Trying to assimilate the meaning of the differential I have looked for different examples of functions which:
- Admits all directional derivatives but are not continuous $(f: \mathbb{R}^2 \rightarrow \mathbb{R} \quad ,\quad (x,y) \mapsto \begin{cases} 0 & \text{for } (x,y)=(0,0) \\ \frac{xy^4}{x^4+y^8} & \text{for } (x,y) \neq (0,0) \end{cases})$
- Admits all directional derivatives and are continuous but not differentiable:$f$ not differentiable at $(0,0)$ but all directional derivatives exist
For what I am looking for now I will use:
Suppose we have a function $f: \mathbb{R}^n \rightarrow \mathbb{R}^m$
that in a point, let say $x_0$, admits all direccional derivatives. Then, we can consider the map
$\phi_{x_{0}}: \mathbb{R}^n \rightarrow \mathbb{R}^m \quad$ $v \mapsto D_{v}f(x_{0}) \quad$ , which sends a vector to its directional derivative in $x_0$.
If I am not mistaken, if this map is linear, it will be the differential of $f$. What I am searching now is:
A function $f: \mathbb{R}^n \rightarrow \mathbb{R}^m$, continuos at $x_0$, which admits all directional derivatives at this point, but in which $\phi_{x_{0}}$ is not continuous.
I don´t know if this would be possible. In the example 2, the function $\phi_{(0,0)}$ is smooth (continuous, although non linear) and I can´t imagine how you could "break" the smoothness without breaking the continuity of $f$ too (that happens in the example 1). I would appreciate if the example is in 2 dimensions ($f: \mathbb{R}^2 \rightarrow \mathbb{R}$) for it to be possible to visualize.
Thanks in advance