Fourier transform of the linear function

Click For Summary

Discussion Overview

The discussion revolves around the Fourier transform of the linear function, specifically the integral of the form \(\int_{-\infty}^{+\infty} x e^{ikx}\, dx\). Participants explore the meaning and implications of this expression, addressing its mathematical validity and the appearance of the Dirac delta function.

Discussion Character

  • Exploratory, Technical explanation, Debate/contested

Main Points Raised

  • One participant questions the meaning of the Fourier transform of the linear function and suggests it evaluates to \(\frac{\delta(k)}{ik}\).
  • Another participant argues that the expression is intrinsically undefined and provides an alternative approach involving differentiation of the delta function.
  • A later reply expresses confusion about the appearance of a minus sign in the differentiation process.
  • Another participant clarifies the differentiation of the exponential function and notes the necessity of a minus sign to align with the original integral.

Areas of Agreement / Disagreement

Participants express differing views on the validity of the Fourier transform of the linear function, with some asserting it is undefined while others attempt to provide a rationale for its evaluation. The discussion remains unresolved regarding the interpretation of the integral.

Contextual Notes

Participants highlight the dependence on definitions and the potential ambiguity in the treatment of the integral, particularly concerning the Dirac delta function and its derivatives.

Irid
Messages
207
Reaction score
1
Hello,
I was wondering if one can give meaning to the Fourier transform of the linear function:

\int_{-\infty}^{+\infty} x e^{ikx}\, dx

I found that it is \frac{\delta(k)}{ik}, does this make sense?
 
Physics news on Phys.org
This expression doesn't make sense since it's intrinsically undefined. A handwaving way is
\int_{\mathbb{R}} \mathrm{d} x x \exp(\mathrm{i} k x)=-\mathrm{i} \frac{\mathrm{d}}{\mathrm{d} k} \int_{\mathbb{R}} \mathrm{d} x \exp(\mathrm{i} k x)=-2 \pi \mathrm{i} \frac{\mathrm{d}}{\mathrm{d} k} \delta(k).
 
  • Like
Likes   Reactions: 1 person
vanhees71 said:
This expression doesn't make sense since it's intrinsically undefined. A handwaving way is
\int_{\mathbb{R}} \mathrm{d} x x \exp(\mathrm{i} k x)=-\mathrm{i} \frac{\mathrm{d}}{\mathrm{d} k} \int_{\mathbb{R}} \mathrm{d} x \exp(\mathrm{i} k x)=-2 \pi \mathrm{i} \frac{\mathrm{d}}{\mathrm{d} k} \delta(k).
Hmm.. seems to make sense. Why is there a minus sign popping up?
 
d/dk(exp(ikx)) = ixexp(ikx). you need -i to get 1 for the original integral.
 
  • Like
Likes   Reactions: 1 person

Similar threads

  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
12K
  • · Replies 5 ·
Replies
5
Views
2K
Replies
1
Views
3K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 3 ·
Replies
3
Views
9K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K