Coefficient in a Laurent Expansion in terms of an Integral

Click For Summary

Discussion Overview

The discussion revolves around determining the coefficient \(C_n\) in the Laurent expansion of the function \(f(z) = \cos(z + z^{-1})\) about \(z=0\). Participants explore various methods to derive the expression for \(C_n\) and discuss the implications of the function's properties on the coefficients, particularly focusing on the behavior of the integral representation of \(C_n\) for different integer values of \(n\).

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • Some participants propose using the formula for Laurent series coefficients, \(C_n = \frac{1}{2\pi i} \oint_\gamma \frac{f(z)\,\mathrm{d}z}{(z-c)^{n+1}}\), with \(\gamma\) being a closed curve around \(c\).
  • One participant suggests parametrizing the contour by letting \(z = e^{i\theta}\) and transforming the integral accordingly.
  • Another participant notes that the integral for \(C_n\) is zero when \(n\) is odd, but does not provide a definitive explanation for this deduction.
  • A later reply introduces an indirect argument involving residues, indicating that the coefficient \(C_n\) must be zero for odd \(n\) based on the nature of the series expansion.
  • Another approach is presented that utilizes the product of series expansions for \(\cos(z)\) and \(\cos(z^{-1})\), suggesting that all terms are even, which supports the claim that \(C_n\) is zero for odd \(n\).

Areas of Agreement / Disagreement

Participants express differing views on the methods to derive \(C_n\) and the implications of the function's properties. While there is a consensus that \(C_n\) is zero for odd \(n\), the reasoning behind this conclusion is debated, with multiple approaches being presented without a clear resolution on the preferred method.

Contextual Notes

Some participants note that the integral representation of \(C_n\) depends on the analytic properties of the function within the annulus defined by the contour. There are also unresolved aspects regarding the assumptions made in the series expansions and the conditions under which the integrals are evaluated.

shen07
Messages
54
Reaction score
0
Hi guys, i need your help to go about his question,

Question:

$$\text{Show that the coefficient }C_n \text{in the Laurent expansion of }$$
$$f(z)=(z+\frac{1}{z}) \text{ about z=0 is given by}$$
$$C_n=\frac{1}{2\pi}\int^{2\pi}_0 \cos(2cos(\theta))cos(n\theta)\, d\theta ,n\in\mathbb{z}$$
 
Physics news on Phys.org
shen07 said:
Hi guys, i need your help to go about his question,

Question:

$$\text{Show that the coefficient }C_n \text{in the Laurent expansion of }$$
$$f(z)=(z+\frac{1}{z}) \text{ about z=0 is given by}$$
$$C_n=\frac{1}{2\pi}\int^{2\pi}_0 \cos(2cos(\theta))cos(n\theta)\, d\theta ,n\in\mathbb{z}$$

the function is actually $$f(z)=cos(z+z^{-1})$$
 
shen07 said:
the function is actually $$f(z)=cos(z+z^{-1})$$

Hi shen07,

Perhaps you can start by filling in your function into the formula for the Laurent series coefficients?
The formula is:
$$C_n=\frac{1}{2\pi i} \oint_\gamma \frac{f(z)\,\mathrm{d}z}{(z-c)^{n+1}}$$
where the function f(z) is expanded around c and $\gamma$ is a closed counter clockwise curve around c.
 
Let $\gamma$ be the unit circle.

Then since $\gamma$ lies in a annulus in which $ \displaystyle \cos \left( z + \frac{1}{z} \right)$ is analytic,

$ \displaystyle C_{n} = \frac{1}{2\pi i} \int_{\gamma} \frac{\cos (z + \frac{1}{z})}{z^{n+1}} \ dz = \frac{1}{4 \pi i} \int_{\gamma} \frac{e^{i(z + \frac{1}{z})} + e^{-i(z+ \frac{1}{z})}}{z^{n+1}} \ dz $

Parametrize the contour by letting $z = e^{i \theta}$.

$\displaystyle = \frac{1}{4 \pi i} \int_{-\pi}^{\pi} \frac{e^{i (e^{i\theta}+e^{-i\theta})} + e^{i(e^{i\theta}+e^{-i\theta})}}{e^{i(n+1)\theta}} i e^{i\theta} \ d \theta = \frac{1}{4 \pi} \int_{- \pi}^{\pi} \frac{e^{i (2 \cos \theta)} + e^{-i(2 \cos \theta)}}{e^{i n \theta}} \ d \theta$

$ = \displaystyle \frac{1}{2 \pi} \int_{-\pi}^{\pi} \cos(2 \cos \theta) e^{-i n \theta} \ d \theta = \frac{1}{2 \pi} \int_{-\pi}^{\pi} \cos(2 \cos \theta) \cos(n \theta)$ since the imaginary part of the integrand is odd

$ \displaystyle = \frac{(-1)^{n}}{2 \pi} \int_{0}^{2 \pi} \cos(2 \cos u) \cos(n u) \ du$

But since the integral is zero when $n$ is odd,

$ \displaystyle C_{n} = \frac{1}{2 \pi} \int_{0}^{2 \pi} \cos(2 \cos u) \cos(n u) \ \ du$
 
Random Variable said:
$ \displaystyle = \frac{(-1)^{n}}{2 \pi} \int_{0}^{2 \pi} \cos(2 \cos u) \cos(n u) \ du$

But since the integral is zero when $n$ is odd,

How did you deduce that the integral is zero ?
 
ZaidAlyafey said:
How did you deduce that the integral is zero ?

I used magic. And by "magic" I mean Maple.
 
Here's an indirect argument.Let $n$ be an odd integer.$ \displaystyle C_{n} = \frac{1}{2 \pi i} \int_{|z|=1} \frac{\cos (z+ \frac{1}{z})}{z^{n+1}} \ dx = \frac{1}{2 \pi i} \text{Re} \int_{|z|=1} \frac{e^{i(z+\frac{1}{z})}}{z^{n+1}} \ dz = \text{Re} \ \text{Res}\Big[ \frac{e^{i(z+\frac{1}{z})}}{z^{n+1}},0 \Big]$$ \displaystyle \frac{e^{i(z+\frac{1}{z})}}{z^{n+1}} = \frac{1}{z^{n+1}} \sum_{k=0}^{\infty} \frac{[i(z+\frac{1}{z})]^{k}}{k!}= \frac{1}{z^{n+1}} \sum_{k=0}^{\infty} \frac{i^{k}}{k!} \sum_{m=0}^{k} \binom{k}{m} z^{m} \left(\frac{1}{z} \right)^{k-m} $

$ \displaystyle = \sum_{k=0}^{\infty} \frac{i^{k}}{k!} \sum_{m=0}^{k} \binom{k}{m} z^{2m-k-n-1} $We want to the know the coefficient of the $\displaystyle \frac{1}{z}$ term.

So we're only interested when $ \displaystyle m = \frac{k+n}{2}$.

But since $n$ is assumed to be odd, $k$ must also be odd.

But if $k$ can only assume odd values, then every term of the series is imaginary. So $C_{n}$ must be zero when $n$ is odd.
 
Another way would be using

$$\cos(z+z^{-1}) = \cos(z) \cos\left(\frac{1}{z}\right)-\sin(z) \sin \left( \frac{1}{z}\right)=\left(1-\frac{z^2}{2!}+\frac{z^4}{4!}+\cdots \right)\left(1-\frac{1}{2!\,z^2}+\frac{1}{4!\,z^4}+\cdots \right)-\left(z-\frac{z^3}{3!}+\frac{z^5}{5!}\cdots \right)\left(\frac{1}{z}-\frac{1}{3!\,z^3}+\frac{1}{5!\,z^5}+\cdots \right)$$

Evidently all the terms are even . Another way would be using Cauchy product formula.
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 9 ·
Replies
9
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 7 ·
Replies
7
Views
4K
  • · Replies 2 ·
Replies
2
Views
2K
Replies
8
Views
4K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 15 ·
Replies
15
Views
4K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K