Conditions on f for dx/f(x) to be a Measure in L^2

  • Context: Graduate 
  • Thread starter Thread starter Jip
  • Start date Start date
  • Tags Tags
    Conditions Measure
Click For Summary

Discussion Overview

The discussion revolves around the conditions required for the expression dx/f(x) to be a valid measure on the real numbers, particularly in the context of a Hilbert space L^2(R, dx/f(x)). Participants explore the necessary properties of the function f, including its positivity and measurability.

Discussion Character

  • Technical explanation
  • Mathematical reasoning

Main Points Raised

  • One participant inquires about the conditions on the function f, specifically that it must be nowhere zero and positive for dx/f(x) to be a valid measure.
  • Another participant suggests that for dx/f(x) to be a valid measure, 1/f must be measurable and sum to one, while noting that the Hilbert space will only contain functions a for which the integral of a*a/f is finite.
  • It is mentioned that in addition to being positive, f must also be measurable.
  • One participant asserts that if 1/f is measurable, then f is also measurable.

Areas of Agreement / Disagreement

Participants generally agree on the necessity of f being positive and measurable, but there is no consensus on additional specific conditions that may apply to f for the measure to be valid.

Contextual Notes

Some assumptions regarding the positivity and measurability of f are acknowledged, but the discussion does not resolve whether further restrictions on f are necessary for specific classes of functions.

Jip
Messages
20
Reaction score
2
Hi,
Let's say I consider the real numbers and some function real function f, nowhere zero, and positive.
My question is, what are the conditions on f for dx/f(x) to be a valid measure on this space?

(I have to consider a Hilbert space L^2(R, dx/f(x)) with scalar product a.b = \int a^*(x) b(x) \frac{dx}{f(x)}

I'm a physicist, so please excuse me if this is not written in perfect mathematical language!
Thanks
 
Physics news on Phys.org
To get a valid measure you just need it (1/f) (to be measurable and) to sum to one; but your Hilbert space will contain only functions a such that the integral of a*a/f is finite. If you need this to apply to a specific class of functions then you'll have to restricf f accordingly.
 
In addition to you assumptions (everywhere or almost everywhere positive) it needs to be measurable, and that is all.
 
Hawkeye18 said:
In addition to you assumptions (everywhere or almost everywhere positive) it needs to be measurable, and that is all.
Thanks! And when 1/f is measurable, is it also true for f?
 
Yes, ##1/f## is measurable if and only if ##f## is.
 

Similar threads

  • · Replies 5 ·
Replies
5
Views
4K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
Replies
14
Views
2K
  • · Replies 0 ·
Replies
0
Views
2K
  • · Replies 14 ·
Replies
14
Views
3K
  • · Replies 1 ·
Replies
1
Views
5K
  • · Replies 8 ·
Replies
8
Views
6K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K