Skip to main content
Logo image

Section 4.4 Change of Variable

While Cartesian, cylindrical, and spherical coordinates enable integration of some functions over some intervals, they are not convenient for all functions and all intervals. Sometimes it is possible to adjust the function or interval using a carefully chosen change of variable.

Subsection 4.4.1 Review

Activity 64.

The purpose of this activity is to review the method of u-substitution which is the simplest form of change of variable in single variable calculus.
First integrate the following.
(a)
\(\int_0^3 e^{3x} \; dx \)
(b)
\(\int_0^1 \sin(x^2)2x \; dx \)
(c)
\(\int_0^{\pi/2} \sin^{57}\theta \cos\theta \; d\theta \)
Note that your substitutions are all smooth functions.

Subsection 4.4.2 Illustration

As a motivation and method of visualization of change of variable we will consider the effect the change has on an image.
A function that maps points to points is called a planar transformation. Most graphic effects you use in photo editing software fall in this category (e.g., scaling, reflection, translation, fish eye lens). Following is one example. Consider for \(P=(x,y),\) \(C(P)=\frac{1}{x^2+y^2}(x,y).\) Figure 4.4.1 shows the original and transformed versions of an image.
Figure 4.4.1. Transformation Example
The following examples will lead to an understanding of how the transformations works and when they are useful.
Figure 4.4.2. Grid for Transformations
Table 4.4.3. Example Table of Transformed Points
\((x,y)\) \(u_1(x,y)\) \(v_1(x,y)\) \((u,v)\)
\((0,0)\)
\((1,0)\)
\((2,0)\)
\((0,1)\) \(-1\) \(0\) \((-1,0)\)
\((1,1)\)
\((2,1)\)
\((0,2)\)
\((1,2)\)
\((2,2)\)

Activity 65.

The purpose of this activity is to notice a connection between the area of a square before and after it is transformed. The connection is based on the partials of the transformation.
  • \(u_1(x,y)=-y\) and \(v_1(x,y)=x\)
  • \(u_2(x,y)=2x\) and \(v_2(x,y)=3y.\)
  • \(u_3(x,y)=x\) and \(v_3(x,y)=y+x^2.\)
Use the grid in Figure 4.4.2 for these problems. For each of the transformations above complete the following steps.
(a)
Map each of the nine labeled points \((x,y)\) to the transformed points \((u,v).\)
(b)
Sketch the transformed points (on \(u\) and \(v\) axes).
(c)
Describe the effect of the transformation on shape and size.
(d)
Calculate the area of the original square.
(e)
For each transformation calculate the area of the new shape.
(f)
Compute \(\left[ \begin{array}{rr} \frac{\partial u}{\partial x} & \frac{\partial v}{\partial x} \\ \frac{\partial u}{\partial y} & \frac{\partial v}{\partial y} \end{array} \right]\)
(g)
Compute \(D=\frac{\partial u}{\partial x}\frac{\partial v}{\partial y} - \frac{\partial u}{\partial y}\frac{\partial v}{\partial x}. \)
(h)
Compare the area of the original square, the area of the transformed square, and the value \(D.\)

Definition 4.4.4. Jacobian.

For a transformation \(T\) given by \(x=t_1(u,v)\) and \(y=t_2(u,v)\) the Jacobian is
\begin{equation*} \left| \begin{array}{rr} \frac{\partial x}{\partial u} & \frac{\partial x}{\partial v} \\ \frac{\partial y}{\partial u} & \frac{\partial y}{\partial v} \\ \end{array} \right| = \left| \frac{\partial x}{\partial u}\frac{\partial y}{\partial v}-\frac{\partial x}{\partial v}\frac{\partial y}{\partial u} \right|. \end{equation*}

Activity 66.

The purpose of this activity is to determine the effect of order on the Jacobian.
Consider \(x(u,v)=\frac{1}{2}(u+v)\) and \(y(u,v)=\frac{1}{2}(u-v).\)
(a)
Calculate \(\left| \begin{array}{rr} \frac{\partial x}{\partial u} & \frac{\partial x}{\partial v} \\ \frac{\partial y}{\partial u} & \frac{\partial y}{\partial v} \\ \end{array} \right|.\)
(b)
Reading carefully, calculate \(\left| \begin{array}{rr} \frac{\partial x}{\partial v} & \frac{\partial x}{\partial u} \\ \frac{\partial y}{\partial v} & \frac{\partial y}{\partial u} \\ \end{array} \right| \)
(c)
Reading carefully, calculate \(\left| \begin{array}{rr} \frac{\partial y}{\partial u} & \frac{\partial y}{\partial v} \\ \frac{\partial x}{\partial u} & \frac{\partial x}{\partial v} \\ \end{array} \right|.\)
(d)
Compare the results.
(e)
Can any transformation make area negative?

Subsection 4.4.3 Method

For any smooth transformation \(x(u,v),y(u,v),\) the double integral over a region \(S\) is given by
\begin{equation*} \dint_S f(x(u,v),y(u,v))\left| \begin{array}{rr} \frac{\partial x}{\partial u} & \frac{\partial x}{\partial v} \\ \frac{\partial y}{\partial u} & \frac{\partial y}{\partial v} \\ \end{array} \right| \: du \: dv \end{equation*}

Example 4.4.5.

Integrate \(f(x,y)=3x+4y\) over the region with vertices with coordinates (2,1), (10,5), (13,14), and (5,10) using the transformations \(u(x,y)=\frac{3}{5}x-\frac{1}{5}y\) and \(v(x,y)=-\frac{1}{5}x+\frac{2}{5}y\text{.}\)
For the Jacobian and the substitution we need to have \(x(u,v)\) and \(y(u,v)\text{.}\) Our first step is to invert the transformation provided. We do this by solving the system of equations for \(x\) and \(y\text{.}\) We can do this by our favorite method of solving systems of linear equations (Gaussian elimination or substitution both work). The result is \(x(u,v)=2u+v\) and \(y(u,v)=u+3v\text{.}\)
Our second step is to calculate the Jacobian.
\begin{align*} \frac{\partial x}{\partial u} = & 2.\\ \frac{\partial x}{\partial v} = & 1.\\ \frac{\partial y}{\partial u} = & 1.\\ \frac{\partial y}{\partial v} = & 3.\\ \left| \begin{array}{rr} \frac{\partial x}{\partial u} & \frac{\partial x}{\partial v} \\ \frac{\partial y}{\partial u} & \frac{\partial y}{\partial v} \\ \end{array} \right| = & (2)(3)-(1)(1)\\ = & 5. \end{align*}
Our third step is to substitute the transformation into the function.
\begin{align*} f(x(u,v),y(u,v)) = & 3(2u+v)+4(u+3v) \\ = & 10u+15v. \end{align*}
The fourth step before evaluating the integral is to find the region over which we are integrating. For this we must transform each of the vertices of the region which is given in terms of \(u\) and \(v\text{.}\)
\begin{align*} u(2,1) = & \frac{3}{5}(2)-\frac{1}{5}(1) & = 1.\\ v(2,1) = & -\frac{1}{5}(2)+\frac{2}{5}(1) & = 0.\\ u(10,5) = & \frac{3}{5}(10)-\frac{1}{5}(5) & = 5.\\ v(10,5) = & -\frac{1}{5}(10)+\frac{2}{5}(5) & = 0.\\ u(13,14) = & \frac{3}{5}(13)-\frac{1}{5}(14) & = 5.\\ v(13,14) = & -\frac{1}{5}(13)+\frac{2}{5}(14) & = 3.\\ u(5,10) = & \frac{3}{5}(5)-\frac{1}{5}(10) & = 1.\\ v(5,10) = & -\frac{1}{5}(5)+\frac{2}{5}(10) & = 3. \end{align*}
Thus the regions vertices in terms of \(u\) and \(v\) are (1,0), (5,0), (5,3), and (1,3). Notice that this is a rectangular region which we can use directly in the setup of the integral.
Finally we setup and evaluate the integral
\begin{align*} \int_0^3 \int_1^5 (10u+15v)5 \; du \; dv = &\\ 5 \int_0^3 \int_1^5 (10u+15v) \; du \; dv = &\\ 5 \int_0^3 5u^2+15vu |_1^5 dv = &\\ 5 \int_0^3 125+75v-5+15v dv = &\\ 5 \int_0^3 120+90v dv = &\\ 5 \left( 120v+45v^2 |_0^3 \right) = & 3825. \end{align*}
For any smooth transformation \(x(u,v,w),\) \(y(u,v,w),\) \(z(u,v,w),\) the triple integral over a region \(S\) is given by
\begin{equation*} \tint_S f(x(u,v,w),y(u,v,w),z(u,v,w))\left|\frac{\partial(x,y,z)}{\partial(u,v,w)}\right| \: du \: dv \: dw \end{equation*}
where \(\left|\frac{\partial(x,y,z)}{\partial(u,v,w)}\right| \) is the determinant of the \(3 \times 3 \) gradient matrix (Jacobian) given below.
\begin{align*} \left| \begin{array}{rrr} \frac{\partial x}{\partial u} & \frac{\partial x}{\partial v} & \frac{\partial x}{\partial w} \\ \frac{\partial y}{\partial u} & \frac{\partial y}{\partial v} & \frac{\partial y}{\partial w} \\ \frac{\partial z}{\partial u} & \frac{\partial z}{\partial v} & \frac{\partial z}{\partial w} \end{array} \right| = \\ \left| \frac{\partial x}{\partial u}\left[\left(\frac{\partial y}{\partial v}\right)\left(\frac{\partial z}{\partial w}\right)-\left(\frac{\partial z}{\partial v}\right)\left(\frac{\partial y}{\partial w}\right)\right] \right. \\ -\frac{\partial x}{\partial v}\left[\left(\frac{\partial y}{\partial u}\right)\left(\frac{\partial z}{\partial w}\right)-\left(\frac{\partial z}{\partial u}\right)\left(\frac{\partial y}{\partial w}\right)\right]\\ \left. +\frac{\partial x}{\partial w}\left[\left(\frac{\partial y}{\partial u}\right)\left(\frac{\partial z}{\partial v}\right)-\left(\frac{\partial z}{\partial u}\right)\left(\frac{\partial y}{\partial v}\right)\right] \right| \end{align*}

Exercises 4.4.4 Exercises

Integrate the following functions using the specified change of coordinates and intervals.

1.

\(f(x,y)=x+y.\) \(r \in [0,1]\) and \(\theta \in [0,2\pi]\) with transformation \(x(r,\theta)=r\cos\theta\) and \(y(r,\theta)=r\sin\theta.\)

2.

\(f(x,y,z)=x+y+z.\) \(x(\rho,\phi,\theta)=\rho\cos\theta\sin\phi,\) \(y(\rho,\phi,\theta)=\rho\sin\theta\sin\phi,\) \(z(\rho,\phi,\theta)=\rho\cos\phi\) for \(\rho \in [0,1],\) \(\phi \in [0,\pi],\) and \(\theta \in [0,2\pi].\)

3.

\(f(x,y)=xy \) over the convex region with vertices \((1,1),\) \((1,-5),\) \((9,1),\) and \((11,-5)\) using the transformation \(x=1+2v\) and \(y=1-3u.\)

4.

\(\int_0^1 \int_0^{\sqrt{1-y^2}} e^{x^2+y^2} \; dx \: dy \) using a useful transformation of your choosing.