Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Proof Fourier Series Converge

  1. Aug 19, 2013 #1
    Hey. I'm looking for a proof of:
    Theorem: If [itex]f \in C^1(\mathbb{T})[/itex], then the Fourier series converges to f uniformly (and hence also pointwise.)

    I have looked around for it, googled, etc, but I only found proofs which used theorem they did not prove. (Or I misunderstood what they said.)
    I'd really like to truly understand why they converge, be it uniformly or pointwise. If anyone could either link me to a proof or do it, it'd be great. Thanks.
  2. jcsd
  3. Aug 19, 2013 #2
    Check Folland's "Fourier analysis and applications", Theorem 2.5
  4. Aug 28, 2013 #3
    Sorry, I don't know what C^1(T) is. Are these complex functions? And what domain is T?
  5. Sep 8, 2013 #4
    Completely answered my question. Thanks a lot!

    C^k is the set of functions such that: There exist continuous derivatives of 0th, 1st, 2nd... and kth order.
    C^1(T) probably means that f is periodic of period 2π or something of the sorts. The number of senseful meanings is not that big.
  6. Sep 8, 2013 #5
    Thanks Swimmingly. Then[tex] C^1[/tex] would be real functions that have just one continuous derivative and you guess T means they are periodic. I would guess if they are not periodic one could construct examples where the Fourier series wouldn't converge at all. There is always the famous [tex]{x^2sin1/x}[/tex] which has exactly one continuous derivative at 0.
    Last edited: Sep 8, 2013
  7. Sep 8, 2013 #6


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Bold faced T usually refers to a torus - in this case I assume the one dimensional torus (which is the circle, equivalently we are discussing periodic functions on the real line)
  8. Sep 8, 2013 #7
    Office_Shredder- Thanks for the clarification. He most likely meant periodic functions on the real line. I personally am more than willing to consider functions on the unit circle -- quite often it helps.

    Also, for all, please substitute [tex]x^3sin1/x[/tex] for [tex]x^2sin1/x[/tex] as per my previous post. The first does have one continuous derivative at x = 0. The second has only a discontinuous derivative at x = 0.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook