Implicit Function Theorem

Math 32A Week 10 Notes
November 29 and December 1, 2016

Austin Christian
The Implicit Function Theorem

Suppose we have a function of two variables, F (x, y), and we’re interested in its height-c level
curve; that is, solutions to the equation F (x, y) = c. For instance, perhaps F (x, y) = x2 + y 2
and c = 1, in which case the level curve we care about is the familiar unit circle. It would
be nice if choosing a value for x in the equation F (x, y) = c would immediately determine
the value of y — that is, if F (x, y) = c determined y as a function of x. But we know √ that
this isn’t
√ generally true. In the case of the unit circle, fixing a value x leaves both 1 − x2
and − 1 − x2 as possibilities for the value of y. Graphically, this obstruction is represented
by the fact that x2 + y 2 = 1 fails the familiar vertical line test, as can be seen in Figure 1.
Figure 1: The level curve x2 + y 2 = 1. The green segment represents a neighborhood of the
red point on which y is determined by x.
Something else that can be seen in Figure 1, though, is that our graph does pass the
vertical line test locally. That is, at most points on the circle we can choose a small neigh-
borhood where our curve satisfies the vertical line test, and thus determines y as a function
of x. The two points where we cannot choose such a neighborhood are (0, 1) and (0, −1).
Indeed, these are precisely the points exempted from the following important theorem.
The Implicit Function Theorem for R2 . Consider a continuously differentiable

function F : R2 → R and a point (x0 , y0 ) ∈ R2 so that F (x0 , y0 ) = c. If ∂F ∂y
(x0 , y0 ) 6= 0,
then there is a neighborhood of (x0 , y0 ) so that whenever x is sufficiently close to x0 there
is a unique y so that F (x, y) = c. Moreover, this assignment is makes y a continuous
function of x.
1
The theorem says that we can make y a function of x — except when ∂F
∂y
= 0. In our case
Fy = 2y vanishes whenever y = 0, and this happens at two points: the two we’ve already
identified as problems. The theorem also holds in three dimensions:
The Implicit Function Theorem for R3 . Consider a continuously differentiable

function F : R3 → R and a point (x0 , y0 , z0 ) ∈ R3 so that F (x0 , y0 , z0 ) = c. If
∂F
∂z
(x0 , y0 , z0 ) 6= 0, then there is a neighborhood of (x0 , y0 , z0 ) so that whenever (x, y) is
sufficiently close to (x0 , y0 ) there is a unique z so that F (x, y, z) = c. Moreover, this
assignment is makes z a continuous function of x and y.
Colloquially, the upshot of the implicit function theorem is that for sufficiently nice points
on a surface, we can (locally) pretend this surface is the graph of a function. The primary
use for the implicit function theorem in this course is for implicit differentiation. You’ve
already seen the two-variable version of this in your first calculus class. In particular, you
probably did the unit circle example we saw above. The usual way to go about this is to
consider the equation F (x, y) = c and to differentiate both sides with respect to x. For the
unit circle this yields
dy
2x + 2y = 0,
dx
dy
from which we solve to find that dx = −x/y. Of course, this expression doesn’t make any
sense at y = 0, but we’ve already excepted the points where this happens. Now that we
have the multivariable chain rule at our disposal, we can actually address this problem more
generally. Differentiating both sides of F (x, y) = c with respect to x gives
∂F ∂F ∂y
·1+ · = 0.
∂x ∂y ∂x
Solving for y 0 (x) leads us to

Fx
y 0 (x) = −
Fy
whenever Fy 6= 0. While we’re here we can also address the three-variable case. Suppose
we have F : R3 → R and a point (x0 , y0 , z0 ) where the implicit function theorem applies.
∂z ∂z
Writing z = z(x, y), we’re interested in the partial derivatives ∂x and ∂y . As before, we’ll do
this by differentiating the equation F (x, y, z) = c with respect to x, and then with respect
to y. An important point here is that we’re considering z as a function of x and y, but
we’re not considering x and y as depending on each other. We could actually say that we’re
interested in the x- or y-derivative of the equation
F (x, y, z(x, y)) = c.
With respect to x we have

∂F ∂F ∂F ∂z
·1+ ·0+ · = 0.
∂x ∂y ∂z ∂x
2
∂z
Solving for ∂x
gives
∂z ∂F/∂x Fx
=− =− .
∂x ∂F/∂z Fz
∂z
Similarly, ∂y
= −Fy /Fz . Here’s an example (due to Lincoln Chayes) to test our understand-
ing.
Example. Consider two three-variable functions H(x, y, z) and K(x, y, z) and the associated
level surfaces
H(x, y, z) = a and K(x, y, z) = b.
We assume that these surfaces intersect along a curve which contains the point (x0 , y0 , z0 ),
and that on some neighborhood of this point, the curve determines y as a function y(x) of
x. Derive a formula for y 0 (x) near x0 in terms of the partial derivatives of H and K. (We
assume that the denominators involved in this derivation do not vanish.)
(Solution) First we apply the implicit function theorem to H at the point (x0 , y0 , z0 ). This
gives us a function zH (x, y) on some neighborhood of (x0 , y0 ) so that
H(x, y, zH (x, y)) = a.
Because we care about the intersection of the two level surfaces, we may substitute this
function in place of z in the formula K(x, y, z) = b, giving us
K(x, y, zH (x, y)) = b.
The left side is now a function of two variables — x an y — and we may apply the implicit
function theorem to write y as a function of x for values of x near x0 . According to our
formula above for y 0 (x) we have
∂
0 ∂x
(K(x, y, zH (x, y)))
y (x) = −∂ .
∂y
(K(x, y, zH (x, y)))
We apply the chain rule to the numerator to obtain

∂ ∂K ∂K ∂K ∂zH
(K(x, y, zH (x, y))) = ·1+ ·0+ .
∂x ∂x ∂y ∂z ∂x
But according to the three-variable implicit function theorem we have
∂zH Hx
=− ,
∂x Hz
so
∂ Hx
(K(x, y, zH (x, y))) = Kx − Kz .
∂x Hz
3
Similarly,
∂ Hy
(K(x, y, zH (x, y))) = Ky − Kz ,
∂y Hz
so
Kx − Kz Hx /Hz Kx Hz − Kz Hx
y 0 (x) = − =− .
Ky − Kz Hy /Hz Ky Hz − Kz Hy
To see our formula in action, consider
H(x, y, z) = x2 + y 2 and K(x, y, z) = x2 + y 2 + z 2 .
The level surfaces H(x, y, z) = 1 and K(x, y, z) = 1 intersect in the unit circle in the plane,
where we know that y 0 (x) = −x/y from our earlier work. Our new formula gives
(2x)(0) − (2z)(2x) x
y 0 (x) = − =− ,
(2y)(0) − (2z)(2y) y
a welcome reality check. ♦

Implicit Function Theorem

Uploaded by

Implicit Function Theorem

Uploaded by

Math 32A Week 10 Notes

November 29 and December 1, 2016

The Implicit Function Theorem

The Implicit Function Theorem for R2 . Consider a continuously differentiable

The Implicit Function Theorem for R3 . Consider a continuously differentiable

Solving for y 0 (x) leads us to

F (x, y, z(x, y)) = c.

With respect to x we have

H(x, y, zH (x, y)) = a.

K(x, y, zH (x, y)) = b.

We apply the chain rule to the numerator to obtain

H(x, y, z) = x2 + y 2 and K(x, y, z) = x2 + y 2 + z 2 .

a welcome reality check. ♦

You might also like