The Implicit Function Theorem

The Implicit Function Theorem for One Equation

Theorem1

The implicit function existence theorem: suppose the function F(x,y)F(x, y) satisfies:

  1. it has continuous first-order partial derivatives in some neighbourhood of the point (x0,y0)(x_0, y_0);
  2. F(x0,y0)=0F(x_0, y_0) = 0;
  3. Fy(x0,y0)≠0F_y(x_0, y_0) \neq 0.

Then in some neighbourhood of the point (x0,y0)(x_0, y_0), the equation F(x,y)=0F(x, y) = 0 uniquely determines a function y=f(x)y = f(x) with continuous derivative, satisfying y0=f(x0)y_0 = f(x_0), and

dydx=−FxFy\frac{dy}{dx} = -\frac{F_x}{F_y}

The Implicit Function of Three Variables

Similarly, the equation F(x,y,z)=0F(x, y, z) = 0 determines an implicit function z=f(x,y)z = f(x,y) when Fz≠0F_z \neq 0, and

∂z∂x=−FxFz,∂z∂y=−FyFz\frac{\partial z}{\partial x} = -\frac{F_x}{F_z}, \quad \frac{\partial z}{\partial y} = -\frac{F_y}{F_z}

The Implicit Function Theorem for a System of Equations

Theorem2

Suppose the system of equations

{F(x,y,u,v)=0G(x,y,u,v)=0\begin{cases} F(x, y, u, v) = 0 \\ G(x, y, u, v) = 0 \end{cases}

satisfies, in some neighbourhood of the point (x0,y0,u0,v0)(x_0, y_0, u_0, v_0):

  1. F,GF, G have continuous first-order partial derivatives;
  2. F(x0,y0,u0,v0)=0F(x_0,y_0,u_0,v_0)=0 and G(x0,y0,u0,v0)=0G(x_0,y_0,u_0,v_0)=0;
  3. the Jacobian determinant J=∂(F,G)∂(u,v)=∣FuFvGuGv∣≠0J = \frac{\partial(F,G)}{\partial(u,v)} = \begin{vmatrix} F_u & F_v \\ G_u & G_v \end{vmatrix} \neq 0.

Then the system uniquely determines, in a neighbourhood of that point, a set of functions u=u(x,y),v=v(x,y)u=u(x,y), v=v(x,y) with continuous partial derivatives, satisfying u0=u(x0,y0)u_0=u(x_0,y_0) and v0=v(x0,y0)v_0=v(x_0,y_0).

Differentiating a System of Implicit Functions

Take the partial derivative of both sides of the system with respect to xx (treating u,vu,v as functions of x,yx,y):

{Fx+Fu∂u∂x+Fv∂v∂x=0Gx+Gu∂u∂x+Gv∂v∂x=0\begin{cases} F_x + F_u \frac{\partial u}{\partial x} + F_v \frac{\partial v}{\partial x} = 0 \\ G_x + G_u \frac{\partial u}{\partial x} + G_v \frac{\partial v}{\partial x} = 0 \end{cases}

This is a linear system in ∂u∂x,∂v∂x\frac{\partial u}{\partial x}, \frac{\partial v}{\partial x}, and by Cramer’s rule:

∂u∂x=−1J∂(F,G)∂(x,v),∂v∂x=−1J∂(F,G)∂(u,x)\frac{\partial u}{\partial x} = -\frac{1}{J} \frac{\partial(F,G)}{\partial(x,v)}, \quad \frac{\partial v}{\partial x} = -\frac{1}{J} \frac{\partial(F,G)}{\partial(u,x)}

Similarly one obtains ∂u∂y\frac{\partial u}{\partial y} and ∂v∂y\frac{\partial v}{\partial y}.

The Inverse Function Theorem

Theorem3

The inverse function theorem: suppose the vector-valued function y⃗=f⃗(x⃗)\vec{y} = \vec{f}(\vec{x}) (that is, yi=fi(x1,…,xn)y_i = f_i(x_1,\ldots,x_n), i=1,…,ni=1,\ldots,n) has continuous first-order partial derivatives in some neighbourhood of the point x⃗0\vec{x}_0, and the Jacobian determinant

J=∂(f1,…,fn)∂(x1,…,xn)≠0J = \frac{\partial(f_1,\ldots,f_n)}{\partial(x_1,\ldots,x_n)} \neq 0

Then in some neighbourhood of y⃗0=f⃗(x⃗0)\vec{y}_0 = \vec{f}(\vec{x}_0) there exists a unique inverse function x⃗=f⃗−1(y⃗)\vec{x} = \vec{f}^{-1}(\vec{y}) with continuous partial derivatives, and

∂(x1,…,xn)∂(y1,…,yn)=1∂(y1,…,yn)∂(x1,…,xn)\frac{\partial(x_1,\ldots,x_n)}{\partial(y_1,\ldots,y_n)} = \frac{1}{\frac{\partial(y_1,\ldots,y_n)}{\partial(x_1,\ldots,x_n)}}

that is, the Jacobian determinant of the inverse function is the reciprocal of that of the original function.

Worked Examples

Example 1: Differentiating an Implicit Function

Let z=z(x,y)z = z(x,y) be determined by the equation x2+y2+z2=4zx^2 + y^2 + z^2 = 4z. Find ∂2z∂x2\frac{\partial^2 z}{\partial x^2}.

an implicit functionthe second-order partial derivativesdifferentiating an implicit function

Approach:

The general method for differentiating an implicit function is to differentiate both sides of the equation with respect to the variable in question, expanding zz by the chain rule as a function of xx and yy, and then solve for the partial derivative you want.

Steps:

  1. Put F=x2+y2+z2−4zF = x^2+y^2+z^2-4z, so Fx=2xF_x=2x and Fz=2z−4F_z=2z-4
  2. ∂z∂x=−FxFz=x2−z\frac{\partial z}{\partial x} = -\frac{F_x}{F_z} = \frac{x}{2-z}
  3. Taking one more partial derivative with respect to xx: ∂2z∂x2=(2−z)−x(−∂z∂x)(2−z)2=2−z+x⋅x2−z(2−z)2=(2−z)2+x2(2−z)3\frac{\partial^2 z}{\partial x^2} = \frac{(2-z) - x(-\frac{\partial z}{\partial x})}{(2-z)^2} = \frac{2-z + x\cdot\frac{x}{2-z}}{(2-z)^2} = \frac{(2-z)^2 + x^2}{(2-z)^3}

Answer: ∂2z∂x2=(2−z)2+x2(2−z)3\frac{\partial^2 z}{\partial x^2} = \frac{(2-z)^2+x^2}{(2-z)^3}

Example 2: Implicit Functions Determined by a System of Equations

Suppose {x2+y2−uv=0xy−u2+v2=0\begin{cases} x^2 + y^2 - uv = 0 \\ xy - u^2 + v^2 = 0 \end{cases} determines u=u(x,y),v=v(x,y)u=u(x,y), v=v(x,y). Find ∂u∂x\frac{\partial u}{\partial x} and ∂v∂x\frac{\partial v}{\partial x}.

a system of implicit functionsthe Jacobian determinantdifferentiating a system

Approach:

When two equations determine two implicit functions, first differentiate the whole system to obtain a linear system in the partial derivatives you want, and then solve it.

Steps:

  1. Put F=x2+y2−uvF=x^2+y^2-uv and G=xy−u2+v2G=xy-u^2+v^2
  2. J=∂(F,G)∂(u,v)=∣−v−u−2u2v∣=−2v2−2u2=−2(u2+v2)J = \frac{\partial(F,G)}{\partial(u,v)} = \begin{vmatrix} -v & -u \\ -2u & 2v \end{vmatrix} = -2v^2 - 2u^2 = -2(u^2+v^2)
  3. Taking the partial derivative with respect to xx: {2x−vux−uvx=0y−2uux+2vvx=0\begin{cases} 2x - v u_x - u v_x = 0 \\ y - 2u u_x + 2v v_x = 0 \end{cases}
  4. By Cramer’s rule: ux=−1J∣2x−uy2v∣=4xv+yu2(u2+v2)u_x = -\frac{1}{J}\begin{vmatrix} 2x & -u \\ y & 2v \end{vmatrix} = \frac{4xv+yu}{2(u^2+v^2)} vx=−1J∣−v2x−2uy∣=4xu−yv2(u2+v2)v_x = -\frac{1}{J}\begin{vmatrix} -v & 2x \\ -2u & y \end{vmatrix} = \frac{4xu-yv}{2(u^2+v^2)}

Answer: ∂u∂x=4xv+yu2(u2+v2)\frac{\partial u}{\partial x} = \frac{4xv+yu}{2(u^2+v^2)} and ∂v∂x=4xu−yv2(u2+v2)\frac{\partial v}{\partial x} = \frac{4xu-yv}{2(u^2+v^2)}

Example 3: The Jacobian of the Polar Coordinate Transformation

Prove that the Jacobian determinant of the polar coordinate transformation x=rcos⁡θ,y=rsin⁡θx = r\cos\theta, y = r\sin\theta is rr, and find the Jacobian of the inverse function.

the inverse function theoremthe Jacobian determinantpolar coordinates

Approach:

The Jacobian determinant describes the factor by which the transformation scales an area element, so expand it with the determinant formula and simplify; the Jacobian of the inverse function is its reciprocal.

Steps:

  1. ∂(x,y)∂(r,θ)=∣cos⁡θ−rsin⁡θsin⁡θrcos⁡θ∣=rcos⁡2θ+rsin⁡2θ=r\frac{\partial(x,y)}{\partial(r,\theta)} = \begin{vmatrix} \cos\theta & -r\sin\theta \\ \sin\theta & r\cos\theta \end{vmatrix} = r\cos^2\theta + r\sin^2\theta = r
  2. By the inverse function theorem, ∂(r,θ)∂(x,y)=1∂(x,y)∂(r,θ)=1r\frac{\partial(r,\theta)}{\partial(x,y)} = \frac{1}{\frac{\partial(x,y)}{\partial(r,\theta)}} = \frac{1}{r} (r≠0r \neq 0)
  3. Checking: r=x2+y2,θ=arctan⁡yxr = \sqrt{x^2+y^2}, \theta = \arctan\frac{y}{x} ∂(r,θ)∂(x,y)=∣xryr−yr2xr2∣=x2+y2r3=r2r3=1r\frac{\partial(r,\theta)}{\partial(x,y)} = \begin{vmatrix} \frac{x}{r} & \frac{y}{r} \\ -\frac{y}{r^2} & \frac{x}{r^2} \end{vmatrix} = \frac{x^2+y^2}{r^3} = \frac{r^2}{r^3} = \frac{1}{r}

Answer: the Jacobian of the polar coordinate transformation is rr, and the Jacobian of the inverse function is 1r\frac{1}{r} (r>0r>0).


Exercises

Exercise 1

Let z=z(x,y)z = z(x,y) be determined by xz+yz+xy=0xz + yz + xy = 0. Find ∂z∂x\frac{\partial z}{\partial x} and ∂z∂y\frac{\partial z}{\partial y}.

an implicit functiondifferentiating an implicit function

Approach:

Use the implicit differentiation formula with FF taken as the left-hand side of the equation.

Steps:

  1. Put F=xz+yz+xyF = xz+yz+xy, so Fx=z+xy⋅0+y=z+yF_x = z+xy\cdot 0 + y = z+y, Fy=x+zF_y = x+z and Fz=x+yF_z = x+y
  2. ∂z∂x=−FxFz=−y+zx+y\frac{\partial z}{\partial x} = -\frac{F_x}{F_z} = -\frac{y+z}{x+y}
  3. ∂z∂y=−FyFz=−x+zx+y\frac{\partial z}{\partial y} = -\frac{F_y}{F_z} = -\frac{x+z}{x+y}

Answer: ∂z∂x=−y+zx+y\frac{\partial z}{\partial x} = -\frac{y+z}{x+y} and ∂z∂y=−x+zx+y\frac{\partial z}{\partial y} = -\frac{x+z}{x+y}

Exercise 2

Prove that the Jacobian determinant of the spherical coordinate transformation is r2sin⁡φr^2\sin\varphi.

the Jacobian determinantspherical coordinatesa proof

Approach:

Expand the 3×3 determinant and simplify with the standard trigonometric identities.

Steps:

  1. x=rsin⁡φcos⁡θx=r\sin\varphi\cos\theta, y=rsin⁡φsin⁡θy=r\sin\varphi\sin\theta and z=rcos⁡φz=r\cos\varphi
  2. ∂x∂r=sin⁡φcos⁡θ\frac{\partial x}{\partial r}=\sin\varphi\cos\theta, ∂x∂φ=rcos⁡φcos⁡θ\frac{\partial x}{\partial \varphi}=r\cos\varphi\cos\theta, ∂x∂θ=−rsin⁡φsin⁡θ\frac{\partial x}{\partial \theta}=-r\sin\varphi\sin\theta
  3. ∂y∂r=sin⁡φsin⁡θ\frac{\partial y}{\partial r}=\sin\varphi\sin\theta, ∂y∂φ=rcos⁡φsin⁡θ\frac{\partial y}{\partial \varphi}=r\cos\varphi\sin\theta, ∂y∂θ=rsin⁡φcos⁡θ\frac{\partial y}{\partial \theta}=r\sin\varphi\cos\theta
  4. ∂z∂r=cos⁡φ\frac{\partial z}{\partial r}=\cos\varphi, ∂z∂φ=−rsin⁡φ\frac{\partial z}{\partial \varphi}=-r\sin\varphi, ∂z∂θ=0\frac{\partial z}{\partial \theta}=0
  5. Expanding the determinant and simplifying gives ∂(x,y,z)∂(r,φ,θ)=r2sin⁡φ\frac{\partial(x,y,z)}{\partial(r,\varphi,\theta)} = r^2\sin\varphi

Answer: the Jacobian determinant of the spherical coordinate transformation is r2sin⁡φr^2\sin\varphi.


Summary

Symbols Used in This Article

SymbolTypePronunciation/ExplanationMeaning in This Article
FFImplicit function equationimplicit functionThe equation that defines the implicit function
JJJacobian determinantJacobianThe Jacobian determinant
∂(F,G)∂(u,v)\frac{\partial(F,G)}{\partial(u,v)}Jacobian determinantJacobian determinantThe Jacobian of the system of functions
f⃗\vec{f}Vector-valued functionvector-valued functionA multivariable vector-valued function
f⃗−1\vec{f}^{-1}Inverse functioninverse functionThe inverse of the vector-valued function

Found an issue with this content?

Tell us what could be improved.