Does f'(0)=0 where x in [-1,1] and f(x)<=f(0)

  • Thread starter Thread starter HaLAA
  • Start date Start date
Join the discussion
Registration is free. Start your own thread to ask a follow-up.
7 replies · 2K views
HaLAA
Messages
85
Reaction score
0

Homework Statement


If the function f:ℝ→ℝ is differentiable and f(x)<=f(0) for all x ∈[-1,1], then f'(0)=0. True or False.

Homework Equations

The Attempt at a Solution


I think the statement is right. Since f(x)<= f(0) for all x in [-1,1], this tells us f is an even function or a symmetric function, then apply Rolle's theorem, we can prove the statement.
 
Physics news on Phys.org
The function is not necessarily even or symmetric. Consider f(x) that is ##x^3## for ##x\leq 0## and ##-x^2## for ##x>0##.
 
what is the sign of ##\frac{f(x) - f(0)}{x} ## for ##x## positive and negative ?
 
HaLAA said:

Homework Statement


If the function f:ℝ→ℝ is differentiable and f(x)<=f(0) for all x ∈[-1,1], then f'(0)=0. True or False.

Homework Equations

The Attempt at a Solution


I think the statement is right. Since f(x)<= f(0) for all x in [-1,1], this tells us f is an even function or a symmetric function, then apply Rolle's theorem, we can prove the statement.

Your function is differentiable on [-1,1] and has a maximum over [-1,1] at the point x = 0. What does that tell you?
 
  • Like
Likes   Reactions: geoffrey159
Ray Vickson said:
Your function is differentiable on [-1,1] and has a maximum over [-1,1] at the point x = 0. What does that tell you?
Ray Vickson said:
Your function is differentiable on [-1,1] and has a maximum over [-1,1] at the point x = 0. What does that tell you?
let h(x)=(f(x)-f(0))/x, when x>0, h(x)<=0; when x<0, h(x)=>0; f(x) is bounded above and the supremum is at x = 0, so f'(0)=0
 
andrewkirk said:
The function is not necessarily even or symmetric. Consider f(x) that is ##x^3## for ##x\leq 0## and ##-x^2## for ##x>0##.

At x =0, we still have f'(0)=0?
 
HaLAA said:
At x =0, we still have f'(0)=0?
Yes, the conclusion still holds. In the OP you reached the correct conclusion based on an incorrect argument. Ray's and Geoffrey's posts indicate the direction that a valid argument would take.
 
HaLAA said:
let h(x)=(f(x)-f(0))/x, when x>0, h(x)<=0; when x<0, h(x)=>0; f(x) is bounded above and the supremum is at x = 0, so f'(0)=0

Can you clearly explain the link between ##h(x)## and ##f'(0) ## ?EDIT: By the way, my hint is valid, but honestly, Mr Vickson's hint is more elegant
 
Last edited: