1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Partial Derivative, piecewise function

  1. Oct 22, 2008 #1
    1. The problem statement, all variables and given/known data
    Let :
    [tex]f(x,y) = \frac{xy(x^2 - y^2)}{(x^2 + y^2)^2}[/tex] if [tex](x,y) \neq (0,0)[/tex]
    [tex]f(x,y) = 0[/tex] if [tex](x,y) = (0,0)[/tex]

    a) Find [tex]f_{xx}(0,0)[/tex]

    b) Find [tex]f_{xy}(0,0)[/tex]

    c) Find [tex]f_{yx}(0,0)[/tex]

    2. Relevant equations

    3. The attempt at a solution
    I'm not sure how to deal with the piecewise defined function, I just took the first and second partial derivatives w/ respect to x (for part a), but then still I can't evaluate the resulting function at (0,0), because at the point it doesn't exist...
  2. jcsd
  3. Oct 22, 2008 #2


    Staff: Mentor

    We can't tell whether what you got is correct because you didn't tell us what you got.
  4. Oct 23, 2008 #3


    User Avatar
    Science Advisor

    That's not really piecewise. f(0, 0) is defined to be 0 in order to make f continuous and so differentiable. You can do the entire problem just ignoring f(0,0).
  5. Dec 31, 2008 #4
    I have exactly the same question. The equation is even more or less the same. Mine is:

    [tex]f(x,y)=\frac{xy(x^2-y^2)}{x^2+y^2}[/tex] if [tex](x,y) \neq (0,0)[/tex]
    [tex]f(x,y)=0[/tex] if [tex](x,y)=(0,0)[/tex].

    But I'm still puzzled how to threat the point (0,0). According to my textbook, [tex]f_{xy}(0,0)\neq f_{yx}(0,0)[/tex]. So, I know more or less the solution, but I'm not sure that my explanation is correct.

    Ignoring the point (0,0) I have:

    first x, then y:


    first y, then x:


    So, it is obvious that as long [tex]x\neq 0[/tex] and [tex] y\neq 0[/tex] that [tex]\frac{\partial{f}}{\partial{y}\partial{x}}(x,y)=\frac{\partial{f}}{\partial{x}\partial{y}}(x,y)[/tex].

    Now, for (x,y)->(0,0), how do I prove that
    [tex]\frac{\partial{f}}{\partial{y}\partial{x}}(0,0)\neq\frac{\partial{f}}{\partial{x}\partial{y}}(0,0)[/tex]? Clearly, just substituting x and y with 0 results in [tex]\frac{0}{0}[/tex].

    Is it valid to say the following:
    For[tex]\frac{\partial{f}}{\partial{y}\partial{x}}(0,0)[/tex], first let x-> 0 first, resulting in [tex]\frac{\partial{f}}{\partial{y}\partial{x}}(0,y)=\frac{-y^6}{y^6}=-1, \forall y[/tex] and thus[tex]\frac{\partial{f}}{\partial{y}\partial{x}}(0,0)=-1[/tex]?

    For[tex]\frac{\partial{f}}{\partial{x}\partial{y}}(0,0)[/tex], first let y-> 0 first, resulting in [tex]\frac{\partial{f}}{\partial{x}\partial{y}}(x,0)=\frac{x^6}{x^6}=1, \forall x[/tex] and thus[tex]\frac{\partial{f}}{\partial{x}\partial{y}}(0,0)=1[/tex]?

    This is more or less a guess. I have no clue whether it is correct and if it is correct how to justify this. Especially, I have no explanation why for [tex]\frac{\partial{f}}{\partial{y}\partial{x}}(0,0)[/tex], I first let x->0 and not y->0.

    Any help on this would be most welcome and will be highly appreciated.

    (and happy new year to you all)
  6. Dec 31, 2008 #5


    User Avatar
    Science Advisor

    The partial derivatives are defined by the limits
    [tex]f_x(x_0,y_0)= \lim_{h\rightarrow 0} \frac{f(x_0+h, y_0)- f(x_0,y_0)}{h}[/tex]
    [tex]f_y(x_0,y_0)= \lim_{h\rightarrow 0} \frac{f(x_0,y_0+h)- f(x_0,y_0)}{h}[/tex]
    Here, [itex](x_0,y_0)= (0,0)[/itex] and with h non-zero, in calculating the limit, we are not at (0,0) and so use the other formula.

    [tex]f_x(0,0)= \lim_{h\rightarrow 0} \frac{\frac{(x+h)(0)((x+h)^2- 0^2)}{(x+h)^2- 0^2)- 0}}{h}= \lim_{h\rightarrow 0} 0/h= 0[/itex]
    [tex]f_y(0,0)= \lim_{h\rightarrow 0} \frac{\frac{(0)(y+h)(0^2- (y+h)^2)}{0^2-(y+h)^2- 0}}{h}= \lim_{h\rightarrow 0} 0/h= 0[/itex]
    While [itex]f_x[/itex] and [itex]f_y[/itex] for (x,y) not equal to (0,0) is given by differentiating f in the usual way.

    Once you know [itex]f_x[/itex] and [itex]f_y[/itex] for (x,y) close to (0,0) do exactly the same thing with this new function to find the second derivative.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook