Scribd
Upload
46 views3 pages

Lecture16 PDF

This document provides a summary of key concepts from linear algebra and topology that will be used to develop vector calculus in multiple dimensions: 1. It reviews definitions of linear independence, spanning sets, bases, subspaces, linear maps, kernels, images, and the rank-nullity theorem from linear algebra. 2. It introduces the dual space construction, where the dual space V* of a vector space V consists of all linear functions from V to the real numbers R. 3. It describes how a linear map A between vector spaces V and W induces a dual map A* between the dual spaces W* and V*, and explains that the matrix of A* is the transpose of the matrix of

Uploaded by

mukul
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
46 views3 pages

Lecture16 PDF

This document provides a summary of key concepts from linear algebra and topology that will be used to develop vector calculus in multiple dimensions: 1. It reviews definitions of linear independence, spanning sets, bases, subspaces, linear maps, kernels, images, and the rank-nullity theorem from linear algebra. 2. It introduces the dual space construction, where the dual space V* of a vector space V consists of all linear functions from V to the real numbers R. 3. It describes how a linear map A between vector spaces V and W induces a dual map A* between the dual spaces W* and V*, and explains that the matrix of A* is the transpose of the matrix of

Uploaded by

mukul
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Lecture 16

4 Multi­linear Algebra
4.1 Review of Linear Algebra and Topology
In today’s lecture we review chapters 1 and 2 of Munkres. Our ultimate goal (not
today) is to develop vector calculus in n dimensions (for example, the generalizations
of grad, div, and curl).
Let V be a vector space, and let vi ∈ V, i = 1, . . . , k.
1. The vi� s are linearly independent if the map from Rk to V mapping (c1 , . . . , ck )
to c1 v1 + . . . + ck vk is injective.

2. The vi� s span V if this map is surjective (onto).

3. If the vi� s form a basis, then dim V = k.

4. A subset W of V is a subspace if it is also a vector space.

5. Let V and W be vector spaces. A map A : V → W is linear if A(c1 v1 + c2 v2 ) =


c1 A(v1 ) + c2 A(v2 ).

6. The kernel of a linear map A : V → W is

ker A = {v ∈ V : Av = 0}. (4.1)

7. The image of A is
Im A = {Av : v ∈ V }. (4.2)

8. The following is a basic identity:

dim ker A + dim Im A = dim V. (4.3)

9. We can associate linear mappings with matrices. Let v1 , . . . , vn be a basis for


V , and let w1 , . . . , wm be a basis for W . Let

�
Avj = aij wj . (4.4)
i=1

Then we associate the linear map A with the matrix [aij ]. We write this A ∼
[aij ].
�
10. If v1 , . . . , vn is a basis for V and uj = aij wj are n arbitrary vectors in W ,
then there exists a unique linear mapping A : V → W such that Avj = uj .

1
�
11. Know all the material in Munkres section 2 on matrices and determinants.
12. The quotient space construction. Let V be a vector space and W a subspace.
Take any v ∈ V . We define v + W ≡ {v + w : w ∈ W }. Sets of this form are
called W ­cosets. One can check that given v1 + W and v2 + W ,
(a) If v1 − v2 ∈ W , then v1 + W = v2 + W .
(b) If v1 − v2 ∈
/ W , then (v1 + W ) ∩ (v2 + W ) = φ.
So every vector v ∈ V belongs to a unique W ­coset.
The quotient space V /W is the set of all W =cosets.
For example, let V = R2 , and let W = {(a, 0) : a ∈ R}. The W ­cosets are then
vertical lines.
The set V /W is a vector space. It satisfies vector addition: (v1 +W )+(v2 +W ) =
(v1 + v2 ) + W . It also satisfies scaler multiplication: λ(v + W ) = λv + W . You
should check that the standard axioms for vector spaces are satisfied.
There is a natural projection from V to V /W :
π : V → V /W, v → v + W. (4.5)
The map π is a linear map, it is surjective, and ker π = W . Also, Im π = V /W ,
so
dim V /W = dim Im π
= dim V − dim ker π (4.6)
= dim V − dim W.

4.2 Dual Space


13. The dual space construction: Let V be an n­dimensional vector space. Define
V ∗ to be the set of all linear functions � : V → R. Note that if �1 , �2 ∈ V ∗ and
λ1 , λ2 ∈ R, then λ1 �1 + λ2 �2 ∈ V ∗ , so V ∗ is a vector space.
What does V ∗ look like? Let e1 , . . . , en be a basis of V . By item (9), there
exists a unique linear map e∗i ∈ V ∗ such that
� ∗
ei (ei ) = 1,
e∗i (ej ) = 0, if j �= i.
Claim. The set of vectors e∗1 , . . . , e∗n is a basis of V ∗ .
� ∗ � ∗
Proof. Suppose � = ci ei = 0. Then 0 = �(ej ) = ci ei (ej ) = cj , so c1 =
. . . = cn = 0. This proves that the vectors e∗i are�linearly independent. Now,
if � ∈ V ∗ and �(ei ) = cj one can check that � = ci e∗i . This proves that the
∗ ∗
vectors ei span V .

2
The vectors e∗1 , . . . , e∗n are said to be a basis of V ∗ dual to e1 , . . . , en .
Note that dim V ∗ = dim V .
Suppose that we have a pair of vectors spaces V, W and a linear map A : V →
W . We get another map
A∗ : W ∗ → V ∗ , (4.7)
defined by A∗ � = � ◦ A, where � ∈ W ∗ is a linear map � : W → R. So A∗ � is a
linear map A∗ � : V → R. You can check that A∗ : W ∗ → V ∗ is linear.
We look at the matrix description of A∗ . Define the following bases:

e1 , . . . , en a basis of V (4.8)
f1 , . . . , fn a basis of W (4.9)
e∗1 , . . . , e∗n a basis of V∗ (4.10)
f1∗ , . . . , fn∗ a basis of W ∗. (4.11)

Then
A∗ fj∗ (ei ) = fj∗ (Aei )
�
= fj∗ ( aki fk ) (4.12)
k
= aji

So, �
A∗ fj = ajk e∗k , (4.13)
k
∗ t
which shows that A ∼ [aji ] = [aij ] , the transpose of A.

You might also like