Inverse and Implicit Function Theorem

Lecture 4.
5 Inverse, and implicit function theorems.

Among the basic tools of the trade are the inverse and implicit function theorems.
We will first state them in a coordinate dependent fashion. When we develop some
of the basic terminology we will have available a coordinate free version.
Theorem 5.1. Let U be a neighborhood let f : U ⊂ V → W be a smooth

map. Suppose dx f : Rn → Rn is invertible for some x ∈ U . Then there is a
neighborhood U � ⊂ U of x so that
f |U � → f (U � )
is a diffeomorphism. Furthermore
d0 ( f −1 ) = (d0 f )−1 .
Proof. We will construct an inverse for f using the contraction mapping theorem.
It is enough to prove the result in the case that x = 0 and f (0) = 0 and D0 f = I d.
(For the last condition replace f by (D0 f )−1 ◦ f.. Set g(x ) = f (x ) − x (so g is
the “nonlinear” part of f .) The equation f (x ) = y can be rewritten as
x + g(x ) = y
or as the fixed point equation
y − g(x ) = x .
We claim that if f is C 1 then for y in a small enough neighborhood of 0 x � →

y − g(x ) = h y (x ) is a contraction mapping on a small enough ball.
Since D0 h y (x ) = 0 and h y is C 1 there is a neighborhood Br (0) so that
�D0 h y � ≤ 12 . By the mean value theorem for x , x � ∈ Br (0) we have
1
�h y (x ) − h y (x � )� ≤ �x − x � �.
2
9
Furthermore if x ∈ Br (0) and y ∈ Br/2 (0) we have
�h y (x )� ≤�h y (x ) − h y (0)� + �h y (0)�

1
≤ �x � + �y�
2
r r
≤ +
2 2
≤ r.
Thus for y ∈ Br/2 we have h y (Br ) ⊂ Br and h y is a contraction there. The

contraction mapping theorem implies for each y the existence of a unique fixed
point φ (y) which is a least a set wise inverse for f .
We check that φ (y) is continuous.
�φ (y) − φ (y � )� = �h y (φ (y)) − h y � (φ (y � ))�

≤ �g(φ (y)) − g(φ (y � ))� + �y � − y�
1
≤ �φ (y) − φ (y � )� + �y � − y�
2
so
�φ (y) − φ (y � )� ≤ 2�y � − y�
(1)
Now we check that φ is differentiable. Let x = φ (y) and x � = φ (y � )
�φ (y) − φ (y � ) − (dx f )−1 (y − y � )� = �x − x � − (dx f )−1 ( f (x ) − f (x � ))�

≤ �dx f �−1 �(dx f )(x − x � ) − ( f (x ) − f (x � )�
≤ o(�x − x � �)
≤ o(�y − y � �).
where we use the differentiability of f to go from the second to third lines and
and inequality 1 to go from the third to the fourth.
Notice that if f is continuously differentiable then so is φ.
An important corollary of the inverse function theorem is the implicit function
theorem. The implicit function theorem can be stated in various, each useful in
some situation. We will use repeatedly the Open Mapping Theorem which say
that a surjective bounded linear map between Banach spaces is an open mapping
in particular an bounded linear map which is an algebraic isomorphism is an iso
morphism.
10
Theorem 5.2. Let f : U ⊂ V → W be a smooth map with f (0) = 0. Suppose
that for some x in U we have that Dx f is surjective and ker(Dx f ) admits a closed
complement C. Then there are neighborhoods U1 of 0 ∈ ker(Dx f ), U2 of 0 ∈ W
and diffeomorphisms φ : U1 × U2 → U and ψ : U2 → W so that the following
diagram commutes:
f
U −→ W
↑φ ↑ψ
p2
U1 × U2 −→ U2
where p2 denotes the projection on the second factor.
Proof. Write a typical element of U as a pair (k, c) with k ∈ ker (Dx f ) and
c ∈ C . The fact that C is closed means in implies that C is a C a Banach space
in its own right. Then the map K × C → V given by (k, c) �→ k + c is an
isomorphism by the Open Mapping Theorem. The Open Mapping Theorem also
implies that d0,0 f |C : C → W is an isomorphism. Let L : W → C denote its
inverse. Consider the map
F (k, c) = (k, L f (k, c)).
We have that � �
Id K ∗
d(0,0) F =
0 IdC
and again by the Open Mapping Theorem the differential of F at (0, 0) is an
isomorphism. The inverse function theorem implies F has an inverse, φ, in a
neighborhood of (0, 0). Setting ψ = d0,0 f |C we have
f (φ (k, c)) = ψ ( p2 (k, c))
on a sufficiently small neighborhood of (0, 0) since
L f (φ (k, c)) = c
on such a neighborhood.
We call a point x where Dx f is not a surjective a critical point. A point in the
range of f which is not the image of a critical point is called a regular value.
11
Definition 5.3. A subset Y of a manifold X is called submanifold if for all y ∈ Y
there is a neighborhood U of Y and a chart φ : V → B so that φ(Y ∩ U ) is an
open subset of a closed linear subspace admitting a complement.
Having made these definition we have a corollary of the implicit function the
orem.
Corollary 5.4. The preimage of a regular value is a submanifold.
12

Inverse and Implicit Function Theorem

Uploaded by

Copyright:

Available Formats

Inverse and Implicit Function Theorem

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Inverse and Implicit Function Theorem

Uploaded by

Copyright:

Available Formats

What is the inverse function theorem and what does it state?

What is the implicit function theorem and what conditions must be satisfied for it to apply?

Lecture 4.

5 Inverse, and implicit function theorems.

Theorem 5.1. Let U be a neighborhood let f : U ⊂ V → W be a smooth

or as the ﬁxed point equation

We claim that if f is C 1 then for y in a small enough neighborhood of 0 x � →

�h y (x )� ≤�h y (x ) − h y (0)� + �h y (0)�

Thus for y ∈ Br/2 we have h y (Br ) ⊂ Br and h y is a contraction there. The

�φ (y) − φ (y � )� = �h y (φ (y)) − h y � (φ (y � ))�

�φ (y) − φ (y � ) − (dx f )−1 (y − y � )� = �x − x � − (dx f )−1 ( f (x ) − f (x � ))�

where p2 denotes the projection on the second factor.

F (k, c) = (k, L f (k, c)).

f (φ (k, c)) = ψ ( p2 (k, c))

on a sufﬁciently small neighborhood of (0, 0) since

Corollary 5.4. The preimage of a regular value is a submanifold.

You might also like