Understanding theorem in Fourier series

Click For Summary
SUMMARY

The discussion centers on the Fourier series representation of a continuous and periodic function $f$ with period $2\pi$. It establishes that if $f'$ exists and is in $\mathcal{P}\mathcal{C}[-\pi,\pi]$, then the series $\sum\limits_{k = -\infty}^{\infty}\lvert A_k\rvert < \infty$ holds true. The coefficients $A_n'$ are derived using integration by parts, leading to the conclusion that the boundary terms vanish due to the periodicity of $f$, specifically $f(\pi) = f(-\pi)$, and the exponential terms simplify to $(-1)^n$.

PREREQUISITES
  • Understanding of Fourier series and their convergence properties
  • Knowledge of integration by parts in calculus
  • Familiarity with periodic functions and their properties
  • Basic concepts of complex exponentials and their behavior
NEXT STEPS
  • Study the properties of Fourier series convergence in detail
  • Learn advanced techniques in integration by parts
  • Explore the implications of periodicity in function analysis
  • Investigate the role of complex exponentials in signal processing
USEFUL FOR

Mathematicians, engineers, and students studying signal processing or harmonic analysis who seek a deeper understanding of Fourier series and their applications.

Dustinsfl
Messages
2,217
Reaction score
5
Suppose $f$ is continuous and periodic with period $2\pi$ on $(-\infty,\infty)$, and $f'$ exist and is in $\mathcal{P}\mathcal{C}[-\pi,\pi]$.
Then $\sum\limits_{k = -\infty}^{\infty}\lvert A_k\rvert < \infty$.$f'$ has a Fourier series so let's call the coefficients $A_n'$. Then $f' = \sum\limits_{n = -\infty}^{\infty}A_n'e_n$ and $A_n' = \frac{1}{2\pi}\int_{-\pi}^{\pi}f'(\theta)e^{-in\theta}d\theta$. Let's integrate by parts letting $u = e^{-in\theta}$ and $dv = f'(\theta)$. Then $du = -ine^{-in\theta}d\theta$ and $v = f(\theta)$.
\begin{alignat*}{3}
A_n' & = & \frac{1}{2\pi}\left[\left. f(\theta)e^{-in\theta}\right|_{-\pi}^{\pi} + in\int_{-\pi}^{\pi}f(\theta)e^{-in\theta}d\theta\right]\\
& = & \frac{1}{2\pi}\left[\underbrace{f(\pi)e^{-in\theta} - f(-\pi)e^{in\theta}}_{\text{this part}} + in\int_{-\pi}^{\pi}f(\theta)e^{-in\theta}d\theta\right]
\end{alignat*}

I know that $f(\pi) = f(-\pi)$ but why is that piece zero? We have $e^{in\theta}$ and $e^{-in\theta}$.
 
Physics news on Phys.org
dwsmith said:
Suppose $f$ is continuous and periodic with period $2\pi$ on $(-\infty,\infty)$, and $f'$ exist and is in $\mathcal{P}\mathcal{C}[-\pi,\pi]$.
Then $\sum\limits_{k = -\infty}^{\infty}\lvert A_k\rvert < \infty$.$f'$ has a Fourier series so let's call the coefficients $A_n'$. Then $f' = \sum\limits_{n = -\infty}^{\infty}A_n'e_n$ and $A_n' = \frac{1}{2\pi}\int_{-\pi}^{\pi}f'(\theta)e^{-in\theta}d\theta$. Let's integrate by parts letting $u = e^{-in\theta}$ and $dv = f'(\theta)$. Then $du = -ine^{-in\theta}d\theta$ and $v = f(\theta)$.
\begin{alignat*}{3}
A_n' & = & \frac{1}{2\pi}\left[\left. f(\theta)e^{-in\theta}\right|_{-\pi}^{\pi} + in\int_{-\pi}^{\pi}f(\theta)e^{-in\theta}d\theta\right]\\
& = & \frac{1}{2\pi}\left[\underbrace{f(\pi)e^{-in\theta} - f(-\pi)e^{in\theta}}_{\text{this part}} + in\int_{-\pi}^{\pi}f(\theta)e^{-in\theta}d\theta\right]
\end{alignat*}

I know that $f(\pi) = f(-\pi)$ but why is that piece zero? We have $e^{in\theta}$ and $e^{-in\theta}$.
You actually have $e^{in\pi}$ and $e^{-in\pi}$, and these are the same because they are both equal to $(-1)^n.$
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 14 ·
Replies
14
Views
2K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 6 ·
Replies
6
Views
2K