0

Note: Please do not give a solution; I am curious to understand why my solution is incorrect, and would prefer guidance to help me complete the question myself. Thank you.


Problem

Show that if $f$ is an infinitely times differentiable function on $\mathbb{R}$ such that for every $x\in\mathbb{R}$ there exists $n\geq 1$ with $f^{(n)}(x)=0$ then there exists $N\geq 1$ such that for every $x\in\mathbb{R}$ we have $f^{(N)}(x)=0$.

My Proof

My proof relies on Baire's Theorem. However I am unable to show that $C^{\infty}(\mathbb{R})$ is a metric space (I would appreciate some guidance in the form of a hint on this). Hence, for the time being I have assumed this to be true and continued with the proof as follows.

For every $x\in\mathbb{R}$ define the set $O_{n}:=\{f^{(n)}(x)\in C^{\infty}(\mathbb{R})|f^{(n)}(x)\neq 0\text{ and }f^{(n)}(x)\rightarrow_{n\rightarrow\infty}0\}$. Then every $O_{n}$ is open and dense in $C^{\infty}(\mathbb{R})$. Hence, by Baire's Theorem, $\bigcap O_{n}$ is dense in $C^{\infty}(\mathbb{R})$. Now we define $N:=\inf\{n\geq 1|f^{(n)}(x)=0, \forall f^{(n)}(x)\in\bigcap O_{n}\}$ so that $f^{(N)}(x)=0$ for every $x\in\mathbb{R}$.

Second Attempt

Assume $f\in\mathbb{C}^{\infty}$ then for every $x\in\mathbb{R}$ there exists $n\geq 1$ such that $f^{(n)}(x)=0$. Then for every $n\geq 1$ define, \begin{align} X_{n}:=\{x\in\mathbb{R}|f^{(n)}(x)=0\}. \end{align} Since $\mathbb{R}$ is a complete metric space, Baire's Theorem tells us there exists $N\in\mathbb{N}$ such that $X_{N}=\{x\in\mathbb{R}|f^{(N)}(x)=0\}$ is non-empty.

This is great progress, however I am not sure where to continue from here. How do I extend $X_{N}$ so that $X_{N}=\mathbb{R}$?

Zeta-Squared
  • 915
  • 5
  • 10
  • Apply Baire Category Theorem to the real line: $\mathbb R =\cup_n {x:f^{n}(x)=0}$. – Kavi Rama Murthy Mar 18 '19 at 06:05
  • what part of my proof is this hint in relation to? – Zeta-Squared Mar 18 '19 at 06:09
  • You are making things complicated by trying to define a metric on $C^{\infty} (\mathbb R)$. I have just given the first step in the proof. You can conclude that $f$ is a polynomial in some open interval. More generally you can show that every point has a neighborhood on which $f$ is a polynomial. Can you proceed? – Kavi Rama Murthy Mar 18 '19 at 06:13
  • Then $O_n$ you define is equal to $\mathbb R$, is not necessarily dense in $\mathbb R$. – Yiorgos S. Smyrlis Mar 18 '19 at 06:13
  • 1
    You are talking about open sets, dense sets etc without defining a metric or a topology and I am suggesting that you should completely revise the proof. – Kavi Rama Murthy Mar 18 '19 at 06:15
  • I won't give a solution, but a location of one: https://math.stackexchange.com/questions/232665/baire-show-that-f-colon-mathbbr-to-mathbbr-is-a-polynomial-in-an-open-bo also https://math.stackexchange.com/questions/1698792/show-that-f-is-a-polynomial – Gerry Myerson Mar 18 '19 at 08:51

0 Answers0