Fourier transform equation question

Click For Summary

Discussion Overview

The discussion revolves around a Fourier transform equation encountered in quantum field theory (QFT) homework. Participants are analyzing the steps involved in proving an integral identity related to the Fourier transform and addressing the implications of certain mathematical assumptions.

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant presents a proof involving integrals and the Fourier transform, questioning the treatment of a term that appears to vanish.
  • Another participant asserts that the term is a total derivative and that the function is assumed to go to zero at infinity.
  • A subsequent reply seeks clarification on the nature of total derivatives versus partial derivatives, expressing confusion about the assumptions made regarding the function's behavior at infinity.
  • Another participant clarifies that being a total derivative does not exclude it from being a partial derivative, emphasizing that the integrand is a derivative of some function and that assumptions about the function's behavior are standard in this context.

Areas of Agreement / Disagreement

Participants express differing views on the nature of the derivative involved and the validity of the assumptions regarding the function's behavior at infinity. No consensus is reached on these points.

Contextual Notes

The discussion highlights the dependence on assumptions about the behavior of functions in the context of Fourier transforms, which may not be universally applicable without further clarification.

Fosheimdet
Messages
15
Reaction score
2
In my QFT homework I was asked to prove that $$\int d^3x \int \frac{d^3k}{(2\pi)^3} e^{i\mathbf{k} \cdot (\mathbf{x} - \mathbf{y})} k_j f(\mathbf{x}) = i \frac{df}{dx_j}(\mathbf{y}) $$

Using ##\frac{\partial e^{i\mathbf{k} \cdot (\mathbf{x} - \mathbf{y})}}{\partial x^j} = i k_j e^{i\mathbf{k} \cdot (\mathbf{x} - \mathbf{y})}##, the solution is as follows

$$\int d^3x \int \frac{d^3k}{(2\pi)^3} e^{i\mathbf{k} \cdot (\mathbf{x} - \mathbf{y})} k_j f(\mathbf{x}) = -i \int d^3x \int \frac{d^3k}{(2\pi)^3} \frac{\partial e^{i\mathbf{k} \cdot (\mathbf{x} - \mathbf{y})}}{\partial x^j} f(\mathbf{x}) $$

$$ = -i \int d^3x \int \frac{d^3k}{(2\pi)^3} \left( \frac{\partial}{\partial x^j} (e^{i\mathbf{k} \cdot (\mathbf{x} - \mathbf{y})}f(\mathbf{x}) ) - \frac{\partial f(\mathbf x)}{\partial x^j} e^{i\mathbf k \cdot (\mathbf x - \mathbf y)}\right) $$

$$ = i \int d^3x \frac{\partial f(\mathbf x) }{\partial x^j} \int \frac{d^3k}{(2\pi)^3} e^{i\mathbf k \cdot (\mathbf x - \mathbf y)} $$

$$= i \int d^3x \frac{\partial f(\mathbf x)}{\partial x^j} \delta^{(3)}(\mathbf x - \mathbf y) = i \frac{\partial f(\mathbf x)}{\partial x^j} \bigg|_{\mathbf x = \mathbf y}$$

I don't understand what happened to the first term on the RHS of the second equation. Why is ##-i\int d^3x \int \frac{d^3k}{(2\pi)^3} \frac{\partial}{\partial x^j} (e^{i\mathbf{k} \cdot (\mathbf{x} - \mathbf{y})}f(\mathbf{x}))=0?##
 
Last edited by a moderator:
Physics news on Phys.org
It is a total derivative and the function is assumed to go to zero sufficiently fast at infinity.
 
Orodruin said:
It is a total derivative and the function is assumed to go to zero sufficiently fast at infinity.
How is it a total derivative though? Isn't ##\frac {\partial}{\partial x^j}## a partial derivative? Also, why can you make the assumption that the function goes to zero at infinity?
 
Being a total derivative does not prevent being a partial derivative. It just means that the integrand is a derivative of some function.

An assumption is an assumption. We typically choose to work with such functions. It is related to the domain of the Fourier transform.
 

Similar threads

  • · Replies 0 ·
Replies
0
Views
3K
  • · Replies 2 ·
Replies
2
Views
4K
  • · Replies 12 ·
Replies
12
Views
2K
  • · Replies 1 ·
Replies
1
Views
896
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 5 ·
Replies
5
Views
1K
  • · Replies 3 ·
Replies
3
Views
958
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 6 ·
Replies
6
Views
3K