Proof that f is Constant on the Nonnegative Reals

  • Context: Graduate 
  • Thread starter Thread starter SiddharthM
  • Start date Start date
Click For Summary

Discussion Overview

The discussion revolves around the properties of a continuous function \( f \) defined on the nonnegative reals, particularly focusing on whether \( f \) can be proven to be constant under certain conditions involving a sequence of functions \( f_n(t) = f(nt) \) that are equicontinuous on the interval [0,1]. The scope includes mathematical reasoning and proof exploration.

Discussion Character

  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant proposes that if \( f_n \) is equicontinuous on [0,1], then \( f \) must be constant on the nonnegative reals, using a proof based on the definition of equicontinuity and the behavior of sequences approaching zero.
  • Another participant asserts that \( f \) is constant on [0,\infty) by examining the value at \( f(0) \) and showing that for any \( c \) in [0,\infty), the distance \( |f(0) - f(c)| \) can be made arbitrarily small.
  • A later reply acknowledges a misunderstanding regarding the equicontinuity condition, clarifying that the proof only establishes that \( f \) is constant on [0,\infty) rather than on all of [0,1].
  • Another participant corrects earlier statements by asserting that the condition of \( f \) being constant on [0,\infty) is both necessary and sufficient for the equicontinuity of \( f_n \) on [0,1], regardless of the behavior of \( f \) on negative reals.

Areas of Agreement / Disagreement

Participants express differing views on the implications of the equicontinuity condition and the behavior of \( f \) on negative reals. There is no consensus on the broader implications of the findings, and the discussion remains unresolved regarding the necessity of the equicontinuity condition.

Contextual Notes

Some participants note the importance of the equicontinuity condition being restricted to [0,1] and the implications this has on the conclusions drawn about the function \( f \). There is also an acknowledgment of the potential for misunderstanding the scope of the proof.

SiddharthM
Messages
176
Reaction score
0
say a function f is continuous on the reals and there exists a sequence of functions so that f_n (t)=f(nt) for n=1,2,3... also the sequence of functions is equicontinuous on [0,1].

what does show about f? I show f is constant on the nonnegative reals

Let epsilon (from here onwards we shall call it e)>0, then because f_n is equicontinuous there exists a delta>0 s.t. d(x,y)<delta implies |f_n(x)-f_n(y)|<e for all n as long as x,y in [0,1]. Now fix any two positive numbers x,y. since x/n and y/n go to zero there exists a N so that |x/n-y/n|<delta for all n>N and that they are both in [0,1]. Then this means that |f_n(x/n)-f_n(y/n)|<e for all n>N. This means |f(x)-f(y)|<e. Since epsilon is arbitrary we see that f(x)=f(y).

I feel like there should be more to this or maybe that I'm wrong altogether.

What examples satisfy the hypothesis?
 
Last edited:
Physics news on Phys.org
f must be constant on [tex][0,\infty)[/tex]. The way I proved it was by looking at f(0) = f_n(0). Fix any c in [tex][0,\infty)[/tex], and e > 0. Find d > 0 such that |x-y|<d implies |f_n(x)-f_n(y)| < e for all n, x,y in [0,1]. Then choose n such that |c/n| < min{d,1}. Then |f(0)-f(c)| = |f_n(0)-f_n(c/n)| < e. It follows that for any c in [tex][0,\infty)[/tex], and any e > 0, |f(0)-f(c)| < e. Hence f is constant on [tex][0,\infty)[/tex].

(edit: changed the proof to restrict c to [tex][0,\infty)[/tex]).
 
Last edited:
Actually, I didn't notice the restriction "equicontinuous on [0,1]"... when I did this problem I must have glanced over that.. I assumed f_n was equicontinuous on R.

So all I really proved was that f is constant on [tex][0,\infty)[/tex].
 
correction

And after a second look, the condition that f is constant on [tex][0,\infty)[/tex] is necessary and sufficient to satisfy the hypothesis.

That is, you can show that as long as f is constant on [tex][0,\infty)[/tex], regardless of what f does on [tex](-\infty,0)[/tex] (even not continuous), then [tex]f_n(x)=f(nx)[/tex] will be equicontinuous on [0,1]. This is actually easy to see: [tex]f_n(x)=f(nx)=C[/tex] for all x in [0,1], hence it's obviously equicontinuous there.
 
Last edited:

Similar threads

  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 11 ·
Replies
11
Views
2K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 24 ·
Replies
24
Views
7K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 11 ·
Replies
11
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 18 ·
Replies
18
Views
3K