Results 1 to 8 of 8

Math Help - Function Recovery from Partial Derivatives

  1. #1
    Junior Member
    Joined
    Apr 2008
    Posts
    31

    Function Recovery from Partial Derivatives

    Hello,

    If a function has the form u(x,y)=c, where c is a constant and the partial deriviates of u are:

    \frac{\partial u}{\partial x} = \frac{x^2}{y^2}-1-\frac{1}{y^2}
    \frac{\partial u}{\partial y} = \frac{2x}{y}

    Is it possible to find u(x,y)?

    Thank you
    Follow Math Help Forum on Facebook and Google+

  2. #2
    MHF Contributor
    Prove It's Avatar
    Joined
    Aug 2008
    Posts
    11,831
    Thanks
    1602
    Quote Originally Posted by sael View Post
    Hello,

    If a function has the form u(x,y)=c, where c is a constant and the partial deriviates of u are:
    \frac{\partial u}{\partial x} = \frac{x^2}{y^2}-1-\frac{1}{y^2}
    \frac{\partial u}{\partial y} = \frac{2x}{y}
    Is it possible to find u(x,y)?

    Thank you
    Absolutely.

    \frac{\partial u}{\partial x} = \frac{x^2}{y^2} - 1 - \frac{1}{y^2}

    u = \int{\frac{x^2}{y^2} - 1 - \frac{1}{y^2}\,dx}.

    Now remember that since you are taking the integral with respect to x, you keep y constant.

    u = \frac{x^3}{3y^2} - x - \frac{x}{y^2} + f(y).

    Do you see how instead of having a CONSTANT of integration, we have a FUNCTION of integration? This is because if you partially differentiate a function of y with respect to x, it becomes 0.


    You also have

    \frac{\partial u}{\partial y} = \frac{2x}{y}

    u = \int{\frac{2x}{y}\,dy}

    u = 2x\ln{|y|} + g(x).



    This is quite strange. Usually when you do the integrations, you get the same functions, only differing by a function of x and a function of y. Of course, by putting them together you get the entire function. Are you sure you copied down the question correctly?
    Follow Math Help Forum on Facebook and Google+

  3. #3
    Junior Member
    Joined
    Apr 2008
    Posts
    31
    Thank you, that makes sense

    I can't see how to get u without f(y) and g(x). Is it possible to do this?
    Follow Math Help Forum on Facebook and Google+

  4. #4
    MHF Contributor
    Prove It's Avatar
    Joined
    Aug 2008
    Posts
    11,831
    Thanks
    1602
    Like I said, I think there is something wrong with the partial derivatives you posted.

    When you do the integration with respect to x, you get all the functions of x, including the g(x) you were missing, and when you do the integration with respect to y, you get all the functions of y, including the f(y) you were missing.

    Also, the two integrations should give identical functions, with the only difference being that one has an extra function of x and the other has an extra function of y...
    Follow Math Help Forum on Facebook and Google+

  5. #5
    MHF Contributor

    Joined
    Apr 2005
    Posts
    16,441
    Thanks
    1862
    Here is the difficulty: if there exist one function u such that
    <br />
\frac{\partial u}{\partial x} = \frac{x^2}{y^2}-1-\frac{1}{y^2}<br />
    and
    <br />
\frac{\partial u}{\partial y} = \frac{2x}{y}<br />

    Then we must have
    \frac{\partial^2 u}{\partial y\partial x}= -\frac{2x^2}{y^3}+ \frac{2}{y^3}
    and
    \frac{\partial^2 u}{\partial x\partial y}= \frac{2}{y}

    But with continuous derivatives, those "mixed derivatives" must be the same.

    No, you cannot recover u- there is no u(x,y) that satisfies those equations.
    Follow Math Help Forum on Facebook and Google+

  6. #6
    Junior Member
    Joined
    Apr 2008
    Posts
    31
    It is part of a bigger problem. I thought I was just getting the function recovery wrong, today is the first time I tried such a thing.

    The problem is to find the Orthogonal Trajectories of the family of curves x^2+(y-c)^2=1+c^2, c is a constant.

    I then get c by itself:

    u(x,y)=\frac{x^2}{y}+y-\frac{1}{y}=2c

    According to my text the Orthogonal Trajectories v(x,y)=c^* can be obtained from:

    \frac{\partial u}{\partial x} = \frac{\partial v}{\partial y}
    \frac{\partial u}{\partial y} = -\frac{\partial v}{\partial x}

    Which results in my original post except I probably should have used v instead of u for the function

    I do have the text book solution for this problem but it doesn't use this method. It is has me in knots as to why this won't work.
    Follow Math Help Forum on Facebook and Google+

  7. #7
    MHF Contributor

    Joined
    Apr 2005
    Posts
    16,441
    Thanks
    1862
    No, those do NOT give the equations you have before. With that u,
    \frac{\partial u}{\partial x}= \frac{2x}{y}
    \frac{\partial u}{\partial y}= -\frac{x^2}{y}+ 1+ \frac{1}{y^2}

    Now, you want
    \frac{\partial v}{\partial y}= -\frac{x^2}{y^2}+ 1+ \frac{1}{y^2}
    -\frac{\partial v}{\partial x}= \frac{2x}{y}

    Looks to me like you have partials with respect to x and y confused.
    Follow Math Help Forum on Facebook and Google+

  8. #8
    Junior Member
    Joined
    Apr 2008
    Posts
    31
    I still get for v(x,y):

    \frac{\partial v}{\partial y}=\frac{\partial u}{\partial x}=\frac{2x}{y}
    \frac{\partial v}{\partial x}=-\frac{\partial u}{\partial y}=\frac{x^2}{y}-1-\frac{1}{y^2}
    Follow Math Help Forum on Facebook and Google+

Similar Math Help Forum Discussions

  1. Production Function (Partial Derivatives)
    Posted in the Calculus Forum
    Replies: 1
    Last Post: January 15th 2012, 02:24 AM
  2. Partial Derivatives Of Composite Function
    Posted in the Differential Geometry Forum
    Replies: 1
    Last Post: October 14th 2010, 05:07 AM
  3. Partial derivatives for the sign function
    Posted in the Calculus Forum
    Replies: 2
    Last Post: September 20th 2010, 08:05 AM
  4. Partial derivatives of implicit function
    Posted in the Calculus Forum
    Replies: 2
    Last Post: January 22nd 2010, 11:58 AM
  5. Increasing function, partial derivatives
    Posted in the Calculus Forum
    Replies: 2
    Last Post: July 13th 2009, 08:23 PM

Search Tags


/mathhelpforum @mathhelpforum