Suppose $f:\mathbb{R} \rightarrow \mathbb{R}$ is everywhere differentiable and

  1. $\lim_{x \to \infty}f(x)=\lim_{x \to -\infty}f(x)=0$,
  2. there exists $c \in \mathbb{R}$ such that $f(c) \gt 0$.

Can we say anything about $\lim_{x \to \infty}f'(x)$ and $\lim_{x \to -\infty}f'(x)$?

I am tempted to say that $\lim_{x \to \infty}f'(x)$ = $\lim_{x \to -\infty}f'(x)=0$.

I started with the following, but I'm not sure this is the correct approach, $$\lim_{x \to \infty}f'(x)= \lim_{x \to \infty}\lim_{h \to 0}\frac{f(x+h)-f(x)}{h}.$$


No. Try $f(x)=\sin(x^a)/x$ for various values of $a$.


Key fact: on small scales, you can change the derivative of a function hugely without changing its values very much. I can introduce a tiny wiggle in a curve that is barely noticeable in size, but very sharp in derivative.

Throw in a few tiny wiggles arbitrarily far along a decaying function of your choice, and you'll get a fucntion which decays but whose derivative keeps spiking.

As a more concrete hint, consider a differentiable function $f$ such that $|f(x)| < 1$ for all $x$, and take $g(x) = \frac{1}{2}f(2x)$. Then $g(x)$ has $|g(y)| < \frac{1}{2}$ for all $y$, so $g$ takes smaller values than $f$, but $g'(z) = f'(2z)$, so the derivative of $g$ gets just as big as the derivative of $f$.

This shows how very small functions can nonetheless have very large derivatives.


To correct a incorrect attempt, let $f(x) = e^{-x^2} \cos(e^{x^4})$, so

$\begin{align}f'(x) &= e^{-x^2} 4 x^3 e^{x^4}(-\sin(e^{x^3})) -2x e^{-x^2} \cos(e^{x^3})\\ &= -4 x^3 e^{x^4-x^2} \sin(e^{x^3}) -2x e^{-x^2} \cos(e^{x^3})\\ \end{align} $

The $e^{x^4-x^2}$ term makes $f'(x)$ oscillate violently and unboundedly as $x \to \pm \infty$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.