MATH 223 Problems Problem Set for Final, Spring 2017
American University of Beirut · Advanced Calculus · 3 pages · Dr. Giuseppe Della Sala
More MATH 223 files
Show the text of this final exam
Math 223 - Advanced Calculus, Spring 2017
Review exercises, part 3
Exercise 1. (1) Let f : R2 → R be a function of class Ck. Suppose
that f(x, 0) = 0 for all x ∈ R.
Show that there exists a Ck−1
function g : R2 → R such that f(x, y) = yg(x, y). (Hint: consider
the function g(x, y) =
� 1
0
∂f
∂y(x, ty)dt).
(2) Let ρ(x, y) = y − h(x) be of class Ck and put Z = {ρ =
0} ⊂ R2. Let f : R2 → R be a Ck function vanishing on Z (this
means that f(v) = 0 for all v ∈ Z). Show that there exists a Ck−1
function g : R2 → R such that f = ρg. (Hint: similar to (1) consider
g(x, y) =
� 1
0
∂f
∂y(x, (1 − t)h(x) + ty)dt).
(3) Let ρ : R2 → R be a Ck function Z = {ρ = 0}, and suppose
that 0 ∈ Z is not a critical point for ρ. Let f be as in (2). Show that
in a neighborhood of 0 we have f = ρg for a certain Ck−1 function
g. (Hint: use the implicit function theorem - you can assume that
the function obtained by using it is Ck).
(4) With the notation of (3), show that ∇f(v) is a multiple of
∇ρ(v) for all v ∈ Z.
Exercise 2. Let Z = {(x, y) ∈ R2 : y = x2}. Using the Lagrange
multipliers method (established in the sixth assignment, Ex. 4), find
the local maxima and local minima on Z of the following functions
f : R2 → R:
(1) f(x, y) = −y;
(2) f(x, y) = x + y;
(3) f(x, y) = x2 + y2 − 2y + 1.
Exercise 3. Using the Lagrange multipliers method, find the local
maxima and local minima on Z of the following functions f : R2 →
R:
(1) f(x, y) = x + y, Z = {(x, y) ∈ R2 : x2 + y2 = 1}.
(2) f(x, y) = x2 + y2, Z = {(x, y) ∈ R2 : xy = 1}.
Exercise 4. Let U ⊂ Rn be an open domain, and let Ωp(U) be the
vector space of p-forms.
(1) Define ⋆ : Ω1(U) → Ωn−1(U) in the following way: if ω =
�n
j=1 fjdxj, ⋆ω = �n
j=1(−1)j−1fjdx1 . . . dxj−1dxj+1 . . . dxn. Show
that ⋆ is a linear isomorphism.
(2) Associate in the usual way a vector field F = (f1, . . . , fn) to
the 1-form ωF = �
j fjdxj. Assume n = 3, and let F, G be vector
fields. Show that ⋆ωF×G = ωFωG (here F × G is the cross product).
(3) From the definition, the product ω(⋆ω) is an n-form. Show
that
�
U ω(⋆ω) ≥ 0 for all ω ∈ Ω1(U), and that the equality holds if
and only if ω = 0.
(4)** Show that ∥ω∥ =
�
U ω(⋆ω) defines a norm on Ω1(U).
Exercise 5. Let ψ : Rn → Rn be a C1 diffeomorphism, and let ω
be a p-form on Rn. We say that ω is ψ-invariant if ψ∗ω = ω.
(1) Fix a diffeomorphism ψ. Show that the set of ψ-invariant
forms in Ωp(Rn) is a vector space.
(2) Show that if ω, η are ψ-invariant then ψη and dω are also
ψ-invariant.
(3) Given θ ∈ [0, 2π], define the (linear) diffeomorphism ψθ :
R2 → R2 as ψθ(x, y) = ((cos θ)x − (sin θ)y, (sin θ)x + (cos θ)y) (the
rotation of angle θ). Show that the 1-form (x2 + y2)(xdy − ydx) is
ψθ-invariant for all θ ∈ [0, 2π]. (Hint: the easiest way is to show
that x2 + y2 and xdy − ydx are ψθ-invariant, and to use (2)).
(4)* Let ψ be a diffeomorphism satisfying ψ ◦ ψ(v) = v for all
v ∈ Rn. Show that any p-form ω can be written as ω = η1 + η2,
where ψ∗η1 = η1 and ψ∗η2 = −η2.
Exercise 6. Let φ : [−1, 1] × [−1, 1] → R3 be the function defined
as φ(s, t) = (s + t, s − t, st), and put Σ = φ([−1, 1] × [−1, 1]).
(1) Show that Σ is a regular surface and compute the normal
vector at any point.
(2) Let ω be the 2-form ω = −dx2dx3 −x1dx1dx2. Compute
�
Σ ω
(with the orientation on Σ given by φ).
(3) Define the vector field F as F(x1, x2, x3) = (−1, 0, −x1).
Check that
�
Σ F =
�
Σ ω.
Exercise 7. Let f : R3 → R be a C1 function, and suppose that
f(0) = 0.
Show that there exist continuous functions g1, g2, g3 :
R3 → R such that f = x1g1 +x2g2 +x3g3. (Hint: fixed (x1, x2, x3) ∈
R3 define the function ℓ : R → R as ℓ(t) = f(tx1, tx2, tx3). Check
that f(x1, x2, x3) =
� 1
0 ℓ′(t)dt. Find ℓ′(t) by using the chain rule.)
Exercise 8. Recall that a regular curve is a C1 function γ : [a, b] →
Rn such that γ′(t) ̸= 0 for t ∈ [a, b].
(1) Let h : R → R be C1, and define Z = {(x, y) ∈ R2 : y =
h(x)}. Show that there exists a regular curve γ : R → R2 such that
γ(R) = Z.
(2) Let ρ : R2 → R be C1, and let Z = {(x, y) ∈ R2 : ρ(x, y) = 0}.
Assume that 0 ∈ Z and ∇ρ(0) ̸= 0. Show that there exists ϵ > 0
and a regular curve γ : [−ϵ, ϵ] → R2 such that γ([−ϵ, ϵ]) ⊂ Z. (Hint:
use the implicit function theorem and (1)).
(3)* Let Z = {(x, y) ∈ R2 : x2 = y3}. Show that there exists no
regular curve γ : R → R2 such that γ(R) = Z.
Exercise 9. Let ρ be a C1 real function in Rn, let Z = {ρ = 0} and
suppose that ∇ρ does not vanish on Z. Let γ : R → Rn be a C1
curve such that γ(t) ∈ Z for all t ∈ R. Show that γ′(t) is orthogonal
(w.r.t. the usual Euclidean product) to ∇ρ(γ(t)) for all t ∈ R.
Let us recall some notation: in Rn, we denote by ∇ the operator
∇ = ( ∂
∂x1, . . . ,
∂
∂xn). If f is a C1 function, the vector field ∇f =
( ∂f
∂x1, . . . , ∂f
∂xn) is called the gradient of f, while if G = (g1, . . . , gn)
is a C1 vector field, the function ∇ · G =
∂g1
∂x1 + . . . + ∂gn
∂xn (also
written simply ∇G) is called the divergence of G. The composition
∇2 = ∇◦∇, also denoted by ∆, is the Laplace operator or Laplacian;
for any C2 function f one has ∆f = ∂2f
∂x2
1 + . . . + ∂2f
∂x2n.
Exercise 10. (1) Let g : Rn → R be a continuous function. Suppose
that
�
B g = 0 for any ball B ⊂ Rn. Show that g = 0.
(2) A C2 function h : Rn → R is called harmonic if ∆h = 0.
Let now f : Rn → R be a C2 function, and suppose that we have
�
∂B ∇f = 0 for any ball B ⊂ Rn. Show that f is harmonic.
Exercise 11. Let f, g : R3 → R be functions of class C2, and let
U be a domain of R3 whose boundary bU = Σ is a regular surface.
Show that
�
U
f∆g +
�
U
∇f · ∇g =
�
Σ
f∇g.
Exercise 12. We say that a function f : Rn → R has compact
support if there exists R > 0 such that f(v) = 0 for all v ∈ Rn
such that ∥v∥ > R (in other words f is zero outside a large enough
ball). We also say that a p-form ω = �
i1,...,ip fi1...ipdxi1 . . . dxip has
compact support if all the functions fi1...ip have compact support.
(1) Suppose that ω ∈ Ω1(R2) has compact support. Show that
there exists R > 0 such that
�
BR(0) dω = 0.
(2)* Suppose that f, g : R3 → R have compact support. Show
that for R > 0 large enough we have
�
BR(0)
f∆g =
�
BR(0)
g∆f.