Sard theorem in dimension one.

  • Thread starter Thread starter boboYO
  • Start date Start date
  • Tags Tags
    Dimension Theorem
Click For Summary

Homework Help Overview

The discussion revolves around the Sard theorem in one dimension, specifically focusing on proving that the set of critical values of a continuously differentiable function has measure zero. The original poster seeks hints for their proof approach and expresses uncertainty about the necessity of the continuity of the derivative.

Discussion Character

  • Exploratory, Conceptual clarification, Mathematical reasoning

Approaches and Questions Raised

  • The original poster attempts to establish that the image of the set of critical points under the function has measure zero, using a construction involving intervals and bounds on the measure. They question whether their proof adequately utilizes the continuity of the derivative.
  • Another participant suggests defining a set based on the derivative being less than a small epsilon, proposing a different approach to show the measure of the image of this set is controlled by the measure of the set itself.
  • A third participant references the continuity of the derivative and the Mean Value Theorem to argue that the measure of the image of intervals can be bounded, contributing to the overall proof.

Discussion Status

Contextual Notes

Participants are navigating the implications of the continuity of the derivative and its role in the proof, with some exploring different definitions and approaches to the problem. The original poster's uncertainty about their proof's completeness highlights the complexity of the topic.

boboYO
Messages
106
Reaction score
0
Hi, I need some hints for the proof of the sard theorem in 1 dimension:

Prove that the set of critical values (f(x) where f'(x)=0) of continuously differentiable f:[a,b]->R has measure 0.My attempt:Fix [tex]\varepsilon[/tex], Let Crit(f) be the set of critical points of f. We want to show that f(Crit(f)) has measure 0.
Let [tex]C_k[/tex] be the set of all [tex]x[/tex] in Crit(f) such that [tex]|x-y|<\frac{1}{k} \implies |fx-fy|<\varepsilon|x-y|[/tex].

Clearly [tex]\bigcup_{k\in\mathbb{N}} C_k \supset \mathrm{Crit}(f)[/tex], so it suffices to show that [tex]f\left(\bigcup_{k\in\mathbb{N}} C_k \right)[/tex] has measure 0.

For any [tex]C_k[/tex], evenly split [tex][a,b][/tex] into intervals [tex]I_1,\dots,I_n[/tex] such that [tex]\frac{b-a}{n}<\frac{1}{k}[/tex].For each [tex]I_i[/tex], if there is a point of [tex]C_k[/tex] in it, say [tex]x[/tex], then

[tex]|fx-fy|<\varepsilon|x-y|[/tex] for all y in [tex]I_i[/tex].

Thus if [tex]I_i[/tex] contains a point of [tex]C_k[/tex], then [tex]f(I_i)[/tex] is contained in a open interval of at most [tex]2w \varepsilon[/tex] where w is the width of the interval.

Thus [tex]f(C_k)[/tex] is contained in an open set of length at most [tex]2w\varepsilon n =2(b-a) \varepsilon=K\varepsilon[/tex].

Since each [tex]f(C_k)[/tex] has measure bounded above by [tex]K \varepsilon[/tex], and [tex]C_{k}\subset C_{k+1}[/tex] for all k, then

[tex]\bigcup_{k\in\mathbb{N}} f(C_k) = f\left(\bigcup_{k\in\mathbb{N}} C_k \right)[/tex] is also bounded above by [tex]K\varepsilon[/tex]. Since [tex]\varepsilon[/tex] was arbitrary we are done.
--

I am not completely convinced because I did not use the hypothesis that the derivative is continuous. Did I make a mistake? Is there a simpler way do this?
 
Last edited:
Physics news on Phys.org
I'm really having a hard time following that. It's not even clear to me that you even used that f is differentiable. Why don't you define C(epsilon) to be the set of all x such that f'(x)<epsilon. Then Crit(f) is contained in C(epsilon). Can you show the measure of f(C(epsilon)) is less than or equal to epsilon*measure(C(epsilon))?
 
[tex](-\epsilon,\epsilon)[/tex] is open and f'(x) is continuous so [tex]C_\epsilon[/tex] is open too. An open set is a (at most) countable union of disjoint open intervals:

[tex]C_\epsilon=\bigcup_{i=1}^{\infty} I_i[/tex]

For every [tex]I_i[/tex], if x and y are 2 points in [tex]I_i[/tex],

[tex]|fx-fy|<\epsilon|x-y|< \epsilon\mu(I_i)[/tex], due to MVT. So [tex]\mu(f(I_i))<\epsilon\mu(I_i)[/tex]

Taking the union over i yields measure(f(Cε)) <ε*measure(Cε)<=ε(b-a) and we are done.

Thanks for your help :)
 
Last edited:
Now that I understand. And it even looks right.
 

Similar threads

  • · Replies 12 ·
Replies
12
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 10 ·
Replies
10
Views
3K
  • · Replies 9 ·
Replies
9
Views
10K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
Replies
18
Views
3K
Replies
4
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 11 ·
Replies
11
Views
2K