For a function $f(x)$ where $x\in[0,1]$, the Fourier series has a fundamental frequency of $2\pi$. But I noticed that in the Fourier series expansion of the solution of heat equation (with same domain and homogeneous boundary conditions), the fundamental frequency was $\pi$. But why is that the case? Won't the set of functions having $2\pi$ as the fundamental frequency already form a complete set of basis? So, shouldn't the extra frequencies and the corresponding terms ($\sin \pi x, \sin 3\pi x, \sin 5\pi x$....) be redundant?

  • $\begingroup$ If you want to expand in functions with a period of $1$, then you want to use $\{ e^{2\pi i nx }\}_{n=-\infty}^{\infty}$. $\endgroup$ – COVID-20 Sep 15 at 3:04

$\{ \sin(n\pi x) \}_{n=1}^{\infty}$ is a complete orthogonal basis of $L^2[0,1]$. And $\{ e^{2\pi i nx} \}_{n=-\infty}^{\infty}$ is a complete orthonormal basis of $L^2[0,1]$. You can expand $\sin(\pi x)$ in $L^2[0,1]$ using a series $\sum_{n=-\infty}^{\infty}a_n e^{2\pi inx}$. That might seem a little strange, but no more strange than being able to expand $\cos(\pi x)$ in an $L^2$ convergent series of functions $\{ \sin(n\pi x) \}_{n=0}^{\infty}$. You'll get $L^2$ convergence, but obviously that won't translate to pointwise convergence at every point of $[0,1]$, and it doesn't have to in order to get $L^2[0,1]$ convergence.

In the same way, you can expand $\cos(\pi x/19)$ in a series of $\{ \sin(n\pi x) \}_{n=1}^{\infty}$, and the series will converge in $L^2[0,1]$. It all seems unlikely at first glance, but it's all part of the Mathemagic.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.