disk, we have
2𝜋 2𝜋
1 𝑖𝜃 1
0 = 𝑓 (0) = ∫ 𝑓 (𝑒 )d𝜃 = ∫ log(1 − 𝑎𝑒 𝑖𝜃 )d𝜃
2𝜋 0 2𝜋 0
for all 𝑎 such that |𝑎| < 1. Then, only taking the real part, we have
2𝜋

∫ log |1 − 𝑎𝑒 𝑖𝜃 |d𝜃 = 0
0

as desired. If 𝑎 = 1, then note that we have ||1 − 𝑒 𝑖𝜃 || = 2(1 − cos(𝜃)) = 2| sin 𝜃/2|. Now,
we have
2𝜋 𝜋/2
| 𝑖𝜃 |
log |1 − 𝑒 | d𝜃 = 4 ∫ log(2 sin(𝜃))d𝜃 = 0

0 0
since
𝜋/2
𝜋
∫ log(sin(𝜃))d𝜃 = − log(2).
0 2
Therefore, the equality holds even when |𝑎| = 1. ■
4 ATHARV SAMPATH

Exercise 15
b.
Proof. Write 𝑆 ∶= {0 < Arg(𝑧) < 𝜙} and choose 𝑛 large enough so that
𝔻 ⊆ 𝑆 ∪ 𝜁𝑛 𝑆 ∪ 𝜁𝑛2 𝑆 ∪ ⋯ ∪ 𝜁𝑛𝑛−1 𝑆
so that the rotated sectors cover the disc. By uniform convergence of 𝑓 to 0 on 𝑆, choose
𝑟 < 1 small enough so that |𝑓 (𝑧)| < 𝜀 for |𝑧| < 𝑟 in 𝑆. Note that 𝔻𝑟 ⊆ ⋃𝑛−1 𝑘
𝑘=0 𝜁𝑛 𝑆𝑟 , where
𝑆𝑟 ∶= {𝑧 ∈ 𝑆 ∣ |𝑧| ≤ 𝑟} is a subsector of radius 𝑟. By the maximum modulus priciple, let 𝑀
be the maximum of 𝑓 on 𝔻, which is (WLOG) attained at some point on 𝑆. Then |𝑓 | < 𝑀
on every 𝜁𝑛𝑘 𝑆𝑟 . Now define
𝑛−1 𝑛−1
𝑘
𝑔(𝑧) ∶= 𝑓 (𝑧) ∏ 𝑓 (𝜁𝑛 𝑧 ) ∶= 𝑓 (𝑧) ∏ 𝑓𝑘 (𝑧).
𝑘=1 𝑘=1
Note that |𝑓 (𝑧)| ≤ 𝜀 and |𝑓𝑘 (𝑧)| ≤ 𝑀, so
𝜀→0
|𝑔(𝑧)| ≤ 𝜀 ⋅ 𝑀 𝑛−1 ⟶ 0
since 𝑀 is a constant. So 𝑔(𝑧) ≡ 0 on 𝔻𝑟 , and by the identity principle, on 𝔻. Thus some
factor 𝑓𝑘 (𝑧) is identically zero. But if 𝑓 (𝜁𝑛𝑘 𝑧 ) ≡ 0 on 𝔻, then 𝑓 (𝑧) ≡ 0 on 𝔻, since every
𝑧 ∈ 𝔻 can be written as 𝜁𝑛𝑘 𝑤 for some 𝑤 ∈ 𝔻. ■
d.
Proof. Write 𝑓 = 𝑢 + 𝑖𝑣 where by assumption 𝑢 is bounded. Both 𝑢 and 𝑣 are harmonic,
so if |𝑢| ≤ 𝑀 on ℂ, there is some disc where |𝑢| = 𝑀 for some point in the interior. By the
maximum modulus principle, 𝑢 is constant on ℂ. So 𝑢𝑥 , 𝑢𝑦 = 0, and by Cauchy-Riemann,
𝑣𝑥 , 𝑣𝑦 = 0, so 𝑣′ = 0 and 𝑣 is constant, making 𝑓 constant. ■
Exercise 16
a.
Proof. Note that if 𝜖 is sufficiently small, then |𝜖𝑔(𝑧)| < |𝑓 (𝑧)|. Then, we can apply Rouche’s
theorem to see that 𝑓 and 𝑓 + 𝜖𝑔 have the same number of zeros inside the unit disk.
Since 𝑓 has a unique zero in the unit disk, so does 𝑓 + 𝜖𝑔. ■
b.
Proof. By the argument principle we have
𝑓 ′ (𝑧)
𝑧𝜖 = ∫ 𝑧 d𝑧,
𝛾 𝑓 (𝑧)

which shows that 𝜖 ↦ 𝑧𝜖 is continuous as desired. ■


COMPLEX ANALYSIS HOMEWORK 7 5

Exercise 17
a.
Proof. Suppose 𝑓 (𝑧) does not have a zero in the unit disc, 𝔻. Then 1/𝑓 (𝑧) is holomorphic
in 𝔻. Note that since |𝑓 (𝑧)| = 1 whenever |𝑧| = 1, |1/𝑓 (𝑧)| = 1/|𝑓 (𝑧)| = 1 whenever
|𝑧| = 1 as well. But 𝑓 (𝑧) is holomorphic in 𝔻, implying |𝑓 (𝑧)| ≤ 1 in 𝔻 by the maximum
modulus principle since |𝑓 (𝑧)| = 1 on the boundary of 𝔻. We find 1 ≤ |𝑓 (𝑧)| ≤ 1 in
the unit disk, which implies that our function is constant as its modulus is constant, a
contradiction. Let 𝑤0 ∈ 𝔻. Consider the constant function 𝑔(𝑧) = −𝑤0 . On the unit circle,
|𝑓 (𝑧)| = 1 > |𝑤0 | = |𝑔(𝑧)| for all |𝑧| = 1. By Rouché’s theorem, 𝑓 (𝑧) and 𝑓 (𝑧) + 𝑔(𝑧) have
the same number of zeros inside the unit circle (in 𝔻). But we have shown that 𝑓 (𝑧) has
at least one zero, thus for some 𝑧𝑤 , 0 = 𝑓 (𝑧𝑤 ) + 𝑔 (𝑧𝑤 ) = 𝑓 (𝑧𝑤 ) − 𝑤0 . Thus for all 𝑤0 ∈ 𝔻,
there exists 𝑧𝑤 such that 𝑓 (𝑧𝑤 ) = 𝑤0 . Thus the image of 𝑓 (𝑧) contains the unit disc. ■
b.
Proof. Since |𝑓 | is continuous on the compact closed unit disk, the modulus of 𝑓 attains
a minimum on the closed unit disk. By the maximum modulus principle, the minimum
must be attained inside the open unit disk since 𝑓 (𝑧0 ) < 1 but 𝑓 (𝑧) ≥ 1 when |𝑧| = 1.
Now assume that the minimum is attained at 𝑧 ′ ∈ 𝔻 and |𝑓 (𝑧 ′ )| > 0. Then 𝑓 (𝑧) is never
1
zero inside 𝔻 so 𝑓 (𝑧) is holomorphic in 𝔻. Now we have that
1
≥1 𝑧∈𝔻
|𝑓 (𝑧)|
by the maximum modulus principle but this contradicts the existance of the point 1 >
|𝑓 (𝑧 ′ )| > 0 Therefore, there is a point 𝑧 ∈ 𝔻 at which 𝑓 (𝑧) = 0. Now consider 𝑓 (𝑧) − 𝑤0
for 𝑤0 ∈ 𝔻 Since |𝑓 (𝑧)| ≥ |−𝑤0 | when 𝑧 ∈ 𝜕𝔻 by rouche’s theorem, 𝑓 (𝑧) − 𝑤 has the same
number of zeros as 𝑓 which has at least one. Therefore, finishing in the same way as
above, the unit disk is contained in the image of 𝑓 . ■

You might also like