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!

Change of variables

  1. Feb 3, 2006 #1
    Hi, I would like some help with the following question.

    Q. Let f be continuous on [0,1] and let R be the triangular region with vertices (0,0), (1,0) and (0,1). Show that:

    \int\limits_{}^{} {\int\limits_R^{} {f\left( {x + y} \right)} dA = \int\limits_0^1 {uf\left( u \right)} } du

    By making the substitutions u = x+y and v = y, I got it down to:

    \int\limits_{}^{} {\int\limits_R^{} {f\left( {x + y} \right)} dA = } \int\limits_0^1 {\int\limits_0^u {f\left( u \right)} dvdu}

    The above leads to the given result. However, I was stuck on trying to get bounds for the integrals so I'm not sure if I've justified those limits of integration properly.

    From the boundary line y = 1 - x, the substitution u = x+y gives u = 1. From the boundary line x = 0, the substitutions yield u = 0 + y = v so that v = u. The boundary line y = 0 yields v = 0. The lower limit for u is what I am lacking in. The three boundaries that I've just obtained completely describe the region R anyway so I decided to say that at the origin x,y = 0 which gives u = 0. I'm not sure whether I should've done something else to obtain the lower u limit.

    Any help would be good.
    Last edited: Feb 3, 2006
  2. jcsd
  3. Feb 3, 2006 #2


    User Avatar
    Homework Helper

    If the original bounds are 0<=y<=1 and 0<=x<=1-y

    then for u=x+y and v=y we have

    0<=x<=1-y ==> y<=x+y<=1 ==> v<=u<=1

    and, of chourse 0<=v<=1
  4. Feb 3, 2006 #3


    User Avatar
    Homework Helper

    A better way: (take all inequalities as inclusive)

    Let L1: x+y=1, 0<y<1 ==> L1': u=1, 0<v<1
    Let L2: y=0, 0<x<1 ==> L2': v=0, 0<u-v<1, but v=0, so 0<u-0<1 or just 0<u<1
    Let L3: x=0, 0<y<1 ==> L3': u-v=0 i.e. u=v, 0<v<1
  5. Feb 3, 2006 #4


    User Avatar
    Homework Helper

    So the transformed region is /| = a right triangle in the uv-plane formed by cutting the unit square along u=v, lower triangle
  6. Feb 3, 2006 #5
    Thanks for the help Benorin. I just thought of another way to justify setting u = 0 as a lower bound for the u integral. I've already established 3boundary lines which already account for the 'shape' of the original region in the xy plane. From a quick sketch I can see that adding the line y = - x as a boundary line still gives the same region so y = - x => y + x = 0 = u. But that's kind of a fudge method, your method is the right one.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Change of variables
  1. Change of Variables (Replies: 0)

  2. Change of Variables (Replies: 4)

  3. Change in variables (Replies: 7)

  4. Change of variable (Replies: 5)

  5. Change of Variables (Replies: 1)