1
$\begingroup$

$f:\mathbb{R}^2\rightarrow \mathbb{R}, (x,y)\mapsto \begin{cases} \frac{\sin^4(x)\sin^4(y)}{x^2+y^2} &\text{if $(x,y) \neq (0,0)$ }\\0&\text{if $ (x,y)=(0,0)$}\end{cases}$

(i) Show that $f$ is differentiable on its domain of definition and calculate $f'(0,0)$

(ii) Determine the Gradient in $P=(\frac{\pi}{2},\frac{\pi}{2})$

(iii) Find a direction $\vec {v}$, whose Directional derivative $\frac{\partial f}{\partial \vec {v} } $ in $P$ is $0$

For (i) my idea was, to show that its partial derivative is continuous:

$f'(x,y) = \frac{2 \sin^3(x) \cos^4(y) (2 (x^2+y^2) \cos(x)-x \sin(x))}{(x^2+y^2)^2} \approx \frac{2 x^3 y^4 (2 (x^2+y^2) x-x^2)}{(x^2+y^2)^2} $ for small $x$ and $y$ is this correct? How to go on?

To (iii)

$ \langle \nabla f (\frac{\pi}{2},\frac{\pi}{2}), \vec {v} \rangle = \langle \begin {pmatrix} -\frac{\pi}{2} \\ -\frac{\pi}{2} \end{pmatrix}, \begin {pmatrix} v_1 \\ v_2 \end{pmatrix} \rangle = -\frac{\pi}{2}(v_1+v_2)=0 \Rightarrow v_1=v_2$ So $\vec {v}=(1,1)^T$ would be a solution. Is this correct?

$\endgroup$
1
  • $\begingroup$ In (i): "...and f'$(0,0)$" what? $\endgroup$
    – 2'5 9'2
    Commented Nov 16, 2014 at 23:16

1 Answer 1

1
$\begingroup$

about part 1. Find df/dx in (0,0) using the limits-definition. Actually using L'Hoital rule you can see that this derivative is 0 regardless of the value of y. For df/dy (0,0) is the same.

about part 2. Find the partial derivatives as a functions of x and y and find the functional values in x=Pi/2, and y=Pi/2

about part 3. consider cosW df/dx(Pi/2,Pi/2) + sinW df/dy(Pi/2, Pi/2) = 0 and find value for W.

$\endgroup$
9
  • 1
    $\begingroup$ For part 1, the existence of the partial derivatives doesn't necessarily imply differentiability. $\endgroup$
    – Teepeemm
    Commented Nov 17, 2014 at 0:39
  • $\begingroup$ Why is it allowed, to disregard $y$ and just consider $x$ using L'Hôpital's rule? $\endgroup$
    – fear.xD
    Commented Nov 17, 2014 at 1:03
  • $\begingroup$ Take a look at the limit lim x->0 (f(x,y)-f(0,y))/(x-0) and consider y is a constant ... $\endgroup$
    – mr-fotev
    Commented Nov 17, 2014 at 1:09
  • $\begingroup$ But this only shows partial derivation? It's still necessary to show, that its continuous $\endgroup$
    – fear.xD
    Commented Nov 17, 2014 at 1:29
  • $\begingroup$ It is true that the existence of the partial derivatives doesn't necessarily imply differentiability. But as soon as these partial derivatives are well defined and continuous in every neighbourhood of a given point, this implies differentiability. Consider the following example: f(x,y)={y^2/(x+y) if (x,y)=/= (0,0) and 0 if (x,y)=(0,0)}. Consider x->+0. Now df/dy = 1 - 1/(1+y/x)^2 and if y->-0 as -x then it's obvious that df/dy is not continuos in all neighbourhoods of (0,0). Therafore the function is not differentiable. $\endgroup$
    – mr-fotev
    Commented Nov 17, 2014 at 1:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .