4

If $f$ is continuous on [$a,b$] then the area function $A(x)=\int_{a}^{x} f(t)dt,$ for $a \le x \le b$, is continuous on [$a,b$] and differntiable on $(a,b)$.

My question is: why is it not differntiable on [$a,b$]? Why is it an open interval and not a closed interval?

Jason
  • 3,563

1 Answers1

4

The limit of the difference quotient has to exist for approaches from arbitrary directions (in $\mathbb{R}$ the limit approaching from above as well as from below). For boundary points in closed sets, this fails for at least one direction.

Take $f: [0,\infty) \rightarrow [0,\infty),\,f(x) = x$. One would immediately say that $f$ is differentiable on $[0,\infty)$ and $f'(x) = 1 \hspace{3mm} \forall x \in [0,\infty)$. But, in fact, if you extend $f$ to

$$f(x) =\left\{\begin{array}{lll} x & x > 0 \\ 0 & x=0 \\ -x & x < 0 \end{array}\right.$$

you get f(x) = |x| which is not differentiable at $x=0$.

root
  • 1,134
  • I think it makes more sense in our usual coordinate system to say "from the right as well as from the left". Still, +1. – Ian Aug 27 '15 at 23:34
  • One can define derivative at the interval boundary. It is just a matter of taste, if one does it not. There is no must. – user251257 Aug 28 '15 at 00:26