Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I am working on my understanding of various transforms and I have been thinking about the Fourier transform, what i "does" to the function it is applied to.

The way I see it:

The function $f$ that is transformed is multiplied with $\exp(icx)$ which essentially describes a rotating vector in the complex plane.

  • If $f$ is periodic cosine and the period of $f$ does not match the period of $\exp(icx)$ the "terms" in the integral will vary and cancel each other leading to a value of zero for the transform.

  • If $f$ is periodic cosine and the period of $f$ matches the period of $\exp(icx)$ the "terms" in the integral will be constant and the value of the transform will be $\infty$.

  • If $f$ is periodic but not a cosine it can be decomposed to a sum of cosines and the different cosine terms of this sum will work as above resulting in a spectra for $f$.

If this is somewhat correct I wonder:

  1. What about aperiodic functions?

  2. Is there a similar way of thinking about the Laplace transform?

Please forgive the non mathematical language, I'm neither a math major nor is English my mother tongue.

share|improve this question
    
I really think that en.wikipedia.org/wiki/Fourier_transform will answer all your questions. –  Josh Dec 12 '12 at 10:53
    
You're not completely wrong but I don't think that's the best way to think at it either. Periodic functions are to be represented with Fourier series, period (pun intended :-) ). Now there clearly is relationship between Fourier series and Fourier integrals, but that's another story. –  Giuseppe Negro Jan 28 '13 at 23:43
add comment

1 Answer

I think that a better intuition is that of thinking at Fourier decompositions (series or integrals) as a change-of-perspective operation. It is something similar in spirit to a change of coordinate in geometry.

Suppose you have a function $f\colon \mathbb{R}\to \mathbb{C}$ which is $T$-periodic. Then the following family of sinusoidal functions is compatible with $f$ in the sense that a period of $f$ contains an integer number of periods of each one of them: $$\sin\left( \frac{2\pi n}{T}t\right),\quad \cos\left( \frac{2\pi n}{T}t\right),\quad n=0, 1, 2, \ldots$$ Now the theory of Fourier series tells us that we can decompose $f$ as the superposition of those sinusoidal functions: $$f(t)=\frac{a_0}{2}+\sum_{n=1}^\infty \left( a_n\cos\left( \frac{2\pi n}{T}t\right)+b_n\sin\left( \frac{2\pi n}{T}t\right)\right),$$ which is usually rewritten in complex notation $$\tag{D} f(t)=\sum_{n=-\infty}^\infty c_n e^{i\frac{2\pi n}{T}t}.$$ (Here "D" stands for "discrete"). The map $f\to (c_n)$ is a kind of Fourier transform. It is a coordinate change in the sense that $f(t)$ is a representation of the original information as a function of $t$ (say, time) while $c_n$ is a representation of the same thing as a function of $n$, which is a number indicating frequency.

If $f$ is not periodic, we may regard it as a function having "infinite period", and construct an analogue of the previous decomposition by letting $T\to \infty$ with some care. The result is the following formula: $$\tag{C}f(t)=\int_{-\infty}^\infty \tilde{f}(\nu)e^{2\pi i \nu t}\, d\nu,$$ ("C" stands for "continuum") where $$\tilde{f}(\nu)=\int_{-\infty}^\infty f(t)e^{-2\pi i t\nu}\, dt$$ is the Fourier transform. Exactly as before, the map $f(t)\to \tilde{f}(\nu)$ is a coordinate change. It serves the purpose of switching between a representation of a piece of information as a function of $t$ (time) and a representation of the same information as a function of $\nu$, which is a frequency.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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