Problem List 5

Multivariate Calculus
Unit 3 - Inverse and Implicit function theorems, Lagrange
multipliers
Lecturer: Prof. Sonja Hohloch, Exercises: Joaquim Brugués

1. For each of the following functions in the specified domains, assert whether they satisfy the con-
ditions of the inverse function theorem. Determine if these functions are injective or surjective.
If the function is globally bijective, find the explicit expression of the inverse.

(a) F (x) = x + a, where Dom(F ) = Rn and a ∈ Rn .

(b) G(x, y) = (x + 2y, x − y), where Dom(G) = R2 .


(c) H(x, y) = x2 − y 2 , 2xy , where Dom(H) = R2 \{(0, 0)}.
2. Consider the function F : R2 → R2 given by the expression F (x, y) = (x2 y, xy 2 ). Show that F
is locally invertible around all points (x, y) such that x 6= 0 and y 6= 0. Compute the differential
of the local inverse of F at the point F (2, 1).


3. Show that the function G(x, y, z) = x + y + z 2 , x − y + z, 2x + y − z has a local differentiable
inverse around the point (1, −1, 2). Compute DG−1 (4, 4, −1).
Does G admit a global inverse?

4. Prove that the function F (x, y) = (ex cos y, ex sin y) has a local inverse at each point of its
domain, but it does not have a global inverse. Find a maximal open subset of R2 such that F
is a diffeomorphism with its image.
5. Consider the curve defined by the equation x2 + y + cos(xy) = 1.

(a) Does it exist a differentiable function g : I → R (where I is a neighbourhood of 0) such

that y = g(x) describes the same curve?
(b) Does it exist a differentiable function h : J → R (where J is a neighbourhood of 0) such
that x = h(y) describes the same curve?
6. Prove that there exists an open interval U ⊂ R around −1 and functions u, v ∈ C 1 (U, R) such
that u(−1) = v(−1) = 1 and
xeu(x) + u(x)ev(x) = 0

.
xev(x) + v(x)eu(x) = 0
√ 
7. Prove that in a neighbourhood of the point e, √1e the curve defined by the equation
 
x
xy = log
y
√
is the graphic of a function y = f (x). Prove that the function f has a critical point at x = e.

1
 ∂F ∂F
8. Let F : R3 → R a C 1 function, and let P ∈ R3 such that F (P ) = 0, ∂x (P ) 6= 0, ∂y (P ) 6= and
∂F
∂z (P ) 6= 0.

(a) Justify that you can take locally x = x(y, z), y = (x, z) or z = z(x, y) around P in such a
way that F (x, y, z) = 0.
(b) Using the notations of the last exercise, prove that

∂y ∂x ∂z
= −1.
∂x ∂z ∂y

9. Find the critical points of the function f when the variables are restricted to the given relations:
(a) f (x, y) = x + y; x2 + y 2 = 1.
(b) f (x, y) = x2 + y 2 + 1; 2x + 3y = 0.
(c) f (x, y, z) = x + y + z; x2 − y 2 = 1, 2x + z = 1.
(d) f (x, y, z) = x − y + z; x2 + y 2 + z 2 = 2.
10. What is the distance1 between the point (0, b) and the parabola x2 − 4y = 0, where b ∈ R?
What is the point or points of the parabola closest to (0, b)?

11. Find all the critical points of the function f (x, y) = x − x2 − y 2 in the disk D = {(x, y) ∈
R2 | x2 + y 2 ≤ 1}.
12. Compute the maximal volume of a parallelepiped inscribed on the ellipsoid given by the equa-
2 2 2
tion xa2 + yb2 + zc2 = 1, where a, b, c ∈ R.

1 Recall that the distance between two sets A and B is d(A, B) = inf{d(x, y) | x ∈ A, y ∈ B}. The distance between

a set A and a point p is d(A, {p}).

2
Solutions
1. (a) The differential is
DF (x) = Id ∈ L(Rn , Rn ),
so the function satisfies the condition of the inverse function theorem at all points of the
domain. Moreover, it is injective and surjective, and its explicit inverse is

F −1 (y) = y − a.

(b) The differential is  

1 2
DG(x, y) = ,
1 −1
so the function satisfies the condition of the inverse function theorem at all points of the
domain. Moreover, it is injective and surjective, and its explicit inverse is
1
G−1 (x, y) = (x + 2y, x − y) .
3

(c) The differential is  

2x −2y
DH(x, y) = ,
2y 2x
and det(DH(x, y)) = 4(x2 + y 2 ). This determinant is 0 if and only if (x, y) = (0, 0),
but this point lies outside of Dom(H). Thus, the function satisfies the conditions of the
inverse function theorem at all points of the domain.


The function is surjective, as it can be proved that H R2 \{(0, 0)} = R2 \{(0, 0)}. The
function is not injective, because H(−x, −y) = H(x, y).
2. The differential of the function is
x2
 
2xy
DF (x, y) = ,
y2 2xy

so det (DF (x, y)) = 3x2 y 2 . The differential of the inverse at F (2, 1) is
 
−1 1 4 −4
DF (F (2, 1)) = .
12 −1 4

3. The differential of G is  
1 1 2z
DG(x, y, z) = 1 −1 1 ,
2 1 −1
and det(DG(x, y, z)) = 3(1 + 2z). At (1, −1, 2), then det(DG(1, −1, 2)) = 15, so the function
has a local inverse.  
0 5 5
1
DG−1 (4, 4, −1) = 3 −9 3  .
15
3 1 −2
The function is not globally invertible, because det(DG(x, y, z)) = 0 for all points such that
z = − 21 .

3
4. The differential is
ex cos y −ex sin y
 
DF (x, y) = ,
ex sin y ex cos y
and det(DF (x, y)) = e2x (cos2 y + sin2 y) = e2x > 0. The function F is surjective, but it is not
injective, because F (x, y + 2π) = F (x, y) ∀(x, y) ∈ R2 .
Thus, it is a diffeomorphism with the image if we take as domain any set of the form R×]y0 , y0 +
2π[.
5. We may use the implicit function theorem with the differentiable function F (x, y) = x2 + y +
cos(xy) around the particular solution (0, 0):
 
∂F ∂F
DF (x, y) = (x, y), (x, y) = (2x − y sin(xy), 1 − x sin(xy)) ,
∂x ∂y
so
∂F ∂F
=0, = 1.
∂x (0,0) ∂y (0,0)

This means that we can apply the implicit function theorem to express y = g(x). However, we
cannot apply this theorem to take x = h(y).
Moreover, F (−x, y) = F (x, y) for all points, so if (x, y) is a solution to our equation, then
(−x, y) is a solution as well. Thus, we cannot define a single function x = h(y) defining our
solution, even locally, around the origin.
6. Let us take the function F : R3 → R2 given by F = (F1 , F2 ), where

F1 (x, u, v) = xeu + uev , F2 (x, u, v) = xev + veu .

We can apply the implicit function theorem with the function F with the variables (u, v). If
we compute the appropriate derivatives,
 u
xe + ev uev
  
0 e
D(u,v) F (x,u,v) = ⇒ D(u,v) F (−1,1,1) = ,
veu xev + eu e 0
 
so det D(u,v) F (−1,1,1)
= −e2 6= 0.
Thus, the conditions of the implicit function theorem are true.
 
7. Let us take the function F (x, y) = xy − log xy = xy − log x + log y, so

∂F 1 ∂F 1
(x, y) = y − , (x, y) = x + .
∂x x ∂y y
Then,
√ √ √
   
∂F 1 ∂F 1
e, √ =0, e, √ = 2 e,
∂x e ∂y e
so we can apply the implicit function theorem to write y = f (x) in a neighbourhood of the
√ 1 
point e, √e .
Moreover,
−1
√ √ √
   
0 ∂F 1 ∂F 1
f ( e) = − e, √ e, √ = 0,
∂y e ∂x e
√
so f has a critical point at e.

4
 8. (a) This is a direct application of the implicit function theorem.
(b) We need to take into account that
 −1  −1  −1
∂y ∂F ∂F ∂x ∂F ∂F ∂z ∂F ∂F
=− , =− , =− .
∂x ∂y ∂x ∂z ∂x ∂z ∂y ∂z ∂y

Thus,
 −1  −1  −1
∂y ∂x ∂z ∂F ∂F ∂F ∂F ∂F ∂F
=− = −1.
∂x ∂z ∂y ∂y ∂x ∂x ∂z ∂z ∂y
   
9. (a) There are two critical points, √1 , √1 and − √1 , − √1 .
2 2 2 2

(b) There is a unique critical point, (0, 0).

(c) There are no critical points of f in the given set.
q q q   q q q 
3 3 3 3 3 3
(d) There are two critical points, 8 , − 8 , 8 and − 8 , 8 , − 8 .

10. There are two cases depending on the value of b:

• If b ≤ 2, then (0, 0) is the point of the parabola closest to (0, b), and the distance is equal
to b.
√
• If b > 2, then ±2 b − 2, b − 2 are the two points closest to (0, b), and the distance is


√
2 b − 1.

11. We can consider separately the critical points in the interior and the boundary of D:
• In the interior there is a unique critical point, 12 , 0 , which is a maximum.

• In the boundary there are two critical points, (1, 0) and (−1, 0).
12. If a parallelepiped has vertices (±x, ±y, ±z), then its volume is given by the function f (x, y, z) =
8xyz.
 q q q 
The maximal volume is attained when the vertices are the points ±a 13 , ±b 13 , ±c 13 , and
8abc
this volume is V = √ .
27