# Exercises in Advanced Linear and Multilinear Algebra John M.

Erdman Portland State University Version May 25, 2009 c 2009 John M. Erdman

Contents
Preface Some Algebraic Objects Chapter 1. Abelian Groups 1.1. Background 1.2. Exercises (Due: Wed. Jan. 7) Chapter 2. Functions and Diagrams 2.1. Background 2.2. Exercises (Due: Fri. Jan. 9) Chapter 3. Rings 3.1. Background 3.2. Exercises (Due: Mon. Jan. 12) Chapter 4. Vector Spaces and Subspaces 4.1. Background 4.2. Exercises (Due: Wed. Jan. 14) Chapter 5. Linear Combinations and Linear Independence 5.1. Background 5.2. Exercises (Due: Fri. Jan. 16) Chapter 6. Bases for Vector Spaces 6.1. Background 6.2. Exercises (Due: Wed. Jan. 21) Chapter 7. Linear Transformations 7.1. Background 7.2. Exercises (Due: Fri. Jan. 23) Chapter 8. Linear Transformations (continued) 8.1. Background 8.2. Exercises (Due: Mon. Jan. 26) Chapter 9. Duality in Vector Spaces 9.1. Background 9.2. Exercises (Due: Wed. Jan. 28) Chapter 10. Duality in Vector Spaces (continued) 10.1. Background 10.2. Exercises (Due: Fri. Jan. 30) Chapter 11. The Language of Categories 11.1. Background 11.2. Exercises (Due: Mon. Feb. 2)
iii

vii ix 1 1 2 3 3 5 7 7 8 9 9 10 13 13 14 17 17 19 21 21 22 25 25 26 27 27 28 31 31 32 33 33 35

iv

CONTENTS

Chapter 12. Direct Sums 12.1. Background 12.2. Exercises (Due: Wed. Feb. 4) Chapter 13. Products and Quotients 13.1. Background 13.2. Exercises (Due: Mon. Feb. 9) Chapter 14. Products and Quotients (continued) 14.1. Background 14.2. Exercises (Due: Wed. Feb. 11) Chapter 15. Projection Operators 15.1. Background 15.2. Exercises (Due: Fri. Feb. 13) Chapter 16. Algebras 16.1. Background 16.2. Exercises (Due: Mon. Feb. 16) Chapter 17. Spectra 17.1. Background 17.2. Exercises (Due: Wed. Feb. 18) Chapter 18. Polynomials 18.1. Background 18.2. Exercises (Due: Fri. Feb. 20) Chapter 19. Polynomials (continued) 19.1. Background 19.2. Exercises (Due: Mon. Feb. 23) Chapter 20. Invariant Subspaces 20.1. Background 20.2. Exercises (Due: Wed. Feb. 25) Chapter 21. The Spectral Theorem for Vector Spaces 21.1. Background 21.2. Exercises (Due: Fri. Feb. 27) Chapter 22. The Spectral Theorem for Vector Spaces (continued) 22.1. Background 22.2. Exercises (Due: Mon. Mar. 2) Chapter 23. Diagonalizable Plus Nilpotent Decomposition 23.1. Background 23.2. Exercises (Due: Wed. Mar. 4) Chapter 24. Inner Product Spaces 24.1. Background 24.2. Exercises (Due: Mon. Mar. 9) Chapter 25. Orthogonality and Adjoints 25.1. Background 25.2. Exercises (Due: Wed. Mar. 11) Chapter 26. Orthogonal Projections

37 37 39 41 41 43 45 45 46 49 49 50 51 51 52 55 55 56 59 59 61 63 63 64 67 67 68 69 69 70 71 71 72 73 73 74 77 77 78 81 81 82 85

CONTENTS

v

26.1. Background 26.2. Exercises (Due: Fri. Mar. 13) Chapter 27. The Spectral Theorem for Inner Product Spaces 27.1. Background 27.2. Exercises (Due: Mon. Mar. 30) Chapter 28. Multilinear Maps 28.1. Background Diﬀerential calculus Permutations Multilinear maps 28.2. Exercises (Due: Wed. Apr. 1) Chapter 29. Determinants 29.1. Background 29.2. Exercises (Due: Fri. Apr. 3) Chapter 30. Free Vector Spaces 30.1. Background 30.2. Exercises (Due: Mon. Apr. 6) Chapter 31. Tensor Products of Vector Spaces 31.1. Background 31.2. Exercises (Due: Wed. Apr. 8) Chapter 32. Tensor Products of Vector Spaces (continued) 32.1. Background 32.2. Exercises (Due: Fri. Apr. 10) Chapter 33. Tensor Products of Linear Maps 33.1. Background 33.2. Exercises (Due: Mon. April 13) Chapter 34. Grassmann Algebras 34.1. Background 34.2. Exercises (Due: Wed. April 15) Chapter 35. Graded Algebras 35.1. Background 35.2. Exercises (Due: Fri. April 17) Chapter 36. Existence of Grassmann Algebras 36.1. Background 36.2. Exercises (Due: Mon. April 20) Chapter 37. The Hodge ∗-operator 37.1. Background 37.2. Exercises (Due: Wed. April 22) Chapter 38. Diﬀerential Forms 38.1. Background 38.2. Exercises (Due: Fri. April 24) Chapter 39. The Exterior Diﬀerentiation Operator 39.1. Background

85 86 89 89 90 93 93 93 93 94 95 97 97 98 99 99 100 101 101 102 103 103 104 105 105 106 107 107 108 109 109 110 111 111 112 113 113 114 115 115 117 119 119

vi

CONTENTS

39.2.

Exercises

(Due: Mon. April 27)

120 121 121 122 125 125 126 127 127 128 129 129 130 131 131 134 135 135 137 139 139 141 143 143 145 147 147 148 149 149 150 151 151 152 153 153 154 155 157

Chapter 40. Diﬀerential Calculus on R3 40.1. Background 40.2. Exercises (Due: Wed. April 29) Chapter 41. Closed and Exact Forms 41.1. Background 41.2. Exercises (Due: Mon. May 4) Chapter 42. The de Rham Cohomology Group 42.1. Background 42.2. Exercises (Due: Wed. May 6) Chapter 43. Cochain Complexes 43.1. Background 43.2. Exercises (Due: Fri. May 8) Chapter 44. Simplicial Homology 44.1. Background 44.2. Exercises (Due: Wed. May 13) Chapter 45. Simplicial Cohomology 45.1. Background 45.2. Exercises (Due: Fri. May 15) Chapter 46. Integration of Diﬀerential Forms 46.1. Background 46.2. Exercises (Due: Mon. May 18) Chapter 47. Stokes’ Theorem 47.1. Background 47.2. Exercises (Due: Wed. May 20) Chapter 48. Quadratic Forms 48.1. Background 48.2. Exercises (Due: Fri. May 22) Chapter 49. Deﬁnition of Cliﬀord Algebra 49.1. Background 49.2. Exercises (Due: Wed. May 27) Chapter 50. Orthogonality with Respect to Bilinear Forms 50.1. Background 50.2. Exercises (Due: Fri. May 29) Chapter 51. Examples of Cliﬀord Algebras 51.1. Background 51.2. Exercises (Due: Mon. June 1) Bibliography Index

Preface
This collection of exercises is designed to provide a framework for discussion in a two-term senior/ﬁrst year graduate level class in linear and multilinear algebra such as the one I have conducted fairly regularly at Portland State University. Most recently I have been using Douglas R. Farenick’s Algebras of Linear Transformations (Springer-Verlag 2001) as a text for parts of the course. (In these Exercises it is referred to as AOLT.) For the more elementary parts of linear algebra there is certainly no shortage of readily available texts. In particular there are now a number of excellent online texts which are available free of charge. Among the best are Linear Algebra [14] by Jim Heﬀeron, http://joshua.smcvt.edu/linearalgebra A First Course in Linear Algebra [2] by Robert A. Beezer, http://linear.ups.edu/download/fcla-electric-2.00.pdf and Linear Algebra [9] by Paul Dawkins. http://tutorial.math.lamar.edu/pdf/LinAlg/LinAlg_Complete.pdf Another very useful online resource is Przemyslaw Bogacki’s Linear Algebra Toolkit [4]. http://www.math.odu.edu/~bogacki/lat And, of course, many topics in linear algebra are discussed with varying degrees of thoroughness in the Wikipedia [28] http://en.wikipedia.org and Eric Weisstein’s Mathworld [27]. http://http://mathworld.wolfram.com For more advanced topics in linear algebra some references that I particularly like are Paul Halmos’s Finite-Dimensional Vector Spaces [13], Hoﬀman and Kunze’s Linear Algebra [15], Charles W. Curits’s Linear Algebra: An Introductory Approach [8], Steven Roman’s Advanced Linear Algebra [23], and William C. Brown’s A Second Course in Linear Algebra [5]. Readable introductions to some topics in multilinear algebra (exterior, graded, and Cliﬀord algebras, for example) are a bit harder to come by. For these I will suggest speciﬁc sources later in these Exercises. The short introductory Background section which precedes each assignment is intended to ﬁx notation and provide “oﬃcial” deﬁnitions and statements of important theorems for ensuing discussions.

vii

Some Algebraic Objects
Let S be a nonempty set. Consider the following axioms: (1) + : S × S → S. ( + is a binary operation, called addition, on S) (2) (x + y) + z = x + (y + z) for all x, y, z ∈ S. (associativity of addition) (3) There exists 0S ∈ S such that x + 0S = 0S + x = x for all x ∈ S. (existence of an additive identity) (4) For every x ∈ S there exists −x ∈ S such that x + (−x) = (−x) + x = 0S . (existence of additive inverses) (5) x + y = y + x for all x, y ∈ S. (commutativity of addition) (6) m : S × S → S : (x, y) → xy. (the map (x, y) → xy is a binary operation, called multiplication, on S) (7) (xy)z = x(yz) for all x, y, z ∈ S. (associativity of multiplication) (8) (x+y)z = xz +yz and x(y +z) = xy +xz for all x, y, z ∈ S. (multiplication distributes over addition) (9) There exists 1S in S such that x 1S = 1S x = x for all x ∈ S. (existence of a multiplicative identity or unit) (10) 1S = 0S . (11) For every x ∈ S such that x = 0S there exists x−1 ∈ S such that xx−1 = x−1 x = 1S . (existence of multiplicative inverses) (12) xy = yx for all x, y ∈ S. (commutativity of multiplication) Deﬁnitions. • (S, +) is a semigroup if it satisﬁes axioms (1)–(2). • (S, +) is a monoid if it satisﬁes axioms (1)–(3). • (S, +) is a group if it satisﬁes axioms (1)–(4). • (S, +) is an Abelian group if it satisﬁes axioms (1)–(5). • (S, +, m) is a ring if it satisﬁes axioms (1)–(8). • (S, +, m) is a commutative ring if it satisﬁes axioms (1)–(8) and (12). • (S, +, m) is a unital ring (or ring with identity) if it satisﬁes axioms (1)–(9). • (S, +, m) is a division ring (or skew field) if it satisﬁes axioms (1)–(11). • (S, +, m) is a field if it satisﬁes axioms (1)–(12). Remarks. • A binary operation is often written additively, (x, y) → x + y, if it is commutative and multiplicatively, (x, y) → xy, if it is not. This is by no means always the case: in a commutative ring (the real numbers or the complex numbers, for example), both addition and multiplication are commutative. • When no confusion is likely to result we often write 0 for 0S and 1 for 1S . • Many authors require a ring to satisfy axioms (1)–(9). • It is easy to see that axiom (10) holds in any unital ring except the trivial ring S = {0}.

ix

CHAPTER 1

Abelian Groups
1.1. Background Topics: groups, Abelian groups, group homomorphisms.

1.1.1. Definition. Let G and H be Abelian groups. A map f : G → H is a homomorphism if f (x + y) = f (x) + f (y) for all x, y ∈ G. We will denote by Hom(G, H) the set of all homomorphisms from G into H and will abbreviate Hom(G, G) to Hom(G). 1.1.2. Definition. Let G and H be Abelian groups. For f and g in Hom(G, H) we deﬁne f + g : G → H : x → f (x) + g(x).

1

2

1. ABELIAN GROUPS

1.2. Exercises

(Due: Wed. Jan. 7)

1.2.1. Exercise. In AOLT, page 2, the author says that a vector space is, among other things, “an additive Abelian group”. What sort of Abelian group is an additive Abelian group? Give an example of a very common Abelian group which is not written additively. 1.2.2. Exercise. Prove that the identity element in an Abelian group is unique. 1.2.3. Exercise. Let x be an element of an Abelian group. Prove that the inverse of x is unique. 1.2.4. Exercise. Let x be an element of an Abelian group. Prove that if x + x = x, then x = 0. 1.2.5. Exercise (†). Let x be an element of an Abelian group. Prove that −(−x) = x. 1.2.6. Exercise. Let f : G → H be a homomorphism of Abelian groups. Show that f (0) = 0. 1.2.7. Exercise (†). Let f : G → H be a homomorphism of Abelian groups. Show that f (−x) = −f (x) for each x ∈ G. 1.2.8. Exercise. Let E2 be the Euclidean plane. It contains points (which do not have coordinates) and lines (which do not have equations). A directed segment is an ordered pair of points. Deﬁne two directed segments to be equivalent if they are congruent (have the same length), lie on parallel lines, and have the same direction. This is clearly an equivalence relation on the set DS − − → of directed segments in the plane. We denote by P Q the equivalence class containing the directed segment (P, Q), going from the point P to the point Q. Deﬁne an operation on DS by − − → − − → − → P Q + QR = P R. Show that this operation is well deﬁned and that under it DS is an Abelian group. − − → −→ − 1.2.9. Exercise. Suppose that A, B, C, and D are points in the plane such that AB = CD. − → −→ − Show that AC = BD. 1.2.10. Exercise. Let G and H be Abelian groups. Show that with addition as deﬁned in 1.1.2 Hom(G, H) is an Abelian group.

CHAPTER 2

Functions and Diagrams
2.1. Background Topics: functions, commutative diagrams. 2.1.1. Definition. It is frequently useful to think of functions as arrows in diagrams. For example, the situation h : R → S, j : R → T , k : S → U , f : T → U may be represented by the following diagram. R
h j

/T
f

S 

k 

/U

The diagram is said to commute if k ◦ h = f ◦ j. Diagrams need not be rectangular. For instance,
?? ?? ??d h ?? ?? ?  /U S
k

R?

is a commutative diagram if d = k ◦ h. 2.1.2. Example. Here is one diagrammatic way of stating the associative law for composition of functions: If h : R → S, g : S → T , and f : T → U and we deﬁne j and k so that the triangles in the diagram
/T ?   g  f h       /U S

R

j

k

commute, then the square also commutes. 2.1.3. Convention. If S, T , and U are sets we will not distinguish between (S × T ) × U , S × (T × U ), and S × T × U . That is, the ordered pairs (s, t), u and s, (t, u) and the ordered triple (s, t, u) will be treated as identical. 2.1.4. Notation. Let S be a set. The map idS : S → S : x → x is the identity function on S. When no confusion will result we write id for idS . 2.1.5. Notation. Let f : S → U and g : T → V be functions between sets. Then f × g denotes the map f × g : S × T → U × V : (s, t) → f (s), g(t) .
3

4

2. FUNCTIONS AND DIAGRAMS

2.1.6. Convention. We will have use for a standard one-element set, which, if we wish, we can regard as the Cartesian product of an empty family of sets. We will denote it by 1. For each set S there is exactly one function from S into 1. We will denote it by εS . If no confusion is likely to arise we write ε for εS . 2.1.7. Notation. We denote by δ the diagonal mapping of a set S into S × S. That is, δ : S → S × S : s → (s, s). 2.1.8. Notation. Let S be a set. We denote by σ the interchange (or switching) operation on the S × S. That is, σ : S × S → S × S : (s, t) → (t, s). 2.1.9. Notation. If S and T are sets we denote by F(S, T ) the family of all functions from S into T . We will use F(S) for F(S, R), the set of all real valued functions on S.

2.2. EXERCISES

(DUE: FRI. JAN. 9)

5

2.2. Exercises

(Due: Fri. Jan. 9)

2.2.1. Exercise. Let S be a set and a : S × S → S be a function such that the diagram S×S×S
a×id id ×a

/ /S×S

a

/S

(D1)

commutes. What is (S, a)? Hint. Interpret a as, for example, addition (or multiplication). 2.2.2. Exercise. Let S be a set and suppose that a : S × S → S and η : 1 → S are functions such that both diagram (D1) above and the diagram (D2) which follows commute.
O s9 S eKKKK ss KKK ss f ss KKg KKK a ss ss KKK s ss KK ss / S×S o 1×S S×1
η×id id ×η

(D2)

(Here f and g are the obvious bijections.) What is (S, a, η)? 2.2.3. Exercise (†). Let S be a set and suppose that a : S × S → S and η : 1 → S are functions such that the diagrams (D1) and (D2) above commute. Suppose further that there is a function ι : S → S for which the following diagram commutes. S HH
/ a / / S×S v: S HH id ×ι vv HH vv HH v HH vv HH vv vv ε HHH vv η HH vv HH v HH vv \$ vv
δ

/ S×S

ι×id

(D3)

1

What is (S, a, η, ι)? 2.2.4. Exercise. Let S be a set and suppose that a : S × S → S, η : 1 → S, and ι : S → S are functions such that the diagrams (D1), (D2), and (D3) above commute. Suppose further that the following diagram commutes. σ / S×S S×S ?
?? ?? ?? a ?? ? ??       a  

S

(D4)

What is (S, a, η, ι, σ)? 2.2.5. Exercise. Let f : G → H be a function between Abelian groups. Suppose that the diagram G×G
+ f ×f

/ H ×H
+

G 

f 

/H

commutes. What can be said about the function f ? 2.2.6. Exercise (††). Let S be a set with exactly one element. Discuss the cardinality of (that is, the number of elements in) the sets F(∅, ∅), F(∅, S),F(S, ∅), and F(S, S),

CHAPTER 3

Rings
3.1. Background Topics: rings; zero divisors; cancellation property; ring homomorphisms.

3.1.1. Definition. An element a of a ring is left cancellable if ab = ac implies that b = c. It is right cancellable if ba = ca implies that b = c. A ring has the cancellation property if every nonzero element of the ring is both left and right cancellable. 3.1.2. Definition. A nonzero element a of a ring is a zero divisor (or divisor of zero) if there exists a nonzero element b of the ring such that (i) ab = 0 or (ii) ba = 0. Most everyone agrees that a nonzero element a of a ring is a left divisor or zero if it satisﬁes (i) for some nonzero b and a right divisor of zero if it satisﬁes (ii) for some nonzero b. There agreement on terminology ceases. Some authors ([7], for example) use the deﬁnition above for divisor of zero; others ([17], for example) require a divisor of zero to be both a left and a right divisor of zero; and yet others ([18], for example) avoid the issue entirely by deﬁning zero divisors only for commutative rings. Palmer in [22] makes the most systematic distinctions: a zero divisor is deﬁned as above; an element which is both a left and a right zero divisor is a two-sided zero divisor ; and if the same nonzero b makes both (i) and (ii) hold a is a joint zero divisor. 3.1.3. Definition. A function f : R → S between rings is a (ring) homomorphism if f (x + y) = f (x) + f (y) and f (xy) = f (x)f (y) for all x and y in R. If in addition R and S are unital rings and f (1R ) = 1S then f is a unital (ring) homomorphism.

7

8

3. RINGS

3.2. Exercises

(Due: Mon. Jan. 12)

3.2.1. Exercise. Show that the additive identity of a ring is an annihilator. That is, show that for every element a of a ring 0a = a 0 = 0. 3.2.2. Exercise. Show that if a and b are elements of a ring, then (−a)b = a(−b) = −ab and (−a)(−b) = ab. 3.2.3. Exercise. Every division ring has the cancellation property. 3.2.4. Exercise. A division ring has no zero divisors. That is, if ab = 0 in a division ring, then a = 0 or b = 0. 3.2.5. Exercise (†). A ring has the cancellation property if and only if it has no zero divisors. 3.2.6. Exercise. Let G, H, and J be Abelian groups. If f ∈ Hom(G, H) and g ∈ Hom(H, J), then the composite of g with f belongs to Hom(G, J). Note: This composite is usually written as gf rather than as g ◦ f . 3.2.7. Exercise. Let G be an Abelian group. Then Hom(G) is a unital ring (under the operations of addition and composition). 3.2.8. Exercise. Let (G, +) be an Abelian group, F a ﬁeld, and M : F → Hom(G) be a unital ring homomorphism. What is (G, +, M )?

CHAPTER 4

Vector Spaces and Subspaces
4.1. Background Topics: vector spaces, linear (or vector) subspaces.

4.1.1. Notation. F = F[a, b] = {f : f is a real valued function on the interval [a, b]} P = P[a, b] = {p : p is a polynomial function on [a, b]} P4 = P4 [a, b] = {p ∈ P : the degree of p is three or less} Q4 = Q4 [a, b] = {p ∈ P : the degree of p is equal to 4} C = C[a, b] = {f ∈ F : f is continuous} D = D[a, b] = {f ∈ F : f is diﬀerentiable} K = K[a, b] = {f ∈ F : f is a constant function} B = B[a, b] = {f ∈ F : f is bounded} J = J [a, b] = {f ∈ F : f is integrable} (A function f ∈ F is bounded if there exists a number M ≥ 0 such that |f (x)| ≤ M for all x in b [a, b]. It is (Riemann) integrable if it is bounded and the Riemann integral a f (x) dx exists.) 4.1.2. Notation. For a vector space V we will write U V to indicate that U is a subspace of V . To distinguish this concept from other uses of the word “subspace” (topological subspace, for example) writers frequently use the expressions linear subspace, vector subspace, or linear manifold. 4.1.3. Definition. If A and B are subsets of a vector space then the sum of A and B, denoted by A + B, is deﬁned by A + B := {a + b : a ∈ A and b ∈ B}.

9

10

4. VECTOR SPACES AND SUBSPACES

4.2. Exercises

(Due: Wed. Jan. 14)

4.2.1. Exercise (†). Show that condition (b)(ii) in the author’s deﬁnition of vector space (AOLT, page 2) is redundant. Using only the other vector space axioms and exercise 1.2.4 prove that in fact αv = 0 if and only if α = 0 or v = 0. 4.2.2. Exercise. Let x be an element of a vector space. Using only the vector space axioms and the preceding exercises prove that (−1)x is the additive inverse of x. That is, show that (−1)x = −x. 4.2.3. Exercise. Let V be the set of all real numbers. Deﬁne an operation of “addition” by x y = the maximum of x and y α x = αx and the set V is a vector for all x, y ∈ V . Deﬁne an operation of “scalar multiplication” by for all α ∈ R and x ∈ V . Prove or disprove: under the operations space.

4.2.4. Exercise (†). Let V be the set of all real numbers x such that x > 0. Deﬁne an operation of “addition” by x y = xy for all x, y ∈ V . Deﬁne an operation of “scalar multiplication” by α x = xα and the set V is a vector for all α ∈ R and x ∈ V . Prove or disprove: under the operations space.

4.2.5. Exercise. Let V be R2 , the set of all ordered pairs (x, y) of real numbers. Deﬁne an operation of “addition” by (u, v) (x, y) = (u + x + 1, v + y + 1) α (x, y) = (αx, αy) and the set V is a vector for all (u, v) and (x, y) in V . Deﬁne an operation of “scalar multiplication” by for all α ∈ R and (x, y) ∈ V . Prove or disprove: under the operations space.

4.2.6. Exercise. Find the smallest subspace of R3 containing the vectors (2, −3, −3) and (0, 3, 2). 4.2.7. Exercise. The author’s deﬁnition of “subspace” (AOLT, page 2) is not quite correct (especially in view of the sentence following the deﬁnition). Explain why. Provide a correct deﬁnition. 4.2.8. Exercise. Let α, β, and γ be real numbers. Prove that the set of all solutions to the diﬀerential equation αy + βy + γy = 0 is a subspace of the vector space of all twice diﬀerentiable functions on R. 4.2.9. Exercise. For a ﬁxed interval [a, b], which sets of functions in the notation list 4.1.1 are vector subspaces of which? 4.2.10. Exercise (††). Let U and V be subspaces of a vector space W , and let {Vλ : λ ∈ Λ} be a (perhaps inﬁnite) family of subspaces of W . Consider the following subsets of W .

4.2. EXERCISES

(DUE: WED. JAN. 14)

11

(a) U ∩ V . (b) U ∪ V . (c) U + V . (d) U − V . (e) ∩λ∈Λ Vλ . Which of these are subspaces of W ? 4.2.11. Exercise. Let R∞ denote the (real) vector space of all sequences x = (xk ) = (xk )∞ = k=1 (x1 , x2 , x3 , . . . ) of real numbers. Deﬁne addition and scalar multiplication pointwise. That is, x + y = (x1 + y1 , x2 + y2 , x3 + y3 , . . . ) for all x, y ∈ R∞ and αx = (αx1 , αx2 , αx3 , . . . ) for all α ∈ R and x ∈ R∞ . Which of the following are subspaces of R∞ ? (a) the sequences which have inﬁnitely many zeros. (b) the sequences which have only ﬁnitely many nonzero terms. (c) the decreasing sequences. (d) the convergent sequences. (e) the sequence which converge to zero. (f) the arithmetic progressions (that is, sequences such that xk+1 − xk is constant). xk+1 is constant). (g) the geometric progressions (that is, sequences such that xk (h) the bounded sequences (that is sequences for which there is a number M > 0 such that ∞ |xk | ≤ M for all k). |xk | < ∞). (i) the absolutely summable sequences (that is, sequences such that
k=1

CHAPTER 5

Linear Combinations and Linear Independence
5.1. Background Topics: linear combinations, linear independence, span. (See AOLT, p. 2.)

5.1.1. Definition. Let A be a subset of a vector space V . The span of A is the set of all linear combinations of elements of A. It is denoted by span(A) (or by spanF (A) if we wish to emphasize the role of the scalar ﬁeld F). The subset A spans the space V if V = span(A). In this case we also say that A is a spanning set for V .

13

14

5. LINEAR COMBINATIONS AND LINEAR INDEPENDENCE

5.2. Exercises

(Due: Fri. Jan. 16)

5.2.1. Exercise. Prove that if A is a nonempty subset of a vector space V , then span A is a subspace of V . (This is the claim made in AOLT, page 2, lines 19–20.) 5.2.2. Exercise. Let A be a nonempty subset of a vector space V . What, exactly, do we mean when we speak of the smallest subspace of V which contains A? 5.2.3. Exercise. Let A be a nonempty subset of a vector space V . Then the following sets are equal: (a) span(A); (b) the intersection of the family of all subspaces of V which contain A; and (c) the smallest subspace of V which contains A. 5.2.4. Exercise. Verify that supersets of linearly dependent sets are linearly dependent and that subsets of linearly independent sets linearly independent. That is, show that if V is a vector space, A is a linearly dependent subset of V , and B is a subset of V which contains A, then B is linearly dependent. Also show that if B is a linearly independent subset of V and A is contained in B, then A is linearly independent. 5.2.5. Exercise. Let w = (1, 1, 0, 0), x = (1, 0, 1, 0), y = (0, 0, 1, 1), and z = (0, 1, 0, 1). (a) Show that {w, x, y, z} does not span (that is, generate) R4 by ﬁnding a vector u in R4 such that u ∈ span(w, x, y, z). / (b) Show that {w, x, y, z} is a linearly dependent set of vectors by ﬁnding scalars α, β, γ, and δ—not all zero—such that αw + βx + γy + δz = 0. (c) Show that {w, x, y, z} is a linearly dependent set by writing z as a linear combination of w, x, and y. 5.2.6. Exercise. In the vector space C[0, π] of continuous functions on the interval [0, π] deﬁne the vectors f , g, and h by f (x) = x g(x) = sin x h(x) = cos x for 0 ≤ x ≤ π. Show that f , g, and h are linearly independent. 5.2.7. Exercise. In the vector space C[0, π] of continuous functions on [0, π] let f , g, h, and j be the vectors deﬁned by f (x) = 1 g(x) = x h(x) = cos x x j(x) = cos2 2 for 0 ≤ x ≤ π. Show that f , g, h, and j are linearly dependent by writing j as a linear combination of f , g, and h. 5.2.8. Exercise. Let a, b, and c be distinct real numbers. Show that the vectors (1, 1, 1), (a, b, c), and (a2 , b2 , c2 ) form a linearly independent subset of R3 . 5.2.9. Exercise (†). In the vector space C[0, 1] deﬁne the vectors f , g, and h by f (x) = x g(x) = ex h(x) = e−x

5.2. EXERCISES

(DUE: FRI. JAN. 16)

15

for 0 ≤ x ≤ 1. Are f , g, and h linearly independent? 5.2.10. Exercise. Let u = (λ, 1, 0), v = (1, λ, 1), and w = (0, 1, λ). Find all values of λ which make {u, v, w} a linearly dependent subset of R3 . 5.2.11. Exercise (†)). Suppose that {u, v, w} is a linearly independent set in a vector space V . Show that the set {u + v, u + w, v + w} is linearly independent in V .

CHAPTER 6

Bases for Vector Spaces
6.1. Background Topics: basis, dimension, partial ordering, comparable elements, linear ordering, chain, maximal element, largest element, axiom of choice, Zorn’s lemma. (See AOLT, page 2 and section 1.6.)

6.1.1. Definition. A relation on a set S is a subset of the Cartesian product S × S. If the relation is denoted by ≤, then it is conventional to write x ≤ y (or equivalently, y ≥ x) rather than (x, y) ∈ ≤ . 6.1.2. Definition. A relation ≤ on a set S is reflexive if x ≤ x for all x ∈ S. It is transitive if x ≤ z whenever x ≤ y and y ≤ z. It is antisymmetric if x = y whenever x ≤ y and y ≤ x. A relation which is reﬂexive, transitive, and antisymmetric is a partial ordering. A partially ordered set is a set on which a partial ordering has been deﬁned. 6.1.3. Example. The set R of real numbers is a partially ordered set under the usual relation ≤. 6.1.4. Example. A family A of subsets of a set S is a partially ordered set under the relation ⊆. When A is ordered in this fashion it is said to be ordered by inclusion. 6.1.5. Example. Let F(S) be the family of real valued functions deﬁned on a set S. For f , g ∈ F(S) write f ≤ g if f (x) ≤ g(x) for every x ∈ S. This is a partial ordering on F(S). 6.1.6. Definition. Let A be a subset of a partially ordered set S. An element u ∈ S is an upper bound for A if a ≤ u for every a ∈ A. An element m in the partially ordered set S is maximal if there is no element of the set which is strictly greater than m; that is, m is maximal if c = m whenever c ∈ S and c ≥ m. An element m in S is the largest element of S if m ≥ s for every s ∈ S. 6.1.7. Example. Let S = {a, b, c} be a three-element set. The family P(S) of all subsets of S is partially ordered by inclusion. Then S is the largest element of P(S)—and, of course, it is also a maximal element of P(S). The family Q(S) of all proper subsets of S has no largest element; but it has three maximal elements {b, c}, {a, c}, and {a, b}. 6.1.8. Definition. A linearly independent subset of a vector space V which spans V is a (Hamel) basis for V . We make a special convention that the empty set is a basis for the trivial vector space {0}. 6.1.9. Definition. Let A be a subset of a partially ordered set S. An element l ∈ S is a lower bound for A if l ≤ a for every a ∈ A. An element l in the partially ordered set S is minimal if there is no element of the set which is strictly less than l; that is, l is minimal if c = l whenever c ∈ S and c ≤ l. An element l in S is the smallest element of S if l ≤ s for every s ∈ S.
17

18

6. BASES FOR VECTOR SPACES

6.1.10. Definition. Two elements x and y in a partially ordered set are comparable if either x ≤ y or y ≤ x. A chain (or linearly ordered subset) is a subset of a partially ordered set in which every pair of elements is comparable. 6.1.11. Axiom (Axiom of Choice). If A = {Aλ : λ ∈ Λ} is a nonempty family of nonempty pairwise disjoint sets, then there exists a function f : Λ → A such that f (λ) ∈ Aλ for every λ ∈ Λ. 6.1.12. Axiom (Zorn’s Lemma). If each chain in a nonempty partially ordered set S has an upper bound in S, then S has a maximal element. (See AOLT, page 31, lines 4–7.) The following theorem is an extremely important item in the set-theoretic foundations of mathematics. Its proof is diﬃcult and can be found in any good book on set theory. 6.1.13. Theorem. The axiom of choice and Zorn’s lemma are equivalent.

6.2. EXERCISES

(DUE: WED. JAN. 21)

19

6.2. Exercises

(Due: Wed. Jan. 21)

6.2.1. Exercise. A linearly independent subset of a vector space V is a basis for V if and only if it is a maximal linearly independent subset. (This is AOLT, proposition 1.35.) 6.2.2. Exercise (††). A spanning subset for a nontrivial vector space V is a basis for V if and only if it is a minimal spanning set for V . 6.2.3. Exercise. Let {e1 , . . . , en } be a basis for a vector space V . Then for each v ∈ V there exist unique scalars α1 , . . . , αn such that v = n αk ek . k=1 6.2.4. Exercise. Let {e1 , . . . , en } be a basis for a vector space V and v = n αk ek be a k=1 vector in V . If αp = 0 for some p, then {e1 , . . . , ep−1 , v, ep+1 , . . . , en } is a basis for V . 6.2.5. Exercise. If some basis for a vector space V contains n elements, then every linearly independent subset of V with n elements is also a basis. Hint. Suppose {e1 , . . . , en } is a basis for V and {v1 , . . . , vn } is linearly independent in V . Start by using exercise 6.2.4 to show that (after perhaps renumbering the ek ’s) the set {v1 , e2 , . . . , en } is a basis for V . 6.2.6. Exercise. In a ﬁnite dimensional vector space V any two bases have the same number of elements. (This establishes the claim made in AOLT, page 2, lines 25–27.) This number is the dimension of V , and is denoted by dim V . 6.2.7. Exercise. Let S3 be the vector space of all symmetric 3 × 3 matrices of real numbers. (A matrix [aij ] is symmetric if aij = aji for all i and j.) (a) What is the dimension of S3 ? (b) Find a basis for S3 . 6.2.8. Exercise. AOLT, section 1.8, exercise 1. 6.2.9. Exercise (†). Let U be the set of all matrices of real numbers of the form and V be the set of all real matrices of the form u −u − x 0 x

v 0 . Find bases for U, V, U + V, and U ∩ V. w −v

6.2.10. Exercise. Let A be a linearly independent subset of a vector space V . Then there exists a basis for V which contains A. (This is AOLT, Theorem 1.36. Can you think of anyway to shorten the proof?) 6.2.11. Exercise. Every vector space has a basis. (This is AOLT, Corollary 1.37.)

CHAPTER 7

Linear Transformations
7.1. Background Topics: linear maps, kernel, range, invertible linear map, isomorphism.

7.1.1. Definition. Let V and W be vector spaces over the same ﬁeld F. A function T : V → W is linear if T (x + y) = T x + T y and T (αx) = αT x for all x, y ∈ V and α ∈ F. 7.1.2. Notation. If V and W are vector spaces the family of all linear functions from V into W is denoted by L(V, W ). Linear functions are usually called linear transformations, linear maps, or linear mappings. When V = W we condense the notation L(V, V ) to L(V ) and we call the members of L(V ) operators. 7.1.3. Definition. Let T : V → W be a linear transformation between vector spaces. Then ker T , the kernel (or nullspace) of T is deﬁned to be the set of all x in V such that T x = 0. Also, ran T , the range of T (or the image of T ), is the set of all y in W such that y = T x for some x in V . The rank of T is the dimension of its range and the nullity of T is the dimension of its kernel. 7.1.4. Definition. Let T : V → W be a linear transformation between vector spaces and let U be a subspace of V . Deﬁne T → (U ) := {T x : x ∈ U }. This is the (direct) image of U under T . 7.1.5. Definition. Let T : V → W be a linear transformation between vector spaces and let U be a subspace of W . Deﬁne T ← (U ) := {x ∈ V : T x ∈ U }. This is the inverse image of U under T . 7.1.6. Definition. A linear map T : V → W between vector spaces is left invertible (or has a left inverse, or is a section) if there exists a linear map L : W → V such that LT = 1V . (Note: 1V is the identity mapping v → v on V . See AOLT, page 4, lines -13 to -10, convention. Other notations for the identity map on V are idV and IV .) The map T is right invertible (or has a right inverse, or is a retraction) if there exists a linear map R : W → V such that T R = 1W . We say that T is invertible (or has an inverse, or is an isomorphism) if it is both left invertible and right invertible. If there exists an isomorphism between two vector spaces V and W , we say that the spaces are isomorphic and we write V ∼ W . = Note: This is not the deﬁnition of isomorphism given in AOLT (deﬁnition on page 3, lines 17–18). You will prove in exercise 7.2.11 that the two deﬁnitions are equivalent.

21

22

7. LINEAR TRANSFORMATIONS

7.2. Exercises

(Due: Fri. Jan. 23)

7.2.1. Exercise. Use the notation of exercise 3.2.8 and suppose that (V, +, M ) and (W, +, M ) are vector spaces over a common ﬁeld F and that T ∈ Hom(V, W ) is such that the diagram V
Mα T

/W

V 

T 

/W

commutes for every α ∈ F. What can be said about the homomorphism T ? 7.2.2. Exercise. Let T : R3 → R3 : x → (x1 + 3x2 − 2x3 , x1 − 4x3 , x1 + 6x2 ). (a) Identify the kernel of T by describing it geometrically and by giving its equation(s). (b) Identify the range of T by describing it geometrically and by giving its equation(s). 7.2.3. Exercise. Let T be the linear map from R3 to R3 deﬁned by T (x, y, z) = (2x + 6y − 4z, 3x + 9y − 6z, 4x + 12y − 8z). Describe the kernel of T geometrically and give its equation(s). Describe the range of T geometrically and give its equation(s). 7.2.4. Exercise. Let C = C[a, b] be the vector space of all continuous real valued functions on the interval [a, b] and C 1 = C 1 [a, b] be the vector space of all continuously diﬀerentiable real valued functions on [a, b]. (Recall that a function is continuously differentiable if it has a derivative and the derivative is continuous.) Let D : C 1 → C be the linear transformation deﬁned by Df = f and let T : C → C1 be the linear transformation deﬁned by
x

(T f )(x) =
a

f (t) dt

for all f (a) (b) (c) (d)

∈ C and x ∈ [a, b]. Compute (and simplify) (DT f )(x). Compute (and simplify) (T Df )(x). Find the kernel of T . Find the range of T .

7.2.5. Exercise. Let T : V → W be a linear map between vector spaces and U V . Show that T → (U ) is a subspace of W . Conclude that the range of a linear map is a subspace of the codomain of the map. (This establishes the claim made in AOLT, page 3, lines 15–16.) 7.2.6. Exercise. Let T : V → W be a linear map between vector spaces and U W . Show that T ← (U ) is a subspace of V . Conclude that the kernel of a linear map is a subspace of the domain of the map. (This establishes the claim made in AOLT, page 3, lines 13–14.) 7.2.7. Exercise (†). Let T : V → W be a linear map between vector spaces. Show that T is injective (that is, one-to-one) if and only if ker T = {0}. 7.2.8. Exercise. Show that an operator T ∈ L(V ) is invertible if it satisﬁes the equation T 2 − T + 1V = 0. 7.2.9. Exercise. Let s be the vector space of all sequences of real numbers. (Addition and scalar multiplication are deﬁned pointwise.) Deﬁne a linear map T by T : s → s : (a1 , a2 , a3 , . . . ) → (0, a1 , a2 , . . . ). Is T left invertible? right invertible? injective? surjective?

7.2. EXERCISES

(DUE: FRI. JAN. 23)

23

7.2.10. Exercise. If a linear map T : V → W between two vector spaces has both a left inverse and a right inverse, then these inverses are equal; so there exists a linear map T −1 : W → V such that T −1 T = 1V and T T −1 = 1W . 7.2.11. Exercise. A linear map between vector spaces is invertible if and only if it is bijective (that is, one-to-one and onto).

CHAPTER 8

Linear Transformations (continued)
8.1. Background Topics: matrix representations of linear maps.

8.1.1. Convention. The space Pn ([a, b]) of polynomial functions of degree strictly less than n ∈ N on the interval [a, b] (where a < b) is a vector space of degree n. For each n = 0, 1, 2, . . . let pn (t) = tn for all t ∈ [a, b]. Then we take {p0 , p1 , p2 , . . . , pn−1 } to be the standard basis for Pn ([a, b]).

25

26

8. LINEAR TRANSFORMATIONS (CONTINUED)

8.2. Exercises

(Due: Mon. Jan. 26)

8.2.1. Exercise (†). Let P4 be the vector space of polynomials of degree less than 4. Consider the linear transformation D2 : P4 → P4 : f → f . (a) Find the matrix representation of D2 (with respect to the usual basis {1, t, t2 , t3 } for P 4 ). (b) Find the kernel of D2 . (c) Find the range of D2 . 8.2.2. Exercise. Let T : P4 → P5 be the linear transformation deﬁned by (T p)(t) = (2+3t)p(t) for every p ∈ P4 and t ∈ R. Find the matrix representation of T with respect to the usual bases for P4 and P5 . 8.2.3. Exercise. Show that an operator T on a vector space V is invertible if it has a unique right inverse. Hint. Consider ST + S − idV , where S is the unique right inverse for T . 8.2.4. Exercise. Let T : V → W be a linear map between vector spaces. If A is a subset of V such that T → (A) is linearly independent, then A is linearly independent. Can you think of a way of saying this without using any symbols? 8.2.5. Exercise (†). Let T : V → W be a linear map between vector spaces. If T is injective and A is a linearly independent subset of V , then T → (A) is a linearly independent subset of W . 8.2.6. Exercise. Let T : V → W be a linear map between vector spaces and A ⊆ V . Then T → (span A) = span T → (A). 8.2.7. Exercise. Let V and W be vector spaces over a ﬁeld F and E be a basis for V . Every function f : E → W can be extended in a unique fashion to a linear map T : V → W . 8.2.8. Exercise. Let T : V → W be a linear map between vector spaces. If T is injective and B is a basis for a subspace U of V , then T → (B) is a basis for T → (U ). 8.2.9. Exercise. Let T : V → W be a linear map between vector spaces. If V is spanned by a set B of vectors and T → (B) is a basis for W , then B is a basis for V and T is an isomorphism. 8.2.10. Exercise. Prove that a linear transformation T : R3 → R2 cannot be one-to-one and that a linear transformation S : R2 → R3 cannot be onto. What is the most general version of these assertions that you can invent (and prove)? 8.2.11. Exercise. Suppose that V and W are ﬁnite dimensional vector spaces of the same ﬁnite dimension and that T : V → W is a linear map. Then the following are equivalent; (a) T is injective; (b) T is surjective; and (c) T is invertible.

CHAPTER 9

Duality in Vector Spaces
9.1. Background Topics: linear functionals, dual space, dual basis. (See AOLT, pages 4–6.)

9.1.1. Definition. Let S be a nonempty set, F be a ﬁeld, and f be an F valued function on S. The support of f , denoted by supp f , is {x ∈ S : f (x) = 0}. 9.1.2. Notation. Let S be a nonempty set and F be a ﬁeld. We denote by l(S) (or by l(S, F), or by FS , or by F(S, F)) the family of all functions α : S → F. For x ∈ l(S) we frequently write the value of x at s ∈ S as xs rather than x(s). (Sometimes it seems a good idea to reduce the number of parentheses cluttering a page.) Furthermore we denote by lc (S) (or by lc (S, F), or by Fc (S)) the family of all functions α : S → F with ﬁnite support; that is, those functions on S which are nonzero at only ﬁnitely many elements of S.

27

28

9. DUALITY IN VECTOR SPACES

9.2. Exercises by

(Due: Wed. Jan. 28)

9.2.1. Exercise (†). Let V be a vector space (over a ﬁeld F) with basis B. Deﬁne Ω : lc (B) → V Ω(v) =
e∈B

v(e) e.

Prove that l(B) (under the usual pointwise operations) is a vector space, that lc (B) is a subspace of l(B), and that Ω is an isomorphism. 9.2.2. Exercise. A vector v in a vector space V with basis B can be written in one and only one way as a linear combination of elements in B. 9.2.3. Exercise. Let V and W be vector spaces and B be a basis for V . Every function T0 : B → W can be extended in one and only one way to a linear transformation T : V → W . 9.2.4. Convention. If in exercise 9.2.1 we write v for Ω(v) then we see that v=
e∈B

v(e) e.

(*)

The isomorphism Ω creates a one-to-one correspondence between functions v in lc (B) and vectors v in V . This is an extension of the usual notation in Rn where we write a vector v in terms of its components:
n

v=
k=1

vk ek .

We will go even further and use Ω to identify V with lc (B) and write v=
e∈B

v(e) e

instead of (*). That is, in a vector space with basis we will treat a vector as a scalar valued function on its basis. (Of course, if this identiﬁcation ever threatens to cause confusion we can always go back to the v or Ω(v) notation.) 9.2.5. Exercise. According to convention 9.2.4 above, what is the value of f (e) when e and f are elements of the basis B? 9.2.6. Exercise. Let V be a vector space with basis B. For every v ∈ V deﬁne a function v ∗ on V by v ∗ (x) = x(e) v(e) for all x ∈ V .
e∈B

Then

v∗

is a linear functional on V .

9.2.7. Notation. In the preceding exercise 9.2.6 the value v ∗ (x) of v ∗ at x is often written x, v . 9.2.8. Exercise. Consider the notation 9.2.7 above in the special case that the scalar ﬁeld F = R. Then , is an inner product on the vector space V . (See the deﬁnition of inner product on page 11 of AOLT.) 9.2.9. Exercise. In the special case that the scalar ﬁeld F = C, things above are usually done a bit diﬀerently. For v ∈ V the function v ∗ is deﬁned by v ∗ (x) = x, v =
e∈B

x(e) v(e).

Why do you think things are done this way?

9.2. EXERCISES

(DUE: WED. JAN. 28)

29

9.2.10. Exercise. Let v be a nonzero vector in a vector space V and E be a basis for V which contains the vector v. Then there exists a linear functional φ ∈ V ∗ such that φ(v) = 1 and φ(e) = 0 for every e ∈ E \ {v}. 9.2.11. Corollary. If v is a vector in a vector space V and φ(v) = 0 for every φ ∈ V ∗ , then v = 0. 9.2.12. Exercise. Let V be a vector space with basis B. The map Φ : V → V ∗ : v → v ∗ (see exercise 9.2.6) is linear and injective. 9.2.13. Exercise (Riesz-Fr´chet theorem for vector spaces). (†) In the preceding exercise 9.2.12 e the map Φ is an isomorphism if V is ﬁnite dimensional. Thus for every φ ∈ V ∗ there exists a unique vector a ∈ V such that a∗ = φ. 9.2.14. Exercise. If V is a ﬁnite dimensional vector space with basis {e1 , . . . , en }, then {e∗ , . . . , e∗ } n 1 is the dual basis for V ∗ . (See AOLT, page 5, line -7 and Proposition 1.1.)

CHAPTER 10

Duality in Vector Spaces (continued)
10.1. Background Topics: annihilators, pre-annihilators, trace. (See AOLT, page 7.)

10.1.1. Notation. Let V be a vector space and M ⊆ V . Then M ⊥ := {f ∈ V ∗ : f (x) = 0 for all x ∈ M } We say that M ⊥ is the annihilator of M . (The reasons for using the familiar “orthogonal complement” notation M ⊥ will become apparent when we study inner product spaces, where “orthogonality” actually makes sense.) 10.1.2. Notation. Let V be a vector space and F ⊆ V ∗ . Then F⊥ := {x ∈ V : f (x) = 0 for all f ∈ F } We say that F⊥ is the pre-annihilator of F .

31

32

10. DUALITY IN VECTOR SPACES (CONTINUED)

10.2. Exercises

(Due: Fri. Jan. 30)

10.2.1. Exercise. In exercise 9.2.12 we showed that the map Φ : V → V ∗ : v → v∗ is always an injective linear map. In exercise 9.2.13 we showed that if V is ﬁnite dimensional, then so is V ∗ and Φ is an isomorphism between V and V ∗ . Now prove that if V is inﬁnite dimensional, then Φ is not an isomorphism. Hint. Let B be a basis for V . Is there a functional ψ ∈ V ∗ such that ψ(e) = 1 for every e ∈ B? Could such a functional be Φ(x) for some x ∈ V ? 10.2.2. Exercise (†). Let V be a vector space over a ﬁeld F. For every x in V deﬁne x : V ∗ → F : φ → φ(x) . (a) Show that x ∈ V ∗∗ for each x ∈ V . (b) Let ΓV be the map from V to V ∗∗ which takes x to x. (When no confusion is likely we write Γ for ΓV , so that Γ(x) = x for each x ∈ V .) Prove that Γ is linear. (c) Prove that Γ is injective. (See AOLT, page 33, exercise 5.) 10.2.3. Exercise. Prove that if V is a ﬁnite dimensional vector space, then the map Γ : V → V ∗∗ deﬁned in the preceding exercise 10.2.2 is an isomorphism. Prove also that if V is inﬁnite dimensional, then Γ is not an isomorphism. Hint. Let B be a basis for V and ψ ∈ V ∗ be as in exercise 10.2.1. Show that if we let C0 be {e∗ : e ∈ B}, then the set C0 ∪ {ψ} is linearly independent and can therefore be extended to a basis C for V ∗ . Find an element τ in V ∗∗ such that τ (ψ) = 1 and τ (φ) = 0 for every other φ ∈ C. Can τ be Γx for some x ∈ V ? (If x = 0 in V = l0 (B), then there is a basis vector f in B such that x(f ) = 0.) 10.2.4. Exercise (†). Find the annihilator in R2 ∗ answer in terms of the standard dual basis for R2 .)

of the vector (1, 1) in R2 . (Express your

10.2.5. Exercise. Let M and N be subsets of a vector space V . Then (a) M ⊥ is a subspace of V ∗ . (b) If M ⊆ N , then N ⊥ M ⊥ . (c) (span M )⊥ = M ⊥ . (d) (M ∪ N )⊥ = M ⊥ ∩ N ⊥ . 10.2.6. Exercise. If M and N be subspaces of a vector space V , then (M + N )⊥ = M ⊥ ∩ N ⊥ . Explain why it is necessary in this exercise to assume that M and N are subspaces of V and not just subsets of V . Hint. Is it always true that span (M + N ) = span M + span N ? 10.2.7. Exercise. In exercises 10.2.5 and 10.2.6 you established some properties of the annihilator mapping M → M ⊥ . See to what extent you can prove similar results about the pre-annihilator mapping F → F⊥ . What can you say about the sets M ⊥ ⊥ and F⊥ ⊥ ?

CHAPTER 11

The Language of Categories
11.1. Background Topics: categories, objects, morphisms, functors, vector space adjoint of a linear map.

11.1.1. Definition. Let A be a class, whose members we call objects. For every pair (S, T ) of objects we associate a set Mor(S, T ), whose members we call morphisms from S to T . We assume that Mor(S, T ) and Mor(U, V ) are disjoint unless S = U and T = V . We suppose further that there is an operation ◦ (called composition) that associates with every α ∈ Mor(S, T ) and every β ∈ Mor(T, U ) a morphism β ◦ α ∈ Mor(S, U ) in such a way that: (1) γ ◦ (β ◦ α) = (γ ◦ β) ◦ α whenever α ∈ Mor(S, T ), β ∈ Mor(T, U ), and γ ∈ Mor(U, V ); (2) for every object S there is a morphism IS ∈ Mor(S, S) satisfying α ◦ IS = α whenever α ∈ Mor(S, T ) and IS ◦ β = β whenever β ∈ Mor(R, S). Under these circumstances the class A, together with the associated families of morphisms, is a category. For a few more remarks on categories and some examples see section 8.1 of my notes [11]. 11.1.2. Definition. If A and B are categories a (covariant) functor F from A to B
/ B) is a pair of maps: an object map F which associates with each object S in (written A A an object F (S) in B and a morphism map (also denoted by F ) which associates with each morphism f ∈ Mor(S, T ) in A a morphism F (f ) ∈ Mor(F (S), F (T )) in B, in such a way that (1) F (g ◦ f ) = F (g) ◦ F (f ) whenever g ◦ f is deﬁned in A; and (2) F (idS ) = idF (S) for every object S in A.
F

/ B diﬀers from the preceding deﬁnition The deﬁnition of a contravariant functor A only in that, ﬁrst, the morphism map associates with each morphism f ∈ Mor(S, T ) in A a morphism F (f ) ∈ Mor(F (T ), F (S)) in B and, second, condition (1) above is replaced by (1 ) F (g ◦ f ) = F (f ) ◦ F (g) whenever g ◦ f is deﬁned in A. For a bit more on functors and some examples see section 15.4 of my notes [11].
F

11.1.3. Definition. The terminology for inverses of morphisms in categories is essentially the
/ T and T / S be morphisms in a category. If β ◦ α = IS , then β same as for functions. Let S is a left inverse of α and, equivalently, α is a right inverse of β. We say that the morphism
α β / S which is both a left α is an isomorphism (or is invertible) if there exists a morphism T −1 and is called the inverse of α. and a right inverse for α. Such a function is denoted by α β

11.1.4. Remark. All the categories that are of interest in this course are concrete categories. A concrete category is, roughly speaking, one in which the objects are sets with additional structure (algebraic operations, inner products, norms, topologies, and the like) and the morphisms are maps (functions) which preserve, in some sense, the additional structure. If A is an object in some concrete category C, we denote by A its underlying set. And if A
33 f

/ B is a morphism

34

11. THE LANGUAGE OF CATEGORIES

in C we denote by f the map from A to B regarded simply as a function between sets. It is easy to see that , which takes objects in C to objects in Set (the category of sets and maps) and morphisms in C to morphisms in Set, is a functor. It is referred to as a forgetful functor. In the category Vec of vector spaces and linear maps, for example, causes a vector space V to “forget” about its addition and scalar multiplication ( V is just a set). And if T : V → W is a linear transformation, then T : V → W is just a map between sets—it has “forgotten” about preserving the operations. For more precise deﬁnitions of concrete categories and forgetful functors consult any text on category theory. 11.1.5. Definition. Let T : V → W be a linear map between vector spaces. For every g ∈ W ∗ let T ∗ (g) = g T . Notice that T ∗ (g) ∈ V ∗ . The map T ∗ from the vector space W ∗ into the vector space V ∗ is the (vector space) adjoint map of T . WARNING: In inner product spaces we will use the same notation T ∗ for a diﬀerent map. If T : V → W is a linear map between inner product spaces, then the (inner product space) adjoint transformation T ∗ maps W to V (not W ∗ to V ∗ ). 11.1.6. Definition. A partially ordered set is order complete if every nonempty subset has a supremum (that is, a least upper bound) and an inﬁmum (a greatest lower bound). 11.1.7. Definition. Let S be a set. Then the power set of S, denoted by P(S), is the family of all subsets of S. 11.1.8. Notation. Let f : S → T be a function between sets. Then we deﬁne f → (A) = {f (x) : x ∈ A} and f ← (B) = {x ∈ S : f (x) ∈ B}. We say that f → (A) is the image of A under f and that f ← (B) is the preimage of B under f .

11.2. EXERCISES

(DUE: MON. FEB. 2)

35

11.2. Exercises

(Due: Mon. Feb. 2)

11.2.1. Exercise. Prove that if a morphism in some category has both a left and a right inverse, then it is invertible. 11.2.2. Exercise (The vector space duality functor). Let T ∈ L(V, W ) where V and W are vector spaces. Show that the pair of maps V → V ∗ and T → T ∗ is a contravariant functor from the category of vector spaces and linear maps into itself. Show that (the morphism map of) this functor is linear. (That is, show that (S + T )∗ = S ∗ + T ∗ and (αT )∗ = αT ∗ .) 11.2.3. Exercise (†). Let T : V → W be a linear map between vector spaces. Show that ker T ∗ = (ran T )⊥ . Is there a relationship between T being surjective and T ∗ being injective? 11.2.4. Exercise. Let T : V → W be a linear map between vector spaces. Show that ker T = (ran T ∗ ) ⊥ . Is there a relationship between T being injective and T ∗ being surjective? 11.2.5. Exercise (The power set functors). Let S be a nonempty set. (a) The power set P(S) of S partially ordered by ⊆ is order complete. (b) The class of order complete partially ordered sets and order preserving maps is a category. (c) For each function f between sets let P(f ) = f → . Then P is a covariant functor from the category of sets and functions to the category of order complete partially ordered sets and order preserving maps. (d) For each function f between sets let P(f ) = f ← . Then P is a contravariant functor from the category of sets and functions to the category of order complete partially ordered sets and order preserving maps.

CHAPTER 12

Direct Sums
12.1. Background Topics: internal direct sum, external direct sum. (See AOLT, pages 8–9.)

12.1.1. Definition. Let M and N be subspaces of a vector space V . We say that the space V is the (internal) direct sum of M and N if M + N = V and M ∩ N = {0}. In this case we write V = M ⊕ N and say that M and N are complementary subspaces. We also say that M is a complement of N and that N is a complement of M . Similarly, if M1 , . . . , Mn are subspaces of a vector space V , if V = M1 + · · · + Mn , and if Mj ∩ Mk = {0} whenever j = k, then we say that V is the (internal) direct sum of M1 , . . . , Mn and we write
n

V = M1 ⊕ · · · ⊕ Mn =
k=1

Mk .

12.1.2. Notation. If the vector space V is a direct sum of subspaces M and N , then, according to exercise 12.2.1, every v ∈ V can be written uniquely in the form m + n where m ∈ M and n ∈ N . We will indicate this by writing v = m ⊕ n. It is important to realize that this notation makes sense only when we have in mind a particular direct sum decomposition M ⊕ N of V . Thus in this context when we see v = a ⊕ b we conclude that v = a + b, that a ∈ M , and that b ∈ N . 12.1.3. Definition. Let V and W be vector spaces over a ﬁeld F. To make the Cartesian product V × W into a vector space we deﬁne addition by (v, w) + (v1 , w1 ) = (v + v1 , w + w1 ) (where v, v1 ∈ V and w, w1 ∈ W ), and we deﬁne scalar multiplication by α(v, w) = (αv, αw) (where α ∈ F, v ∈ V , and w ∈ W ). The resulting vector space we call the (external) direct sum of V and W . It is conventional to use the same notation V ⊕ W for external direct sums that we use for internal direct sums. 12.1.4. Notation. If S, T , and U are nonempty sets and if f : S → T and g : S → U , then we deﬁne the function (f, g) : S → T × U by (f, g)(s) = (f (s), g(s)). Suppose, on the other hand, that we are given a function h mapping S into the Cartesian product T × U . Then for each s ∈ S the image h(s) is an ordered pair, which we will write as h1 (s), h2 (s) . (The superscripts have nothing to do with powers.) Notice that we now have functions h1 : S → T and h2 : S → U . These are the components of h. In abbreviated notation h = (h1 , h2 ). The (external) direct sum of two vector spaces V1 and V2 is best thought of not just as the vector space V1 ⊕V2 deﬁned in 12.1.3 but as this vector space together with two distinguished “projection” mappings deﬁned below. (The reason for this assertion may be deduced from deﬁnition 13.1.1 and exercise 13.2.1.)
37

38

12. DIRECT SUMS

12.1.5. Definition. Let V1 and V2 be vector spaces. For k = 1, 2 deﬁne the coordinate projections πk : V1 ⊕ V2 → Vk by πk (v1 , v2 ) = vk . Notice two simple facts: (i) π1 and π2 are surjective linear maps; and (ii) idV1 ⊕V2 = (π1 , π2 ).

12.2. EXERCISES

(DUE: WED. FEB. 4)

39

12.2. Exercises

(Due: Wed. Feb. 4)

12.2.1. Exercise. If V is a vector space and V = M ⊕ N , then for every v ∈ V there exist unique vectors m ∈ M and n ∈ N such that v = m + n. 12.2.2. Exercise. Suppose that a vector space V is the direct sum of subspaces M1 , . . . , Mn . Show that for every v ∈ V there exist unique vectors mk ∈ Mk (for k = 1, . . . , n) such that v = m1 + · · · + mn . 12.2.3. Exercise. Let U be the plane x + y + z = 0 and V be the line x = y = z in R3 . The purpose of this problem is to conﬁrm that R3 = U ⊕ V . This requires establishing three things: (i) U and V are subspaces of R3 (which is very easy and which we omit); (ii) R3 = U + V ; and (iii) U ∩ V = {0}. (a) To show that R3 = U + V we need R3 ⊆ U + V and U + V ⊆ R3 . Since U ⊆ R3 and V ⊆ R3 , it is clear that U + V ⊆ R3 . So all that is required is to show that R3 ⊆ U + V . That is, given a vector x = (x1 , x2 , x3 ) in R3 we must ﬁnd vectors u = (u1 , u2 , u3 ) in U and v = (v1 , v2 , v3 ) in V such that x = u + v. Find two such vectors. (b) The last thing to verify is that U ∩ V = {0}; that is, that the only vector U and V have in common is the zero vector. Suppose that a vector x = (x1 , x2 , x3 ) belongs to both U and V . Since x ∈ U it must satisfy the equation x1 + x2 + x3 = 0. since x ∈ V it must satisfy the equations x1 = x2 and x2 = x3 . (2) Solve the system of equations (1) and (2). 12.2.4. Exercise. Let U be the plane x + y + z = 0 and V be the line x = − 3 y = 3z. The 4 purpose of this exercise is to see (in two diﬀerent ways) that R3 is not the direct sum of U and V . (a) If R3 were equal to U ⊕ V , then U ∩ V would contain only the zero vector. Show that this is not the case by ﬁnding a vector x = 0 in R3 such that x ∈ U ∩ V . (b) If R3 were equal to U ⊕ V , then, in particular, we would have R3 = U + V . Since both U and V are subsets of R3 , it is clear that U + V ⊆ R3 . Show that the reverse inclusion R3 ⊆ U + V is not correct by ﬁnding a vector x ∈ R3 which cannot be written in the form u + v where u ∈ U and v ∈ V . (c) We have seen in part (b) that U + V = R3 . Then what is U + V ?
1 12.2.5. Exercise. Let U be the plane x + y + z = 0 and V be the line x − 1 = 2 y = z + 2 in R3 . State in one short sentence how you know that R3 is not the direct sum of U and V .

(1)

12.2.6. Exercise (††). Let C = C([0, 1]) be the (real) vector space of all continuous real valued functions on the interval [0, 1] with addition and scalar multiplication deﬁned pointwise. (a) Let f1 (t) = t and f2 (t) = t4 for 0 ≤ t ≤ 1. Let U be the set of all functions of the form αf1 + βf2 where α, β ∈ R. Show that U is a subspace of C. (b) Let V be the set of all functions g in C which satisfy
1 1

tg(t) dt = 0
0

and
0

t4 g(t) dt = 0.

Show that V is a subspace of C. (c) Show that C = U ⊕ V . (d) Let f (t) = t2 for 0 ≤ t ≤ 1. Find functions u ∈ U and v ∈ V such that f = u + v.

40

12. DIRECT SUMS

12.2.7. Exercise. Let C = C[−1, 1] be the vector space of all continuous real valued functions on the interval [−1, 1]. A function f in C is even if f (−x) = f (x) for all x ∈ [−1, 1]; it is odd if f (−x) = −f (x) for all x ∈ [−1, 1]. Let Co = {f ∈ C : f is odd } and Ce = {f ∈ C : f is even }. Show that C = Co ⊕ Ce . 12.2.8. Exercise. Prove that the external direct sum of two vector spaces (as deﬁned in 12.1.3) is indeed a vector space. 12.2.9. Exercise. Let V1 and V2 be vector spaces. Prove that following fact about the direct sum V1 ⊕ V2 : for every vector space W and every pair of linear maps T1 : W → V1 and T2 : W → V2 there exists a unique linear map S : W → V1 ⊕ V2 which makes the following diagram commute. W?    ??? ?? T2  T1  ?? S   ??  ??     / V2 V 1 ⊕ V2 V1 o
π1 π2

12.2.10. Exercise. Every subspace of a vector space has a complement. That is, if M is a subspace of a vector space V , then there exists a subspace N of V such that V = M ⊕ N . 12.2.11. Exercise. Let V be a vector space and suppose that V = U ⊕ W . Prove that if B is a basis for U and C is a basis for W , then B ∪ C is a basis for V . From this conclude that dim V = dim U + dim W . 12.2.12. Exercise (†). Let U , V , and W be vector spaces. Show that if U ∼ W , then U ⊕ V ∼ = = W ⊕ V . Show also that the converse of this assertion need not be true. 12.2.13. Exercise. A linear transformation has a left inverse if and only if it is injective (oneto-one). It has a right inverse if and only if it is surjective (onto).

CHAPTER 13

Products and Quotients
13.1. Background Topics: products, coproducts, quotient spaces, exact sequences. (See AOLT, pages 9–10.)

13.1.1. Definition. Let A1 and A2 be objects in a category C. We say that a triple (P, π1 , π2 ), where P is an object and πk : P → Ak (k = 1, 2) are morphisms, is a product of A1 and A2 if for every object B in C and every pair of morphisms fk : B → Ak (k = 1, 2) there exists a unique map g : B → P such that fk = πk ◦ g for k = 1, 2. B?    ??? ?? f2  f1  ?? g   ??  ??     / A2 A1 o P
π1 π2

Notice that what you showed in exercise 12.2.9 is that the external direct sum is a product in the category of vector spaces and linear maps. A triple (P, j1 , j2 ), where P is an object and jk : Ak → P , (k = 1, 2) are morphisms, is a coproduct of A1 and A2 if for every object B in C and every pair of morphisms Fk : Ak → B (k = 1, 2) there exists a unique map G : P → B such that Fk = G ◦ jk for k = 1, 2. B ? O _??   ???  F1  ??F2  G ??   ??   ?  /P o A2 A1
j1 j2

13.1.2. Definition. Let M be a subspace of a vector space V . Deﬁne an equivalence relation ∼ on V by x∼y if and only if y − x ∈ M. For each x ∈ V let [x] be the equivalence class containing x. Let V /M be the set of all equivalence classes of elements of V . For [x] and [y] in V /M deﬁne [x] + [y] := [x + y] and for α ∈ R and [x] ∈ V /M deﬁne α[x] := [αx]. Under these operations V /M becomes a vector space. It is the quotient space of V by M . The notation V /U is usually read “V mod M ”. The linear map π : V → V /M : x → [x] is called the quotient map.
41

42

13. PRODUCTS AND QUOTIENTS

13.1.3. Definition. A sequence of vector spaces and linear maps ···
/ Vn−1
jn

/ Vn

jn+1

/ Vn+1

/ ···

is said to be exact at Vn if ran jn = ker jn+1 . A sequence is exact if it is exact at each of its constituent vector spaces. A sequence of vector spaces and linear maps of the form 0
/U
j

/V

k

/W

/0

is a short exact sequence. (Here 0 denotes the trivial 0-dimensional vector space, and the unlabeled arrows are the obvious linear maps.)

13.2. EXERCISES

(DUE: MON. FEB. 9)

43

13.2. Exercises

(Due: Mon. Feb. 9)

13.2.1. Exercise. Show that in the category of vector spaces and linear maps the external direct sum is not only a product but also a coproduct. 13.2.2. Exercise (††). Show that in the category of sets and maps (functions) the product and the coproduct are not the same. 13.2.3. Exercise. Show that in a arbitrary category (or in the category of vector spaces and linear maps, if you prefer) that products (and coproducts) are essentially unique. (Essentially unique means unique up to isomorphism. That is, if (P, π1 , π2 ) and (Q, ρ1 , ρ2 ) are both products of two given objects, then P ∼ Q.) = 13.2.4. Exercise (†). Verify the assertions made in deﬁnition 13.1.2. In particular, show that ∼ is an equivalence relation, that addition and scalar multiplication of the set of equivalence classes is well deﬁned, that under these operations V /M is a vector space, and that the quotient map is linear. The following exercise is called the fundamental quotient theorem or the ﬁrst isomorphism theorem for vector spaces. (See AOLT, theorem 1.6, page 10.) 13.2.5. Exercise. Let V and W be vector spaces and M V . If T ∈ L(V, W ) and ker T ⊇ M , then there exists a unique T ∈ L(V /M , W ) which makes the following diagram commute.
CC CC CC T CC π CC CC CC C!  /W V /M

V C C

T Furthermore, T is injective if and only if ker T = U ; and T is surjective if and only if T is. Corollary: ran T ∼ V / ker T . = 13.2.6. Exercise. The sequence 0
/U
j

/V

k

/W

/0

of vector spaces is exact at U if and only if j is injective. It is exact at W if and only if k is surjective. 13.2.7. Exercise. Let U and V be vector spaces. Then the following sequence is short exact: 0
/U
ι1

/U ⊕V

π2

/V

/ 0.

The indicated linear maps are the obvious ones: ι1 : U → U ⊕ V : u → (u, 0) and π2 : U ⊕ V → V : (u, v) → v.

44

13. PRODUCTS AND QUOTIENTS

13.2.8. Exercise. Suppose a < b. Let K be the family of constant functions on the interval [a, b], C 1 be the family of all continuously diﬀerentiable functions on [a, b], and C be the family of all continuous functions on [a, b]. (A function f is said to be continuously differentiable if its derivative f exists and is continuous.) Specify linear maps j and k so that the following sequence is short exact: 0
/K
j

/ C1

k

/C

/ 0.

13.2.9. Exercise. Let C be the family of all continuous functions on the interval [0, 2]. Let E1 be the mapping from C into R deﬁned by E1 (f ) = f (1). (The functional E1 is called evaluation at 1.) Find a subspace F of C such that the following sequence is short exact. 0
/F
ι

/C

E1

/R

/ 0.

13.2.10. Exercise. If j : U → V is an injective linear map between vector spaces, then the sequence 0 is exact. 13.2.11. Exercise. Consider the following diagram in the category of vector spaces and linear maps. 0
/U
f j

/U

j

/V

π

/ V / ran j

/0

/V
g

k

0 

/U

j

/V 

k

/W

/W   h 

/0

/0

If the rows are exact and the left square commutes, then there exists a unique linear map h : W → W which makes the right square commute. 13.2.12. Exercise. Consider the following diagram of vector spaces and linear maps 0
/U
f j

/V
g

k

/W
h

/0

0 

/U

j

/V 

k 

/W

/0

where the rows are exact and the squares commute. Prove the following. (a) If g is surjective, so is h. (b) If f is surjective and g is injective, then h is injective. (c) If f and h are surjective, so is g. (d) If f and h are injective, so is g.

CHAPTER 14

Products and Quotients (continued)
14.1. Background Topics: quotient map in a category, quotient object.

14.1.1. Definition. Let A be an object in a concrete category C. A surjective morphism π / B in C is a quotient map for A if a function g : B → C (in SET) is a morphism (in C) A whenever g ◦ π is a morphism. An object B in C is a quotient object for A if it is the range of some quotient map for A.

45

46

14. PRODUCTS AND QUOTIENTS (CONTINUED)

14.2. Exercises
j

(Due: Wed. Feb. 11)
k

/U /V /W /0 is an exact sequence of vector spaces 14.2.1. Exercise. Show that if 0 ∼ U ⊕ W . Hint. Consider the following diagram and use exercise 13.2.12. and linear maps, then V =

0

/U /U

j

0

i1

/ U ⊕W o

/V  g 

k

/W /

/0. /0

π2 i2

W

The trick is to ﬁnd the right map g. 14.2.2. Exercise (†). Prove the converse of the preceding exercise. That is, suppose that U , V , and W are vector spaces and that V ∼ U ⊕ W ; prove that there exists linear maps j and k such =
/U j /V k /W that the sequence 0 isomorphism. Deﬁne j and k in terms of g.
j

/ 0 is exact. Hint. Suppose g : U ⊕ W → V is an

/U /V /W / 0 is an exact sequence of vector 14.2.3. Exercise. Show that if 0 ∼ V / ran j. Thus, if U spaces and linear maps, then W = V and j is the inclusion map, then W ∼ V /U . Give two diﬀerent proofs of this result: one using exercise 13.2.5 and the other using = exercise 13.2.12.
k

14.2.4. Exercise. Prove the converse of exercise 14.2.3. That is, suppose that j : U → V is an injective linear map between vector spaces and that W ∼ V / ran j; prove that there exists a linear = map k which makes the sequence 0
/U
j

/V

k

/W

/ 0 exact.

14.2.5. Exercise. Let W be a vector space and M V W . Then ∼ W/V. (W/M ) (V /M ) = Hint. Exercise 14.2.3 14.2.6. Exercise. Let M be a subspace of a vector space V . Then the following are equivalent: (a) dim V /M < ∞ ; (b) there exists a ﬁnite dimensional subspace F of V such that V = M ⊕ F ; and (c) there exists a ﬁnite dimensional subspace F of V such that V = M + F . 14.2.7. Exercise. Suppose that a vector space V is the direct sum of subspaces U and W . Some authors deﬁne the codimension of U to be dim W . Others deﬁne it to be dim V /U . Show that these are equivalent. 14.2.8. Exercise. Prove that if M is a ﬁnite dimensional subspace of a vector space V , then dim V /M = dim V − dim M . 14.2.9. Exercise (††). Prove that in the category of vector spaces and linear maps every surjective linear map is a quotient map. 14.2.10. Exercise. Let U , V , and W be vector spaces. If S ∈ L(U, V ), T ∈ L(V, W ), then the sequence 0 is exact.
/ ker S / ker T S / ker T / coker S / coker T S / coker T /0

14.2. EXERCISES

(DUE: WED. FEB. 11)

47

For obvious reasons the next result is usually called the rank-plus-nullity theorem. (See AOLT, theorem 1.8, page 10.) 14.2.11. Exercise. Let T : V → W be a linear map between vector spaces. Give a very simple proof that if V is ﬁnite dimensional, then rank T + nullity T = dim V. 14.2.12. Exercise. Show that if V0 , V1 , . . . , Vn are ﬁnite dimensional vector spaces and the sequence 0
n

/ Vn

dn

/ Vn−1

/ ...

/ V1

d1

/ V0

/0

is exact, then
k=0

(−1)k dim Vk = 0.

CHAPTER 15

Projection Operators
15.1. Background Topics: idempotent maps, projections. (See AOLT, section 3.2, pages 81–86.) If you have not yet discovered Professor Farenick’s web page where he has a link to a list of corrections to AOLT, this might be a good time to look at it. See http://www.math.uregina.ca/~farenick/fixups.pdf

15.1.1. Definition. Let V be a vector space. An operator E ∈ L(V ) is a projection operator if it is idempotent; that is, if E 2 = E. 15.1.2. Definition. Let V be a vector space and suppose that V = M ⊕ N . We know from an earlier exercise 12.2.1 that for each v ∈ V there exist unique vectors m ∈ M and n ∈ N such that v = m + n. Deﬁne a function EN M : V → V by EN M v = m. The function EN M is called the projection of V along N onto M . (This terminology is, of course, optimistic. We must prove that EN M is in fact a projection operator.) 15.1.3. Definition. Let M1 ⊕ · · · ⊕ Mn be a direct sum decomposition of a vector space V . For each k ∈ Nn let Nk be the following subspace of V complementary to Mk : Nk := M1 ⊕ · · · ⊕ Mk−1 ⊕ Mk+1 ⊕ · · · ⊕ Mn . Also (for each k) let Ek := EN
k Mk

be the projection onto Mk along the complementary subspace Nk . The projections E1 , . . . En are the projections associated with the direct sum decomposition V = M1 ⊕ · · · ⊕ Mn .

49

50

15. PROJECTION OPERATORS

15.2. Exercises

(Due: Fri. Feb. 13)

15.2.1. Exercise (†). If E is a projection operator on a vector space V , then V = ran E ⊕ ker E. 15.2.2. Exercise. Let V be a vector space and E, F ∈ L(V ). If E + F = idV and EF = 0, then E and F are projection operators and V = ran E ⊕ ran F . 15.2.3. Exercise. Let V be a vector space and E1 , . . . , En ∈ L(V ). If n Ek = idV and k=1 Ei Ej = 0 whenever i = j, then each Ek is a projection operator and V = n ran Ek . k=1 15.2.4. Exercise. If E is a projection operator on a vector space V , then ran E = {x ∈ V : Ex = x}. 15.2.5. Exercise. Let E and F be projection operators on a vector space V . Then E +F = idV if and only if EF = F E = 0 and ker E = ran F . 15.2.6. Exercise (†). If M ⊕ N is a direct sum decomposition of a vector space V , then the function EN M deﬁned in 15.1.2 is a projection operator whose range is M and whose kernel is N . 15.2.7. Exercise. If M ⊕ N is a direct sum decomposition of a vector space V , then EN M + EM N = idV and EN M EM N = 0. 15.2.8. Exercise. If E is a projection operator on a vector space V , then there exist M , N such that E = EN M . V

15.2.9. Exercise. Let M be the line y = 2x and N be the y-axis in R2 . Find [EM N ] and [EN M ]. 15.2.10. Exercise. Let E be the projection of R3 onto the plane 3x − y + 2z = 0 along the z-axis and let F be the projection of R3 onto the z-axis along the plane 3x − y + 2z = 0. (a) Find [E]. (b) Where does F take the point (4, 5, 1)? 15.2.11. Exercise. Let P be the plane in R3 whose equation is x − y − 2z = 0 and L be the line whose equations are x = 0 and y = −z. Let E be the projection of R3 along L onto P and F be the projection of R3 along P onto L. Then     b b b a b b c  [E] = −a c and [F ] =  a −a −c a −a −a −a a c where a = ,b= , and c = .

15.2.12. Exercise. Suppose a ﬁnite dimensional vector space V has the direct sum decomposition V = M ⊕ N and that E = EM N is the projection along M onto N . Show that E ∗ is the projection in L(V ∗ ) along N ⊥ onto M ⊥ .

CHAPTER 16

Algebras
16.1. Background Topics: algebras, matrix algebras, standard matrix units, ideals, center, central algebra, polynomial functional calculus, quaternions, representation, faithful representation, left regular representation, permutation matrices. (See AOLT, sections 2.1–2.3.)

16.1.1. Definition. Let A be an algebra over a ﬁeld F. The unitization of A is the unital alge˜ bra A = A × F in which addition and scalar multiplication are deﬁned pointwise and multiplication is deﬁned by (a, λ) · (b, µ) = (ab + µa + λb, λµ).

51

52

16. ALGEBRAS

16.2. Exercises

(Due: Mon. Feb. 16)

16.2.1. Exercise. In AOLT at the top of page 43 the author says that the ideal generated by a subset S of an algebra A is the intersection of all ideals containing S. For this to be correct, we must know that the intersection of a family of ideals in A is itself an ideal. Prove this. 16.2.2. Exercise. Also in AOLT at the top of page 43 the author says that the ideal generated by a subset S of an algebra A can be characterized by as
m

aj sj bj : m = 1, 2, 3, . . . , sj ∈ S, and aj , bj ∈ A .
j=1

Prove that for a unital algebra this is correct.. 16.2.3. Exercise. In AOLT at the bottom of page 43 the author constructs the Cartesian product algebra A1 × · · · × An of algebras A1 , . . . An . Show that this is indeed a product in the category of algebras and algebra homomorphisms. Under what circumstances will the product algebra A1 × · · · × An be unital? 16.2.4. Exercise (†). Prove that product algebra Mn (F)×Mn (F) is neither simple nor central. (See AOLT, section 2.7, exercise 3.) 16.2.5. Exercise. AOLT, section 2.7, exercise 4. 16.2.6. Exercise. In AOLT at the bottom of page 44 and the top of page 45 the author constructs the quotient algebra A/J, where A is an algebra and J is an ideal in A. Prove that the operations on A/J are well-deﬁned and that under these operations A/J is in fact an algebra. 16.2.7. Exercise (†). An element a of a unital algebra A is invertible if there exists an element a−1 ∈ A such that aa−1 = a−1 a = 1A . Explain why no proper ideal in a unital algebra can contain an invertible element. 16.2.8. Exercise. Let φ : A → B be an algebra homomorphism. Prove that the kernel of φ is an ideal in A and that the range of φ is a subalgebra of B. ˜ 16.2.9. Exercise. Prove that the unitization A of an algebra A is in fact a unital algebra with ˜ (0, 1) as its identity. Prove also that A is (isomorphic to) a subalgebra of A with codimension 1. (See AOLT, pages 49–50.) 16.2.10. Exercise. Let A and B be algebras and J be an ideal in A. If φ : A → B is an algebra homomorphism and ker φ ⊇ J, then there exists a unique algebra homomorphism φ : A/J → B which makes the following diagram commute.
AA AA AA φ AA AA π AA AA AA  /B A/J

AA

φ Furthermore, φ is injective if and only if ker φ = U ; and φ is surjective if and only if φ is. Corollary: ran φ ∼ A/ ker φ. (See AOLT, page 45, Theorem 2.5.) =

16.2. EXERCISES

(DUE: MON. FEB. 16)

53

16.2.11. Exercise. Let V be a vector space, v and w be in V and φ and ψ be in V ∗ . Then (v ⊗ φ)(w ⊗ ψ) = φ(w)(v ⊗ ψ). 16.2.12. Exercise. Let V be a vector space, v be in V and φ and ψ be in V ∗ . Then (v ⊗ φ)∗ (ψ) = ψ(v)φ.

CHAPTER 17

Spectra
17.1. Background Topics: eigenvalues, spectrum, annihilating polynomial, minimal polynomial, division algebra, centralizer, skew-centralizer, spectral mapping theorem. (See AOLT, sections 2.4–2.5.)

17.1.1. Definition. Let a be an element of a unital algebra A over a ﬁeld F. The spectrum of a, denoted by σA (a) or just σ(a), is the set of all λ ∈ F such that a − λ1 is not invertible. NOTE: This deﬁnition, which is the “oﬃcial” one for this course, diﬀers from—and is not equivalent to—the one given on page 61 of AOLT. 17.1.2. Theorem (Spectral Mapping Theorem). If T is an operator on a ﬁnite dimensional vector space and p is a polynomial, then σ(p(T )) = p(σ(T )). That is, if σ(T ) = {λ1 , . . . , λk }, then σ(p(T )) = {p(λ1 ), . . . , p(λk )}.

55

56

17. SPECTRA

17.2. Exercises

(Due: Wed. Feb. 18)

17.2.1. Exercise. Let V be a vector space, T ∈ L(V ), v ∈ V , and φ ∈ V ∗ . Then T (v ⊗ φ) = (T v) ⊗ φ. 17.2.2. Exercise. Let V be a vector space, T ∈ L(V ), v ∈ V , and φ ∈ V ∗ . Then v ⊗ T ∗ φ = (v ⊗ φ)T. 17.2.3. Exercise. Prove that every ﬁnite rank linear map between vector spaces is a linear combination of rank 1 linear maps. 17.2.4. Exercise (†). Give an example to show that the result stated in AOLT, Theorem 2.24, part 1 need not hold in an inﬁnite dimensional unital algebra. Hint. Let l1 (N, R) be the family of all absolutely summable sequences of real numbers, that is, the set of all sequences (an ) of real numbers such that ∞ |an | < ∞. Consider the operator T in the algebra L(l1 (N, R)) which takes k=1 1 the sequence (an ) to the sequence n an . 17.2.5. Exercise. Give an example to show that the result stated in AOLT, Theorem 2.24, part 2 need not hold in an inﬁnite dimensional unital algebra. Hint. Consider the unilateral shift operator U in the algebra L(l(N, R)) which takes the sequence (an ) to the sequence (0, a1 , a2 , a3 , . . . ). 17.2.6. Exercise. Give an example to show that the result stated in AOLT, Theorem 2.24, part 3 need not hold in an inﬁnite dimensional unital algebra. 17.2.7. Exercise. If z is an element of the algebra C of complex numbers, then σ(z) = {z}. 17.2.8. Exercise. Let f be an element of the algebra C([a, b]) of continuous complex valued functions on the interval [a, b]. Find the spectrum of f . 17.2.9. Exercise. The family M3 (C) of 3 × 3 matrices of complex numbers is a  5 −6 under the usual matrix operations. Show that the spectrum of the matrix −1 4 3 −6 unital algebra  −6 2  is {1, 2}. −4

17.2.10. Exercise. Let a be an element of a unital algebra such that a2 = 1. Then either (i) a = 1, in which case σ(a) = {1}, or (ii) a = −1, in which case σ(a) = {−1}, or (iii) σ(a) = {−1, 1}. 1 Hint. In (iii) to prove σ(a) ⊆ {−1, 1}, consider (a + λ1). 1 − λ2 17.2.11. Exercise. An element a of an algebra is idempotent if a2 = a. Let a be an idempotent element of a unital algebra. Then either (i) a = 1, in which case σ(a) = {1}, or (ii) a = 0, in which case σ(a) = {0}, or (iii) σ(a) = {0, 1}. 1 Hint. In (iii) to prove σ(a) ⊆ {0, 1}, consider a + (λ − 1)1 . λ − λ2

17.2. EXERCISES

(DUE: WED. FEB. 18)

57

 1 −1 0 0 0 17.2.12. Exercise. Let T be the operator on R3 whose matrix representation is  0 −2 2 2 and let p(x) = x3 − x2 + x − 3. Verify the spectral mapping theorem 17.1.2 in this special case by computing separately σ(p(T )) and p(σ(T )). 17.2.13. Exercise. Prove the spectral mapping theorem 17.1.2.

CHAPTER 18

Polynomials
18.1. Background Topics: formal power series, polynomials, convolution, indeterminant, degree of a polynomial, polynomial functional calculus, annihilating polynomial, monic polynomial, minimal polynomial.

18.1.1. Notation. We make the convention that the set of natural numbers N does not include zero but the set Z+ of positive integers does. Thus Z+ = N ∪ {0}. 18.1.2. Notation. If S is a set and A is an algebra, l(S, A) denotes the vector space of all functions from S into A with pointwise operations of addition and scalar multiplication, and lc (S, A) denotes the subspace of functions with ﬁnite support. 18.1.3. Definition. Let A be a unital commutative algebra. On the vector space l(Z+ , A)
n

deﬁne a binary operation ∗ (often called convolution) by (f ∗ g)n =
j+k=n

fj g k =
j=0

fj g n−j

(where f , g ∈ l(Z+ , A) and n ∈ Z+ . An element of l(Z+ , A) is a formal power series (with coeﬃcients in A) and an element of lc (Z+ , A) is a polynomial (with coeﬃcients in A). 18.1.4. Remark. We regard the algebra A as a subset of l(Z+ , A) by identifying the element a ∈ A with the element (a, 0, 0, 0, . . . ) ∈ l(Z+ , A). Thus the map a → (a, 0, 0, 0, . . . ) becomes an inclusion map. (Technically speaking, of course, the map ψ : a → (a, 0, 0, 0, . . . ) is an injective unital homomorphism and A ∼ ran ψ.) = 18.1.5. Convention. In the algebra l(Z+ , A) we will henceforth write ab for a ∗ b. 18.1.6. Definition. Let A be a unital commutative algebra. In the algebra l(Z+ , A) of formal power series the special sequence x = (0, 1A , 0, 0, 0, . . . ) is called the indeterminant of l(Z+ , A). Notice that the sequence xn = x x x · · · x (n factors) has the property that (xn )n = 1 and (xn )k = 0 whenever k = n. It is conventional to take x0 to be the identity (1A , 0, 0, 0, . . . ) of l(Z+ , A). 18.1.7. Remark. The algebra l(Z+ , A) of formal power series with coeﬃcients in a unital commutative algebra A is frequently denoted by A [x] and the subalgebra lc (Z+ , A) of polynomials is denoted by A[x]. For many algebraists scalar multiplication is of little interest so A is taken to be a unital commutative ring, so that A [x] is ring of formal power series (with coeﬃcients in A) and A[x] is the polynomial ring (with coeﬃcients in A). In your text, AOLT, A is always a ﬁeld F. Since a ﬁeld can be regarded as a one-dimensional vector space over itself, it is also an algebra. Thus in the text F[x] is the polynomial algebra with coeﬃcients in F and has as its basis {xn : n = 0, 1, 2, . . . }. 18.1.8. Definition. A nonzero polynomial p, being an element of l0 (Z+ , A), has ﬁnite support. So there exists n0 ∈ Z+ such that pn = 0 whenever n > n0 . The smallest such n0 is the degree
59

60

18. POLYNOMIALS

of the polynomial. We denote it by deg p. A polynomial of degree 0 is a constant polynomial. The zero polynomial (the additive identity of l(Z+ , A)) is also a constant polynomial and many authors assign its degree to be −∞. If p is a polynomial of degree n, then pn is the leading coefficient of p. A polynomial is monic if its leading coeﬃcient is 1.
n k k=0 pk x

18.1.9. Definition. Let A be a unital algebra over a ﬁeld F. For each polynomial p = with coeﬃcients in F deﬁne
n

p: A → A: a →
k=0

pk ak .

Then p is the polynomial function on A determined by the polynomial p. Also for ﬁxed a ∈ A deﬁne Φ : F[x] → A : p → p(a). The mapping Φ is the polynomial functional calculus determined by the element a. 18.1.10. Definition. Let V be a vector space over a ﬁeld F and T ∈ L(V ). A nonzero polynomial p ∈ F[x] such that p(T ) = 0 is an annihilating polynomial for T . A monic polynomial of smallest degree that annihilates T is a minimal polynomial for T . 18.1.11. Notation. If T is an operator on a ﬁnite dimensional vector space over a ﬁeld F, we denote its minimal polynomial in F[x] by mT .

18.2. EXERCISES

(DUE: FRI. FEB. 20)

61

18.2. Exercises

(Due: Fri. Feb. 20)

18.2.1. Exercise. If A is a unital commutative algebra, then under the operations deﬁned in 18.1.3 l(Z+ , A) is a unital commutative algebra (whose multiplicative identity is the sequence (1A , 0, 0, 0, . . . )) and lc (Z+ , A) is a unital subalgebra of l(Z+ , A). 18.2.2. Exercise. If φ : A → B is a unital algebra homomorphism between unital commutative algebras, then the map l(Z+ , φ) : l(Z+ , A) → l(Z+ , B) : f → φ(fn )
∞ n=0

is also a unital homomorphism of unital commutative algebras. The pair of maps A → l(Z+ , A) and φ → l(Z+ , φ) is a covariant functor from the category of unital commutative algebras and unital algebra homomorphisms to itself. 18.2.3. Exercise. Let A be a unital commutative algebra. If p is a nonzero polynomial in lc (Z+ , A), then
n

p=
k=0

pk xk

where n = deg p.

18.2.4. Exercise. If p and q are polynomials with coeﬃcients in a unital commutative algebra A, then (i) deg(p + q) ≤ max{deg p, deg q}, and (ii) deg(pq) ≤ deg p + deg q. If A is a ﬁeld, the equality holds in (ii). 18.2.5. Exercise. Show that if A is a unital commutative algebra, then so is l(A, A) under pointwise operations of addition, multiplication, and scalar multiplication. 18.2.6. Exercise (†). Prove that for each a ∈ A the polynomial functional calculus Φ : F[x] → A deﬁned in 18.1.9 is a unital algebra homomorphism. Show also that the map Ψ : F[x] → l(A, A) : p → p is a unital algebra homomorphism. (Pay especially close attention to the fact that “multiplication” on F[x] is convolution whereas “multiplication” on l(A, A) is deﬁned pointwise.) What is the image under Φ of the indeterminant x? What is the image under Ψ of the indeterminant x? 18.2.7. Exercise. AOLT, section 2.7, exercise 2. 18.2.8. Exercise (†). Give an example to show that the polynomial functional calculus Φ may fail to be injective. Hint. The preceding exercise 18.2.7. 18.2.9. Exercise. Let A be a unital algebra over a ﬁeld F with ﬁnite dimension m. Show that for every a ∈ A there exists a polynomial p ∈ F[x] such that 1 ≤ deg p ≤ m and p(a) = 0. 18.2.10. Exercise. Let V be a ﬁnite dimensional vector space. Prove that every T ∈ L(V ) has a minimal polynomial. Hint. Use exercise 18.2.9.. 18.2.11. Exercise. Let f and d be polynomials with coeﬃcients in a ﬁeld F and suppose that d = 0. Then there exist unique polynomials q and r in F[x] such that (i) f = dq + r and

62

18. POLYNOMIALS

(ii) r = 0 or deg r < deg d. Hint. Let f = k fj xj and d = m dj xj be in standard form. The case k < m is trivial. For j=0 j=0 k ≥ m suppose the result to be true for all polynomials of degree strictly less than k. What can you say about f = f − p where p = (fk d−1 )xk−m d? m 18.2.12. Exercise. Let V be a ﬁnite dimensional vector space and T ∈ L(V ). Prove that the minimal polynomial for T is unique.

CHAPTER 19

Polynomials (continued)
19.1. Background Topics: irreducible polynomial, prime polynomial, Langrange interpolation formula, unique factorization theorem, greatest common divisor, relatively prime.

19.1.1. Definition. A polynomial p ∈ F[x] is irreducible (over F) provided that it is not constant and whenever p = f g with f , g ∈ F[x], then either f or g is constant. 19.1.2. Definition. Let t0 , t1 , . . . , tn be distinct elements of a ﬁeld F. For 0 ≤ k ≤ n deﬁne pk ∈ F[x] by n x − tj pk = . tk − t j
j=0 j=k

19.1.3. Definition. If F is a ﬁeld, a polynomial p ∈ F[x] is reducible over F if there exist polynomials q and r in F[x] both of degree at least one such that p = qr. A polynomial which is not reducible over F is irreducible over F. A nonscalar irreducible polynomial is a prime polynomial over F (or is a prime in F[x]).

63

64

19. POLYNOMIALS (CONTINUED)

19.2. Exercises

(Due: Mon. Feb. 23)

19.2.1. Exercise. Let T be an operator on a ﬁnite dimensional vector space. Show that T is invertible if and only if the constant term of its minimal polynomial is not zero. Explain, for an invertible operator T , how to write its inverse as a polynomial in T . 19.2.2. Exercise (†). Let T be an operator on a ﬁnite dimensional vector space over a ﬁeld F. If p ∈ F[x] and p(T ) = 0, then mT divides p. (If p, p1 ∈ F[x], we say that p1 divides p if there exists q ∈ F[x] such that p = p1 q.) 19.2.3. Exercise. Let T be the operator on the real vector space R2 whose matrix representa0 −1 tion (with respect to the standard basis) is . Find the minimal polynomial mT of T and 1 0 show that it is irreducible (over R). 19.2.4. Exercise. Let T be the operator on the complex vector space C2 whose matrix repre0 −1 sentation (with respect to the standard basis) is . Find the minimal polynomial mT of T 1 0 and show that it is not irreducible (over C). 19.2.5. Exercise. Let p be a polynomial of degree m ≥ 1 in F[x]. If Jp is the principal ideal generated by p, that is if Jp := {pf : f ∈ F[x]} then dim F[x]/Jp = m. Hint. Show that B = { [xk ] : k = 0, 1, . . . , m − 1} is a basis for the vector space F[x]/Jp . 19.2.6. Exercise. Let T be an operator on a ﬁnite dimensional vector space V over a ﬁeld F and Φ : F[x] → L(V ) be the associated polynomial functional calculus. If p is a polynomial of degree m ≥ 1 in F[x] and Jp is the principal ideal generated by p, then the sequence 0 is exact. 19.2.7. Exercise (Lagrange Interpolation Formula). Prove that the polynomials deﬁned in 19.1.2 form a basis for the vector space V of all polynomials with coeﬃcients in F and degree less than or equal to n and that for each polynomial q ∈ V
n

/ Jp

/ F[x]

Φ

/ ran Φ

/0

q=
k=0

q(tk )pk .

19.2.8. Exercise. Use the Lagrange Interpolation Formula to ﬁnd the polynomial with coeﬃcients in R and degree no greater than 3 whose values at −1, 0, 1, and 2 are, respectively, −6, 2, −2, and 6. 19.2.9. Exercise. Let F be a ﬁeld and p, q, and r be polynomials in F[x]. If p is a prime in F[x] and p divides qr, then p divides q or p divides r. 19.2.10. Exercise. Prove the Unique Factorization Theorem: Let F be a ﬁeld. A nonscalar monic polynomial in F[x] can be factored in exactly one way (except for the order of the factors) as a product of monic primes in F[x] .

19.2. EXERCISES

(DUE: MON. FEB. 23)

65

19.2.11. Exercise. Let F be a ﬁeld. Then every nonzero ideal in F[x] is principal. 19.2.12. Exercise. If p1 , . . . , pn are polynomials, not all zero, with coeﬃcients in a ﬁeld F, then there exists a unique monic polynomial d in the ideal generated by p1 , . . . , pn such that d divides pk for each k = 1, . . . , n and, furthermore, any polynomial which divides each pk also divides d. This polynomial d is the greatest common divisor of the pk ’s. The polynomials pk are relatively prime if their greatest common denominator is 1.

CHAPTER 20

Invariant Subspaces
20.1. Background Topics: invariant subspaces, invariant subspace lattice, reducing subspaces, transitive algebra, Burnside’s theorem, triangulable. (See AOLT sections 3.1 and 3.3.)

20.1.1. Definition. An operator T on a vector space V is reduced by a pair (M, N ) of subspaces M and N of V if (i) V = M ⊕ N , (ii) M is invariant under T , and (iii) N is invariant under T .

67

68

20. INVARIANT SUBSPACES

20.2. Exercises

(Due: Wed. Feb. 25)

  3 4 2 20.2.1. Exercise. Let S be the operator on R3 whose matrix representation is 0 1 2. 0 0 0 Find three one dimensional subspaces U , V , and W of R3 which are invariant under S.   0 0 2 20.2.2. Exercise. Let T be the operator on R3 whose matrix representation is 0 2 0. 2 0 0 Find a two dimensional subspace U of R3 which is invariant under T . 20.2.3. Exercise. Find inﬁnitely many subspaces of the vector space of polynomial functions on R which are invariant under the diﬀerentiation operator.  2 0 0 20.2.4. Exercise. Let T be the operator on R3 whose matrix representation is −1 3 2. 1 −1 0 Find a plane and a line in R3 which reduce T . 20.2.5. Exercise. AOLT, section 3.7, exercise 1. 20.2.6. Exercise. AOLT, section 3.7, exercise 2. 20.2.7. Exercise (†). Let M be a subspace of a vector space V and T ∈ L(V ). Show that if M is invariant under T , then ET E = T E for every projection E onto M . Show also that if ET E = T E for some projection E onto M , then M is invariant under T . 20.2.8. Exercise. Suppose a vector space V has the direct sum decomposition V = M ⊕ N . An operator T on V is reduced by the pair (M, N ) if and only if ET = T E, where E = EM N is the projection along M onto N . 20.2.9. Exercise. Let M and N be complementary subspaces of a vector space V (that is, V is the direct sum of M and N ) and let T be an operator on V . Show that if M is invariant under T , then M ⊥ is invariant under T ∗ and that if T is reduced by the pair (M, N ), then T ∗ is reduced by the pair (M ⊥ , N ⊥ ). 20.2.10. Exercise. Let T : V → W be linear and S : W → V a left inverse for T . Then (a) W = ran T ⊕ ker S, and (b) T S is the projection along ker S onto ran T . 20.2.11. Exercise. If V is a vector space, V = M ⊕ N = M ⊕ N , and M ⊆ M , then M = M . 

CHAPTER 21

The Spectral Theorem for Vector Spaces
21.1. Background Topics: similarity of operators, diagonalizable, resolution of the identity, spectral theorem for vector spaces. (See section 1.1 of my notes on operator algebras [12].)

21.1.1. Definition. Suppose that on a vector space V there exist projection operators E1 , . . . , En such that (i) IV = E1 + E2 + · · · + En and (ii) Ei Ej = 0 whenever i = j. Then we say that the family {E1 , E2 , . . . , En } of projections is a resolution of the identity. 21.1.2. Definition. Two operators on a vector space (or two n × n matrices) R and T are similar if there exists an invertible operator (or matrix) S such that R = S −1 T S. 21.1.3. Notation. Let α1 , . . . , αn be elements of a ﬁeld F. Then diag(α1 , . . . , αn ) denotes the n × n matrix whose entries are all zero except on the main diagonal where they are α1 , . . . , αn . k That is, if [d1 , . . . , d n ] = diag(α1 , . . . , αn ), then dk = αk for each k, and djk = 0 whenever j = k. Such a matrix is a diagonal matrix. 21.1.4. Definition. Let V be a vector space of ﬁnite dimension n. An operator T on V is diagonalizable if it has n linearly independent eigenvectors (or, equivalently, if it has a basis of eigenvectors). 21.1.5. Theorem (Spectral Theorem for Vector Spaces). If T is a diagonalizable operator on a ﬁnite dimensional vector space V , then
n

T =
k=1

λ k Ek

where λ1 , . . . , λn are the (distinct) eigenvalues of T and {E1 , . . . En } is the resolution of the identity whose projections are associated with the corresponding eigenspaces M1 , . . . , Mn .

69

70

21. THE SPECTRAL THEOREM FOR VECTOR SPACES

21.2. Exercises

(Due: Fri. Feb. 27)

21.2.1. Exercise. If {E1 , E2 , . . . , En } is a resolution of the identity on a vector space V , then V = n ran Ek . k=1 21.2.2. Exercise. If M1 ⊕ · · · ⊕ Mn is a direct sum decomposition of a vector space V , then the family {E1 , E2 , . . . , En } of the associated projections is a resolution of the identity. 21.2.3. Exercise. If v is a nonzero eigenvector associated with an eigenvalue λ of an operator T ∈ L(V ) and p is a polynomial in F[x], then p(T )v = p(λ)v. 21.2.4. Exercise. Let λ1 , . . . , λn be the distinct eigenvalues of an operator T on a ﬁnite dimensional vector space V and let M1 , . . . Mn be the corresponding eigenspaces. If U = n Mk , k=1 then dim U = n dim Mk . Furthermore, if Bk is a basis for Mk (k = 1, . . . n), then n Bk is a k=1 k=1 basis for U . 21.2.5. Exercise. If R and T are operators on a vector space, is RT always similar to T R? What if R is invertible? 21.2.6. Exercise. Let R and T be operators on a vector space. If R is similar to T and p ∈ F[x] is a polynomial, then p(R) is similar to p(T ). 21.2.7. Exercise. If R and T are operators on a vector space, R is similar to T , and R is invertible, then T is invertible and T −1 is similar to R−1 . 21.2.8. Exercise. Let A be an n × n matrix with entries from a ﬁeld F. Then A, regarded as an operator on Fn , is diagonalizable if and only if it is similar to a diagonal matrix. 21.2.9. Exercise (†). Let E1 , . . . , En be the projections associated with a direct sum decomposition V = M1 ⊕· · ·⊕Mn of a vector space V and let T be an operator on V . Then each subspace Mk is invariant under T if and only if T commutes with each projection Ek . 21.2.10. Exercise. Prove the spectral theorem for vector spaces 21.1.5. 21.2.11. Exercise. Let T be an operator on a ﬁnite dimensional vector space V . If λ1 , . . . , λn are distinct scalars and E1 , . . . , En are nonzero operators on V such that (i) T = n λk Ek , k=1 (ii) I = n Ek , and k=1 (iii) Ej Ek = 0 whenever j = k, then T is diagonalizable, the scalars λ1 , . . . , λn are the eigenvalues of T , and the operators E1 , . . . , En are projections whose ranges are the eigenspaces of T . 21.2.12. Exercise. If T is an operator on a ﬁnite dimensional vector space V over a ﬁeld F and p ∈ F[x], then
n

p(T ) =
k=1

p(λk )Ek

where λ1 , . . . , λn are the (distinct) eigenvalues of T and E1 , . . . En are the projections associated with the corresponding eigenspaces M1 , . . . , Mn .

CHAPTER 22

The Spectral Theorem for Vector Spaces (continued)
22.1. Background Topics: representation, faithful representation, irreducible representation, Cayley-Hamilton theorem. (See AOLT, section 3.4.)

22.1.1. Definition. If A is an n × n matrix, deﬁne the characteristic polynomial cA of A to be the determinant of xI − A. (Note that xI − A is a matrix with polynomial entries.) As you would expect, the characteristic polynomial of an operator on a ﬁnite dimensional space (with basis B) is the characteristic polynomial of the matrix representation of that operator (with respect to B). 22.1.2. Theorem (Cayley-Hamilton Theorem). If T is an operator on a ﬁnite dimensional vector space, then the characteristic polynomial of T annihilates T .

71

72

22. THE SPECTRAL THEOREM FOR VECTOR SPACES (CONTINUED)

22.2. Exercises

(Due: Mon. Mar. 2)

22.2.1. Exercise (†). AOLT, section 3.7, exercise 12. 22.2.2. Exercise. If T is a diagonalizable operator on a ﬁnite dimensional vector space V , then the projections E1 , . . . , En associated with the decomposition of V as a direct sum Mk of its eigenspaces can be expressed as polynomials in T . Hint. Apply the Lagrange interpolation formula with the tk ’s being the eigenvalues of T .   0 0 2 22.2.3. Exercise. Let T be the operator on R3 whose matrix representation is 0 2 0. Use 2 0 0 exercise 22.2.2 to write T as a linear combination of projections.  2 −2 1 22.2.4. Exercise. Let T be the operator on R3 whose matrix representation is −1 1 1. −1 2 0 Use exercise 22.2.2 to write T as a linear combination of projections. 22.2.5. Exercise. Let T be the operator on R3 whose matrix representation is given in exercise 22.2.2. Write T as a linear combination of projections. 22.2.6. Exercise. Let T be an operator on a ﬁnite dimensional vector space over a ﬁeld F. Show that λ is an eigenvalue of T if and only if cT (λ) = 0. 22.2.7. Exercise (††). Let T be an operator on a ﬁnite dimensional vector space. Show that T is diagonalizable if and only if its minimal polynomial is of the form n (x − λk ) for some distinct k=1 elements λ1 , . . . , λn of the scalar ﬁeld F. 22.2.8. Exercise. If T is an operator on a ﬁnite dimensional vector space, then its minimal polynomial and characteristic polynomial have the same roots. 22.2.9. Exercise. Prove the Caley-Hamilton theorem 22.1.2. Hint. A proof is in the text (see AOLT, pages 92–94). The problem here is to ﬁll in missing details and provide in a coherent fashion any background material necessary to understanding the proof. 

CHAPTER 23

Diagonalizable Plus Nilpotent Decomposition
23.1. Background Topics: primary decomposition theorem, diagonalizable plus nilpotent decomposition, Jordan normal form.

23.1.1. Theorem (Primary Decomposition Theorem). Let T ∈ L(V ) where V is a ﬁnite dimensional vector space. Factor the minimal polynomial
n

mT =
k=1

r pk k

into powers of distinct irreducible monic polynomials p1 , . . . , pn and let Wk = ker pk (T rk ) for each k. Then (i) V = n Wk , k=1 (ii) each Wk is invariant under T , and r (iii) if Tk = T W , then mT = pk k .
k k

In the preceding theorem the spaces Wk are the generalized eigenspaces of the operator T . 23.1.2. Theorem. Let T be an operator on a ﬁnite dimensional vector space V . Suppose that the minimal polynomial for T factors completely into linear factors mT (x) = (x − λ1 )d1 . . . (x − λr )dr where λ1 , . . . λr are the (distinct) eigenvalues of T . For each k let Wk be the generalized eigenspace ker(T − λk I)dk and let E1 , . . . , Er be the projections associated with the direct sum decomposition V = W1 ⊕ W2 ⊕ · · · ⊕ Wr . Then this family of projections is a resolution of the identity, each Wk is invariant under T , the operator D = λ 1 E1 + · · · + λ r Er is diagonalizable, the operator N =T −D is nilpotent, and N commutes with D. Furthermore, if D1 is diagonalizable, N1 is nilpotent, D1 + N1 = T , and D1 N1 = N1 D1 , then D1 = D and N1 = N . 23.1.3. Corollary. Every operator on a ﬁnite dimensional complex vector space can be written as the sum of two commuting operators, one diagonalizable and the other nilpotent.

73

74

23. DIAGONALIZABLE PLUS NILPOTENT DECOMPOSITION

23.2. Exercises

(Due: Wed. Mar. 4)

23.2.1. Exercise. Prove the primary decomposition theorem 23.1.1. 23.2.2. Exercise. Prove theorem 23.1.2. 23.2.3. Exercise. Let T be the operator on R2 whose matrix representation is (a) Explain brieﬂy why T is not diagonalizable. (b) Find the diagonalizable and nilpotent parts of T . a b −c c Answer: D = and N = where a = b a −c c 2 1 . −1 4

,b=

, and c = 

.

 0 0 −3 23.2.4. Exercise (†). Let T be the operator on R3 whose matrix representation is −2 1 −2. 2 −1 5 (a) Find D and N , the diagonalizable and nilpotent parts of T . Express these as polynomials in T . (b) Find a matrix S which diagonalizes D.     2 −1 −1 −2 1 −2 (c) Let [D1 ] = −1 2 −1 and [N1 ] = −1 −1 −1. Show that D1 is diagonalizable, −1 −1 2 3 0 3 that N1 is nilpotent, and that T = D1 + N1 . Why does this not contradict the uniqueness claim made in theorem 23.1.2?   0 1 0 −1 −2 3 0 −1  23.2.5. Exercise. Let T be the operator on R4 whose matrix representation is  −2 1 2 −1. 2 −1 0 3 . (a) The characteristic polynomial of T is (λ − 2)p where p = r where r = (b) The minimal polynomial of T is (λ − 2) .   a b b b b a b b  (c) The diagonalizable part of T is D =  and b = .  b b a b  where a = b b b a   −a b c −b −a b c −b  (d) The nilpotent part of T is N =  ,b= , and −a b c −b where a = c= . a −b c b   1 0 0 1 −1 0 1 −2 3 −3   5 whose matrix representation is 0 0 −1 2 −2. 23.2.6. Exercise. Let T be the operator on R   1 −1 1 0 1 1 −1 1 −1 2 (a) Find the characteristic polynomial of T . Answer: cT (λ) = (λ + 1)p (λ − 1)q where p = (b) Find the minimal polynomial of T . Answer: mT (λ) = (λ + 1)r (λ − 1)s where r = and q = and s = . .

23.2. EXERCISES

(DUE: WED. MAR. 4)

75

(c) Find the eigenspaces V1 and V2 of T . Answer: V1 = span{(a, 1, b, a, a)} where a = V2 = span{(1, a, b, b, b), (b, b, b, 1, a)} where a = (d) Find the diagonalizable part of T .   a b b b b  b a −c c −c   Answer: D =  b b −a c −c where a =   b b b a b  b b b b a (e) Find the nilpotent part of  a a a a  Answer: N = a a   b −b b −b T.  a b −b a b −b  a a a  where a =  b −b b  b −b b

and b = and b =

; and .

,b=

, and c =

.

and b =

.

(f) Find a matrix S which diagonalizes the diagonalizable part D of T . What is the diagonal form Λ of D associated with this  matrix?  a b a a a  b a b a a   Answer: S =  b a a b a where a = and b = .   a a a b b  a a a a b   −a 0 0 0 0  0 a 0 0 0   and Λ =  0 0 a 0 0 where a = .    0 0 0 a 0 0 0 0 0 a 23.2.7. Exercise. Prepare and deliver a (30–45 minute) blackboard presentation on the Jordan normal form of a matrix.

CHAPTER 24

Inner Product Spaces
24.1. Background Topics: inner products, norm, unit vector, square summable sequences, the Schwarz (or CauchySchwarz) inequality. (See AOLT, section 1.3 and also section 1.2 of my lecture notes on operator algebras [12].)

24.1.1. Definition. If x is a vector in Rn , then the norm (or length) of x is deﬁned by x = A vector of length 1 is a unit vector 24.1.2. Theorem (Pythagorean theorem). If x and y are perpendicular vectors in an inner product space, then x + y 2 = x 2 + y 2. 24.1.3. Theorem (Parallelogram law). If x and y are vectors in an inner product space, then x+y
2

x, x .

+ x−y

2

=2 x

2

+ 2 y 2.

24.1.4. Theorem (Polarization identity). If x and y are vectors in a complex inner product space, then 1 x, y = 4 ( x + y 2 − x − y 2 + i x + iy 2 − i x − iy 2 ) . What is the correct formula for a real inner product space?

77

78

24. INNER PRODUCT SPACES

24.2. Exercises

(Due: Mon. Mar. 9)

24.2.1. Exercise. Prove the Pythagorean theorem 24.1.2. 24.2.2. Exercise. Prove the parallelogram law 24.1.3. 24.2.3. Exercise. On the vector space C([0, 1]) of continuous real valued functions on the interval [0, 1] the uniform norm is deﬁned by f u = sup{|f (x)| : 0 ≤ x ≤ 1}. Prove that the uniform norm is not induced by an inner product. That is, prove that there is no inner product on f, f for all f ∈ C([0, 1]). Hint. Use exercise 24.2.2 C([0, 1]) such that f u = 24.2.4. Exercise. Prove the polarization identity 24.1.4. 24.2.5. Exercise. If a1 , . . . , an > 0, then
n n

aj
j=1 k=1

1 ak

≥ n2 .

The proof of this is obvious from the Schwarz inequality if we choose x and y to be what? 24.2.6. Exercise (†). Notice that part (a) is a special case of part (b).
1 1 (a) Show that if a, b, c > 0, then 2 a + 1 b + 6 c 3 (b) Show that if a1 , . . . , an , w1 , . . . , wn > 0 and n 2 2

≤ 1 a2 + 1 b2 + 1 c2 . 2 3 6 n wk = 1, then k=1 ak 2 wk .

n

ak wk
k=1

k=1

24.2.7. Exercise. Show that if

∞ 2 k=1 ak

converges, then

∞ −1 k=1 k ak

converges absolutely.

24.2.8. Exercise. A sequence (ak ) of (real or) complex numbers is square summable if < ∞. The vector space of all square summable sequences of real numbers (respectively, complex numbers) is denoted by l2 (R) (respectively, l2 (C)). When no confusion will result, both are denoted by l2 . If a, b ∈ l2 , deﬁne
∞ 2 k=1 |ak | ∞

a, b =
k=1

ak bk .

Show that this deﬁnition makes sense and makes l2 into an inner product space. 24.2.9. Exercise (†). Use vector methods (no coordinates, no major results from Euclidean geometry) to show that the midpoint of the hypotenuse of a right triangle is equidistant from the vertices. Hint. Let ABC be a right triangle and O be the midpoint of the hypotenuse AB. What − → − − − − → → −→ − can you say about < AO + OC, CO + OB >? 24.2.10. Exercise. Use vector methods to show that if a parallelogram has perpendicular diagonals, then it is a rhombus (that is, all four sides have equal length). Hint. Let ABCD be − → −→ − a parallelogram. Express the dot (inner) product of the diagonals AC and DB in terms of the − − → −→ − lengths of the sides AB and BC. 24.2.11. Exercise. Use vector methods to show that an angle inscribed in a semicircle is a right angle.

24.2. EXERCISES

(DUE: MON. MAR. 9)

79

24.2.12. Exercise. Let a be a vector in an inner product space H. If x, a = 0 for every x ∈ H, then a = 0. 24.2.13. Exercise. Let S, T : H → K be linear maps between inner product spaces H and K. If Sx, y = T x, y for every x ∈ H and y ∈ K, then S = T .

CHAPTER 25

25.1. Background Topics: orthogonality, orthogonal complement, adjoint, involution, ∗ -algebra, ∗ -homomorphism, self-adjoint, Hermitian, normal, unitary.

25.1.1. Definition. An involution on an algebra A is a map x → x∗ from A into A which satisﬁes (i) (x + y)∗ = x∗ + y ∗ , (ii) (αx)∗ = α x∗ , (iii) x∗∗ = x, and (iv) (xy)∗ = y ∗ x∗ for all x, y ∈ A and α ∈ E. An algebra on which an involution has been deﬁned is a ∗ -algebra (pronounced “star algebra”). An algebra homomorphism φ between ∗ -algebras which preserves involution (that is, such that φ(a∗ ) = (φ(a))∗ ) is a ∗ -homomorphism (pronounced “star homomorphism”. A ∗ -homomorphism φ : A → B between unital algebras is said to be unital if φ(1A ) = 1B . 25.1.2. Definition. Let H and K be inner product spaces and T : H → K be a linear map. If there exists a function T ∗ : K → H which satisﬁes T x, y = x, T ∗ y for all x ∈ H and y ∈ K, then T ∗ is the adjoint of T . When H and K are real vector spaces, the adjoint of T is usually called the transpose of T and the notation T t is used (rather than T ∗ ). 25.1.3. Definition. An element a of a ∗ -algebra A is self-adjoint (or Hermitian) if a∗ = a. It is normal if a∗ a = aa∗ . And it is unitary if a∗ a = aa∗ = 1. The set of all self-adjoint elements of A is denoted by H(A), the set of all normal elements by N (A), and the set of all unitary elements by U(A).

81

82

25.2. Exercises

(Due: Wed. Mar. 11)

25.2.1. Exercise. Let H be an inner product space and a ∈ H. Deﬁne ψa : V → E by ψa (x) = x, a for all x ∈ H. Then ψa is a linear functional on H. (See the last sentence of the proof of Theorem 1.15, AOLT, page 15.) 25.2.2. Exercise (†). Give a proof of the Riesz representation theorem that is much simpler than the one given in AOLT, page 15, Theorem 1.15. Hint. Use exercises 9.2.9 and 9.2.13. 25.2.3. Exercise. Show by example that the Riesz representation theorem (AOLT, page 15, Theorem 1.15) does not hold (as stated) in inﬁnite dimensional spaces. Hint. Consider the function φ : lc (N) → E : x → ∞ αk where x = ∞ αk ek , the ek ’s being the standard basis vectors for k=1 k=1 l0 (N). 25.2.4. Exercise. Show that if S is a set of mutually perpendicular vectors in an inner product space and 0 ∈ S, then the set S is linearly independent. / 25.2.5. Exercise. Let S and T be subsets of an inner product space H. (a) S ⊥ is a subspace of H. (b) If S ⊆ T , then T ⊥ S ⊥ . (c) (span S)⊥ = S ⊥ . 25.2.6. Exercise. Show that if M is a subspace of a ﬁnite dimensional inner product space H, then H = M ⊕ M ⊥ . Show also that this need not be true in an inﬁnite dimensional space. 25.2.7. Exercise. Let M be a subspace of an inner product space H. (a) Show that M ⊆ M ⊥⊥ . (b) Prove that equality need not hold in (a). (c) Show that if H is ﬁnite dimensional, then M = M ⊥⊥ . 25.2.8. Exercise. Let M and N be subspaces of an inner product space H. (a) Show that (M + N )⊥ = (M ∪ N )⊥ = M ⊥ ∩ N ⊥ . (b) Show that if H is ﬁnite dimensional, then (M ∩ N )⊥ = M ⊥ + N ⊥ . 25.2.9. Exercise. It seems unlikely that the similarity between the results of the exercises 25.2.5 and 25.2.8 and those you obtained in exercises 10.2.5 and 10.2.6 could be purely coincidental. Explain carefully what is going on here. 25.2.10. Exercise. Show that if U is the unilateral shift operator on l2 U : l2 → l2 : (x1 , x2 , x2 , . . . ) → (0, x1 , x2 , . . . ), then its adjoint is given by U ∗ : l2 → l2 : (x1 , x2 , x3 , . . . ) → (x2 , x3 , x4 , . . . ). 25.2.11. Exercise (Multiplication operators). Let (X, A, µ) be a sigma-ﬁnite measure space and L2 (X) be the Hilbert space of all (equivalence classes of) complex valued functions on X which are square integrable with respect to µ. Let φ be an essentially bounded complex valued µ-measurable function on X. Deﬁne Mφ on L2 (X) by Mφ (f ) := φf. Then Mφ is an operator on L2 (X), Mφ = φ ∞ , and Mφ ∗ = Mφ .

25.2. EXERCISES

(DUE: WED. MAR. 11)

83

25.2.12. Exercise. Let a and b be elements of a ∗ -algebra. Show that a commutes with b if and only if a∗ commutes with b∗ . 25.2.13. Exercise. Show that in a unital ∗ -algebra 1∗ = 1. 25.2.14. Exercise. Let a be an element of a unital ∗ -algebra. Show that a∗ is invertible if and only if a is. And when a is invertible we have (a∗ )−1 = a−1 . 25.2.15. Exercise. Let a be an element of a unital ∗ -algebra. Show that λ ∈ σ(a) if and only if λ ∈ σ(a∗ ).

CHAPTER 26

Orthogonal Projections
26.1. Background Topics: orthogonal projections on an inner product space, projections in an algebra with involution, orthogonality of projections in a ∗ -algebra.

26.1.1. Definition. An operator P on an inner product space H is an orthogonal projection if it is self-adjoint and idempotent. (On a real inner product space, of course, the conditions are symmetric and idempotent.) 26.1.2. Definition. A projection in a ∗ -algebra A is an element p of the algebra which is idempotent (p2 = p) and self-adjoint (p∗ = p). The set of all projections in A is denoted by P(A). Notice that projections in *-algebras correspond to orthogonal projections on inner product spaces and not to (the more general) projections on vector spaces. 26.1.3. Definition. Two projections p and q in a ∗ -algebra are orthogonal, written p ⊥ q, if pq = 0.

85

86

26. ORTHOGONAL PROJECTIONS

26.2. Exercises

(Due: Fri. Mar. 13)

26.2.1. Exercise. Let T : H → K be a linear map between inner product spaces. Show that if the adjoint of T exists, then it is unique. (That is, there is at most one function T ∗ : K → H that satisﬁes T x, y = x, T ∗ y for all x ∈ H and y ∈ K.) 26.2.2. Exercise (†). Let T : H → K be a linear map between inner product spaces. Show that if the adjoint of T exists, then it is linear. S∗ 26.2.3. Exercise. Let S and T be operators on an inner product space H. Then (S + T )∗ = + T ∗ and (αT )∗ = αT ∗ for every α ∈ E. (See AOLT, Theorem 1.17.)

26.2.4. Exercise. Let T : H → K be a linear map between inner product spaces. Show that if the adjoint of T exists, then so does the adjoint of T ∗ and T ∗∗ = T . 26.2.5. Exercise. Let S : H → K and T : K → L be linear maps between complex inner product space. Show that if S and T both have adjoints, then so does their composite T S and (T S)∗ = S ∗ T ∗ . (See AOLT, Theorem 1.17 and section 1.8, exercise 12.) 26.2.6. Exercise. Let H be an inner product space and M and N be subspaces of H such that H = M + N and M ∩ N = ∅. (That is, we suppose that H is the vector space direct sum of M and N .) Also let P = EN M be the projection of H along N onto M . Prove that P is self-adjoint (P ∗ exists and P ∗ = P ) if and only if M ⊥ N . 26.2.7. Exercise. If P is an orthogonal projection on an inner product space H, then the space is the orthogonal direct sum of the range of P and its kernel; that is H = ran P + ker P and ran P ⊥ ker P . 26.2.8. Exercise. Let p and q be projections in a ∗ -algebra. Then pq is a projection if and only if pq = qp. 26.2.9. Exercise. Let P and Q be orthogonal projections on an inner product space V . If P Q = QP , then P Q is an orthogonal projection whose kernel is ker P + ker Q and whose range is ran P ∩ ran Q. 26.2.10. Exercise (†). Let T : H → K be a linear map between inner product spaces. Show that ker T ∗ = (ran T )⊥ . Is there a relationship between T being surjective and T ∗ being injective? 26.2.11. Exercise. Let T : H → K be a linear map between inner product spaces. Show that ker T = (ran T ∗ )⊥ . Is there a relationship between T being injective and T ∗ being surjective? 26.2.12. Exercise. It seems unlikely that the similarity between the results of the two preceding exercises and those you obtained in exercises 11.2.3 and 11.2.4 could be purely coincidental. Explain carefully what is going on here.

26.2. EXERCISES

(DUE: FRI. MAR. 13)

87

∗ -algebra

26.2.13. Exercise. A necessary and suﬃcient condition for two projections p and q in a to be orthogonal is that pq + qp = 0.

26.2.14. Exercise. Let p and q be projections in a ∗ -algebra. Then p + q is a projection if and only if p and q are orthogonal. 26.2.15. Exercise. Let P and Q be orthogonal projections on an inner product space V . Then P ⊥ Q if and only if ran P ⊥ ran Q. In this case P + Q is an orthogonal projection whose kernel is ker P ∩ ker Q and whose range is ran P + ran Q.

CHAPTER 27

The Spectral Theorem for Inner Product Spaces
27.1. Background Topics: subprojections, partial isometries, initial and ﬁnal projections, orthogonal resolutions of the identity, the spectral theorem for complex inner product spaces.

27.1.1. Definition. If p and q are projections in a ∗ -algebra we write p ≤ q if p = pq. In this case we say that p is a subprojection of q or that p is smaller than q. (Note: it is easy to see that the condition p = pq is equivalent to p = qp.) 27.1.2. Definition. An element v of a ∗ -algebra is a partial isometry if vv ∗ v = v. If v is a partial isometry, then v ∗ v is the initial projection of v and vv ∗ is its final projection. 27.1.3. Definition. Let M1 ⊕ · · · ⊕ Mn be an orthogonal direct sum decomposition of an inner product space V . For each k let Pk be the orthogonal projection onto Mk . The projections P1 , . . . Pn are the orthogonal projections associated with the orthogonal direct sum decomposition V = M1 ⊕ · · · ⊕ Mn . The family {P1 , . . . , Pn } is an orthogonal resolution of the identity. (Compare this with deﬁnitions 15.1.3 and 21.1.1.) 27.1.4. Definition. Two elements a and b of a ∗ -algebra A are unitarily equivalent if there exists a unitary element u of A such that b = u∗ au. 27.1.5. Definition. An operator T on a complex inner product space V is unitarily diagonalizable if there exists an orthonormal basis for V consisting of eigenvectors of T . 27.1.6. Theorem (Spectral Theorem: Complex Inner Product Space Version). If N is a normal operator on a ﬁnite dimensional complex inner product space V , then N is unitarily diagonalizable and can be written as
n

N=
k=1

λ k Pk

where λ1 , . . . , λn are the (distinct) eigenvalues of N and {P1 , . . . , Pn } is the orthogonal resolution of the identity whose orthogonal projections are associated with the corresponding eigenspaces M1 , . . . , Mn .

89

90

27. THE SPECTRAL THEOREM FOR INNER PRODUCT SPACES

27.2. Exercises

(Due: Mon. Mar. 30)

27.2.1. Exercise. If A is a ∗ -algebra, then the relation ≤ deﬁned in 27.1.1 is a partial ordering on P(A). If A is unital, then 0 ≤ p ≤ 1 for every p ∈ P(A). 27.2.2. Exercise. Let p and q be projections in a ∗ -algebra. Then q − p is a projection if and only if p ≤ q. 27.2.3. Exercise. Let P and Q be orthogonal projections on an inner product space V . Then the following are equivalent: (i) P ≤ Q; (ii) P x ≤ Qx for all x ∈ V ; and (iii) ran P ⊆ ran Q. In this case Q − P is an orthogonal projection whose kernel is ran P + ker Q and whose range is ran Q ran P . Notation: If M and N are subspaces of an inner product space with N M , then M N denotes the orthogonal complement of N in M (so N ⊥ (M N ) and M = N ⊕ (M N ) ). 27.2.4. Exercise. If p and q are commuting projections in a ∗ -algebra, then p ∧ q and p ∨ q exist. In fact, p ∧ q = pq and p ∨ q = p + q − pq. 27.2.5. Exercise. Let v be a partial isometry in a ∗ -algebra, p be its initial projection, and q be its ﬁnal projection. Then (a) v ∗ is a partial isometry, (b) p is a projection, (c) p is the smallest projection such that vp = v, (d) q is a projection, and (e) q is the smallest projection such that qv = v. 27.2.6. Exercise. Let N be a normal operator on a ﬁnite dimensional complex inner product space H. Show that N x = N ∗ x for all x ∈ H. (See AOLT, page 21, Proposition 1.22.) 27.2.7. Exercise. Let N be the operator on C2 whose matrix representation is 0 1 . −1 0 (a) The eigenspace M1 associated with the eigenvalue −i is the span of ( 1 , ). (b) The eigenspace M2 associated with the eigenvalue i is the span of ( 1 , ). (c) The (matrix representations of the) orthogonal projections P1 and P2 onto the eigenspaces a b a −b M1 and M2 , respectively, are P1 = ; and P2 = where a = and −b a b a b= . (d) Write N as a linear combination of the projections found in (c). Answer: [N ] = P1 + P2 . a a (e) A unitary matrix U which diagonalizes [N ] is a and b = .  where  = −b b The associated diagonal form Λ = U ∗ [N ] U of [N ] is  .

27.2. EXERCISES

(DUE: MON. MAR. 30)

91

27.2.8. Exercise. Let H be the self-adjoint matrix

2 1+i . 1−i 3 (a) Use the spectral theorem to write H as a linear combination of orthogonal projections. 1 2 −1 − i Answer: H = αP1 + βP2 where α = ,β = , P1 = , 3 1 1 1+i and P2 = . 3 (b) Find a square root of H. √ 1 4 1+i Answer: H = . 3

  4 + 2i 1 − i 1 − i 1 27.2.9. Exercise. Let N =  1 − i 4 + 2i 1 − i . 3 1 − i 1 − i 4 + 2i (a) The matrix N is normal because N N ∗   a b b = N ∗ N =  b a b  where a = b b a and

. b= (b) According to the spectral theorem N can be written as a linear combination of orthogonal , projections. Written in this form N λ1 P1 + λ2 P where λ1 =  = 2  a a a b −a −a λ2 = , P1 = a a a, and P2 = −a b −a where a = and a a a −a −a b b= .   a −b −c (c) A unitary matrix U which diagonalizes N is a b −c where a = ,b= , a d 2c c , . The associated diagonal form Λ = U ∗ N U of N is =  and d =     . 

27.2.10. Exercise (†). Let A be an n × n matrix A of complex numbers. Then A, regarded as an operator on Cn , is unitarily diagonalizable if and only if it is unitarily equivalent to a diagonal matrix. 27.2.11. Exercise. The real and imaginary parts of an element of a ∗ -algebra are self-adjoint. 27.2.12. Exercise. An element of a ∗ -algebra is normal if and only if its real part and its imaginary part commute. (See AOLT, page 23, proposition 1.25.) 27.2.13. Exercise. AOLT, section 1.8, exercise 19.

CHAPTER 28

Multilinear Maps
28.1. Background Topics: free objects, free vector space, the “little-oh” functions, tangency, diﬀerential, permutations, cycles, symmetric group, multilinear maps, alternating multilinear maps. Diﬀerential calculus The following four deﬁnitions introduce the concepts necessary for a (civilized) discussion of diﬀerentiability of real valued functions on Rn . In these deﬁnitions f and g are functions from Rn into R and a is a point in Rn . 28.1.1. Definition. For every h ∈ Rn let ∆fa := f (a + h) − f (a). 28.1.2. Definition. The function f belongs to the family o of “little-oh” functions if for every c > 0 there exists a δ > 0 such that f (x) ≤ c x whenever x < δ. 28.1.3. Definition. Functions f and g are tangent (at 0) if f − g ∈ o. 28.1.4. Definition. The function f is differentiable at the point a if there exists a linear map dfa from Rn into R which is tangent to ∆fa . We call dfa the differential (or total derivative) of f at a. 28.1.5. Remark. If a function f : Rn → R is diﬀerentiable, then at each point a in Rn the diﬀerential of f at a is a linear map from Rn into R. Thus we regard df : a → dfa (the differential of f ) as a map from Rn into L(Rn , R). It is natural to inquire whether the function df is itself diﬀerentiable. If it is, its diﬀerential at a (which we denote by d 2fa ) is a linear map from Rn into L(Rn , R); that is d 2fa ∈ L(Rn , L(Rn , R)). 2f maps Rn into L(Rn , L(Rn , R)), its diﬀerential (if it exists) belongs to In the same vein, since d L(Rn , L(Rn , L(Rn , R))). It is moderately unpleasant to contemplate what an element of L(Rn , L(Rn , R)) or of L(Rn , L(Rn , L(Rn , R))) might “look like”. And clearly as we pass to even higher order diﬀerentials things look worse and worse. It is comforting to discover that an element of L(Rn , L(Rn , R)) may be regarded as a map from (Rn )2 into R which is bilinear (that is, linear in both of its variables), and that L(Rn , L(Rn , L(Rn , R))) may be thought of as a map from (Rn )3 into R which is linear in each of its three variables. More generally, if V1 , V2 , V3 , and W are arbitrary vector spaces it will be possible to identify the vector space L(V1 , L(V2 , W ))) with the space of bilinear maps from V1 × V2 to W , the vector space L(V1 , L(V2 , L(V3 , W ))) with the trilinear maps from V1 × V2 × V3 to W , and so on (see exercise 28.2.3). Permutations A bijective map σ : X → X from a set X onto itself is a permutation of the set. If x1 , x2 , . . . , xn are distinct elements of a set X, then the permutation of X that maps x1 → x2 , x2 → x3 , . . . , xn−1 → xn , xn → x1 and leaves all other elements of X ﬁxed is a cycle (or cyclic
93

94

28. MULTILINEAR MAPS

permutation) of length n. A cycle of length 2 is a transposition. Permutations σ1 , . . . , σn of a set X are disjoint if each x ∈ X is moved by at most one σj ; that is, if σj (x) = x for at most one j ∈ Nn := {1, 2, . . . , n}. 28.1.6. Proposition. If X is a nonempty set, the set of permutations of X is a group under composition. Notice that if σ and τ are disjoint permutations of a set X, then στ = τ σ. If X is a set with n elements, then the group of permutations of X (which we may identify with the group of permutations of the set Nn ) is the symmetric group on n elements (or on n letters); it is denoted by Sn . 28.1.7. Proposition. Any permutation σ = idX of a set X can be written as a product (composite) of cycles of length at least 2. This decomposition is unique up to the order of the factors. A permutation of a set X is even if it can be written as the product of an even number of transpositions, and it is odd if it can be written as a product of an odd number of transpositions. 28.1.8. Proposition. Every permutation of a ﬁnite set is either even or odd, but not both. The sign of a permutation σ, denoted by sgn σ, is +1 if σ is even and −1 if σ is odd. Multilinear maps 28.1.9. Definition. Let V1 , V2 , . . . , Vn , and W be vector spaces over a ﬁeld F. We say that a function f : V1 × · · · × Vn → W is multilinear (or n-linear) if it is linear in each of its n variables. We ordinarily call 2-linear maps bilinear and 3-linear maps trilinear. We denote by Ln (V1 , . . . , Vn ; W ) the family of all n-linear maps from V1 × · · · × Vn into W . A multilinear map from the product V1 × · · · × Vn into the scalar ﬁeld F is a multilinear form (or a multilinear functional. 28.1.10. Definition. A multilinear map f : V n → W from the n-fold product V × · · · × V of a vector space V into a vector space W is alternating if f (v1 , . . . , vn ) = 0 whenever vi = vj for some i = j. 28.1.11. Definition. If V and W are vector spaces, a multilinear map f : V n → W is skewsymmetric if f (v 1 , . . . , v n ) = (sgn σ)f v σ(1) , . . . , v σ(n) for all σ ∈ Sn .

28.2. EXERCISES

(DUE: WED. APR. 1)

95

28.2. Exercises

(Due: Wed. Apr. 1)

28.2.1. Exercise. Let V and W be vector spaces, u, v, x, y ∈ V , and α ∈ R. (a) Expand T (u + v, x + y) if T is a bilinear map from V × V into W . (b) Expand T (u + v, x + y) if T is a linear map from V ⊕ V into W . (c) Write T (αx, αy) in terms of α and T (x, y) if T is a bilinear map from V × V into W . (d) Write T (αx, αy) in terms of α and T (x, y) if T is a linear map from V ⊕ V into W . 28.2.2. Exercise. Show that composition of operators on a vector space V is a bilinear map on L(V ). 28.2.3. Exercise. Show that if U , V , and W are vector spaces, then so is L2 (U, V ; W ). Show also that the spaces L(U, L(V, W )) and L2 (U, V ; W ) areisomorphic. Hint. The isomorphism is implemented by the map ˆ F : L(U, L(V, W )) → L2 (U, V ; W ) : φ → φ ˆ where φ(u, v) := (φ(u))(v) for all u ∈ U and v ∈ V . (Recall remark 28.1.5.) 28.2.4. Exercise. Let V = R2 and f : V 2 → R : (x, y) → x1 y2 . Is f bilinear? Is it alternating? 28.2.5. Exercise. Let V = R2 and g : V 2 → R : (x, y) → x1 +y2 . Is g bilinear? Is it alternating? 28.2.6. Exercise. Let V = R2 and h : V 2 → R : (x, y) → x1 y2 − x2 y1 . Is h bilinear? Is it alternating? If {e1 , e2 } is the usual basis for R2 , what is h(e1 , e2 )? 28.2.7. Exercise. Let V and W be vector spaces. Then every alternating multilinear map f : V n → W is skew-symmetric. Hint. Consider f (u + v, u + v) in the bilinear case.

CHAPTER 29

Determinants
29.1. Background Topics: determinant of a matrix, determinant function on Mn (A).

29.1.1. Definition. A ﬁeld F is of characteristic zero if n1 = 0 for no n ∈ N. NOTE: In the following material on determinants, we will assume that the scalar ﬁelds underlying all the vector spaces we encounter are of characteristic zero. Thus multilinear functions will be alternating if and only if they are skew-symmetric. (See exercises 28.2.7 and 29.2.3.) 29.1.2. Remark. Let A be a unital commutative algebra. In the sequel we identify the algebra n An = An × · · · × An (n factors) with the algebra Mn (A) of n × n matrices of elements of A by n regarding the term ak in (a1 , . . . , an ) ∈ An as the k th column vector of an n×n matrix of elements of A. There are many standard notations for the same thing: Mn (A), An × · · · × An (n factors), 2 n An , An×n , and An , for example. The identity matrix, which we usually denote by I, in Mn (A) is (e1 , . . . , en ), where e1 , . . . , en are the standard basis vectors for An ; that is, e1 = (1A , 0, 0, . . . ), e2 = (0, 1A , 0, 0, . . . ), and so on. 29.1.3. Definition. Let A be a unital commutative algebra. A determinant function is an alternating multilinear map D : Mn (A) → A such that D(I) = 1A .

97

98

29. DETERMINANTS

29.2. Exercises 29.2.1. Exercise. Let V = Rn . Deﬁne

(Due: Fri. Apr. 3)
1 n (sgn σ)vσ(1) . . . vσ(n) . σ∈Sn

∆ : V n → R : (v 1 , . . . , v n ) →

Then ∆ is an alternating multilinear form which satisﬁes ∆(e1 , . . . , en ) = 1. Note: If A is an n × n matrix of real numbers we deﬁne det A, the determinant of A, to be ∆(v 1 , . . . , v n ) where v 1 , . . . , v n are the column vectors of the matrix A.   1 3 2 29.2.2. Exercise. Let A = −1 0 3. Use the deﬁnition above to ﬁnd det A. −2 −2 1 29.2.3. Exercise. If V and W are vector spaces over a ﬁeld F of characteristic zero and f : V n → W is a skew-symmetric multilinear map, then f is alternating. 29.2.4. Exercise. Let ω be an n-linear functional on a vector space V over a ﬁeld of characteristic zero. If ω(v 1 , . . . , v n ) = 0 whenever v i = v i+1 for some i, then ω is skew-symmetric and therefore alternating. 29.2.5. Exercise. Let f : V n → W be an alternating multilinear map, j = k in Nn , and α be a scalar. Then f (v 1 , . . . , v j + αv k , . . . , v n ) = f (v 1 , . . . , v j , . . . , v n ).
↑ j ↑ j

29.2.6. Exercise. Let A be a unital commutative algebra and n ∈ N. A determinant function exists on Mn (A). Hint. Consider det : Mn (A) → A : (a1 , . . . an ) →
σ∈Sn

(sgn σ)a1 . . . an . σ(1) σ(n)

29.2.7. Exercise. Let D be an alternating multilinear map on Mn (A) where A is a unital commutative algebra and n ∈ N. For every C ∈ Mn (A) D(C) = D(I) det C. 29.2.8. Exercise. Show that the determinant function on Mn (A) (where A is a unital commutative algebra) is unique. 29.2.9. Exercise. Let A be a unital commutative algebra and B, C ∈ Mn (A). Then det(BC) = det B det C. Hint. Consider the function D(C) = D(c1 , . . . , cn ) := det(Bc1 , . . . , Bcn ), where Bck is the product of the n × n matrix B and the k th column vector of C. 29.2.10. Exercise. For an n × n matrix B let B t , the transpose of B, be the matrix obtained from B by interchanging its rows and columns; that is, if B = bj , then B t = bi . Prove that j i det B t = det B.

CHAPTER 30

Free Vector Spaces
30.1. Background Topics: free object, free vector space. (See AOLT, section 6.1.)

30.1.1. Definition. Let F be an object in a (concrete) category C and ι : S → F be a map whose domain is a nonempty set S. We say that the object F is free on the set S (or that F is the free object generated by S) if for every object A in C and every map f : S → A there exists a unique morphism fι : F → A in C such that fι ◦ ι = f . S?
ι / ?? ?? ?? ? f ? ?? ?

F

e fι

F
  e fι

A

A

We will be interested in free vector spaces; that is, free objects in the category VEC of vector spaces and linear maps. Naturally, merely deﬁning a concept does not guarantee its existence. It turns out, in fact, that free vector spaces exist on arbitrary sets. (See exercise 30.2.2.)

99

100

30. FREE VECTOR SPACES

30.2. Exercises

(Due: Mon. Apr. 6)

30.2.1. Exercise. If two objects in some concrete category are free on the same set, then they are isomorphic. 30.2.2. Exercise. Let S be an arbitrary nonempty set and F be a ﬁeld. Prove that there exists a vector space V over F which is free on S. Hint. Given the set S let V be the set of all F-valued functions on S which have ﬁnite support. Deﬁne addition and scalar multiplication pointwise. The map ι : s → χ{s} of each element s ∈ S to the characteristic function of {s} is the desired injection. To verify that V is free over S it must be shown that for every vector space W and every function S diagram commute.
f

/

W there exists a unique linear map V S?
ι / ?? ?? ?? ? f ? ?? ?

e f

/ W which makes the following

V

˜ f

V

˜ f

W

W

30.2.3. Exercise. Prove that every vector space is free. Hint. Of course, part of the problem is to specify a set S on which the given vector space is free. 30.2.4. Exercise. Let S = {a, ∗, #}. Then an expression such as √ 1 3a − 2 ∗ + 2 # is said to be a formal linear combination of elements of S. Make sense of such expressions.

CHAPTER 31

Tensor Products of Vector Spaces
31.1. Background Topics: vector space tensor products, elementary tensors. (See AOLT, section 6.2. A more extensive and very careful exposition of tensor products can be found in chapter 14 of [23].)

31.1.1. Definition. Let U and V be vector spaces. A vector space U ⊗ V together with a bilinear map τ : U × V → U ⊗ V is a tensor product of U and V if for every vector space W and every bilinear map B : U × V → W , there exists a unique linear map B : U ⊗ V → W which makes the following diagram commute. U ×V ??
τ

?? ?? ?? e B B ? ?? ? 

/ U ⊗V

W

101

102

31. TENSOR PRODUCTS OF VECTOR SPACES

31.2. Exercises

(Due: Wed. Apr. 8)

31.2.1. Exercise. Prove that in the category of vector spaces and linear maps if tensor products exist, then they are unique (up to isomorphism). 31.2.2. Exercise. Show that in the category of vector spaces and linear maps tensor products exist. Hint. Let U and V be vector spaces over a ﬁeld F. Consider the free vector space lc (U ×V ) = lc (U × V , F). Deﬁne ∗ : U × V → lc (U × V ) : (u, v) → χ{(u,v)} . Then let S1 = {(u1 + u2 ) ∗ v − u1 ∗ v − u2 ∗ v : u1 , u2 ∈ U and v ∈ V }, S2 = {(αu) ∗ v − α(u ∗ v) : α ∈ F, u ∈ U , and v ∈ V }, S3 = {u ∗ (v1 + v2 ) − u ∗ v1 − u ∗ v2 : u ∈ U and v1 , v2 ∈ V }, S4 = {u ∗ (αv) − α(u ∗ v) : α ∈ F, u ∈ U , and v ∈ V }, S = span(S1 ∪ S2 ∪ S3 ∪ S4 ), and U ⊗ V = lc (U × V )/S . Also deﬁne τ : U × V → U ⊗ V : (u, v) → [u ∗ v]. Then show that U ⊗ V and τ satisfy the conditions stated in deﬁnition 31.1.1. NOTE: It is conventional to write u ⊗ v for τ (u, v) = [u ∗ v]. Tensors of the form u ⊗ v are called elementary tensors (or decomposable tensors or homogeneous tensors). Keep in mind that not every member of U ⊗ V is of the form u ⊗ v. 31.2.3. Exercise. Let u and v be elements of ﬁnite dimensional vector spaces U and V , respectively. Show that if u ⊗ v = 0, then either u = 0 or v = 0. 31.2.4. Exercise. Let u1 , . . . , un be linearly independent vectors in a vector space U and v1 , . . . , vn be arbitrary vectors in a vector space V . Prove that if n uk ⊗ vk = 0, then vk = 0 k=1 for each k ∈ Nn . 31.2.5. Exercise. Let {ei }m and {fj }n be bases for vector spaces U and V , respectively. i=1 j=1 m n Show that the family {ei ⊗ fj }i=1, j=1 is a basis for U ⊗ V . Conclude that if U and V are ﬁnite dimensional, then the dimension of U ⊗ V is the product of the dimensions of U and V . 31.2.6. Exercise. Let U and V be ﬁnite dimensional vector spaces and {fj }n be a basis j=1 for V . Show that for every element t ∈ U ⊗ V there exist unique vectors u1 , . . . un ∈ U such that
n

t=
j=1

uj ⊗ fj .

31.2.7. Exercise. If U and V are vector spaces, then U ⊗ V ∼ V ⊗ U. = 31.2.8. Exercise. If V is a vector space over a ﬁeld F, then V ⊗ F ∼ V ∼ F ⊗ V. = =

CHAPTER 32

Tensor Products of Vector Spaces (continued)
32.1. Background Topics: No additional topics.

103

104

32. TENSOR PRODUCTS OF VECTOR SPACES (CONTINUED)

32.2. Exercises

(Due: Fri. Apr. 10)

32.2.1. Exercise. Let U , V , and W be vector spaces. For every vector space X and every trilinear map k : U × V × W → X there exists a unique linear map k : U ⊗ (V ⊗ W ) : → X such that k u ⊗ (v ⊗ w) = k(u, v, w) for all u ∈ U , v ∈ V , and w ∈ W . 32.2.2. Exercise. If U , V , and W are vector spaces, then U ⊗ (V ⊗ W ) ∼ (U ⊗ V ) ⊗ W. = 32.2.3. Exercise. If U and V are ﬁnite dimensional vector spaces, then U ⊗ V ∗ ∼ L(V, U ). = Hint. Consider the map T : U × V ∗ → L(V, U ) : (u, φ) → T (u, φ) where T (u, φ) (v) = φ(v) u. 32.2.4. Exercise. If U , V , and W are vector spaces, then U ⊗ (V ⊕ W ) ∼ (U ⊗ V ) ⊕ (U ⊗ W ). = 32.2.5. Exercise. If U and V are ﬁnite dimensional vector spaces, then (U ⊗ V )∗ ∼ U ∗ ⊗ V ∗ . = 32.2.6. Exercise. If U , V , and W are ﬁnite dimensional vector spaces, then L(U ⊗ V, W ) ∼ L(U, L(V, W )) ∼ L2 (U, V ; W ). = = 32.2.7. Exercise. Let u1 , u2 ∈ U and v1 , v2 ∈ V where U and V are ﬁnite dimensional vector spaces. Show that if u1 ⊗ v1 = u2 ⊗ v2 = 0, then u2 = αu1 and v2 = βv1 where αβ = 1.

CHAPTER 33

Tensor Products of Linear Maps
33.1. Background Topics: tensor products of linear maps.

33.1.1. Definition. Let S : U → W and T : V → X be linear maps between vector spaces. We deﬁne the tensor product of the linear maps S and T by S ⊗ T : U ⊗ V → W ⊗ X : u ⊗ v → S(u) ⊗ T (v) .

105

106

33. TENSOR PRODUCTS OF LINEAR MAPS

33.2. Exercises

(Due: Mon. April 13)

33.2.1. Exercise. Deﬁnition 33.1.1 deﬁnes the tensor product S ⊗ T of two maps only for homogeneous elements of U ⊗ V . Explain exactly what is needed to convince ourselves that S ⊗ T is well deﬁned on all of U ⊗ V . Then prove that S ⊗ T is a linear map. 33.2.2. Exercise. Some authors hesitate to use the notation S ⊗ T for the mapping deﬁned in 33.1.1 on the (very reasonable) grounds that S ⊗ T already has a meaning; it is a member of the vector space L(U, W ) ⊗ L(V, X). Discuss this problem and explain, in particular, why the use of the notation S ⊗ T in 33.1.1 is not altogether unreasonable. 33.2.3. Exercise. Suppose that R, S ∈ L(U, W ) and that T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Then (R + S) ⊗ T = R ⊗ T + S ⊗ T. 33.2.4. Exercise. Suppose that R ∈ L(U, W ) and that S, T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Then R ⊗ (S + T ) = R ⊗ S + R ⊗ T. 33.2.5. Exercise. Suppose that S ∈ L(U, W ) and that T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Then for all α, β ∈ R (αS) ⊗ (βT ) = αβ(S ⊗ T ). 33.2.6. Exercise. Suppose that Q ∈ L(U, W ), R ∈ L(V, X), S ∈ L(W, Y ), and that T ∈ L(X, Z) where U , V , W , X, Y , and Z are ﬁnite dimensional vector spaces. Then (S ⊗ T )(Q ⊗ R) = SQ ⊗ T R. 33.2.7. Exercise. Suppose that S ∈ L(U, W ) and that T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Show that if S and T are invertible, then so is S ⊗ T and (S ⊗ T )−1 = S −1 ⊗ T −1 . 33.2.8. Exercise. Suppose that S ∈ L(U, W ) and that T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Show that if S ⊗ T = 0, then either S = 0 or T = 0. 33.2.9. Exercise. Suppose that S ∈ L(U, W ) and that T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Then ran(S ⊗ T ) = ran S ⊗ ran T. 33.2.10. Exercise. Suppose that S ∈ L(U, W ) and that T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Then ker(S ⊗ T ) = ker S ⊗ V + U ⊗ ker T. 33.2.11. Exercise. Suppose that S ∈ L(U, W ) and that T ∈ L(V, X) where U , V , W , and X are ﬁnite dimensional vector spaces. Then (S ⊗ T ) t = S t ⊗ T t . 33.2.12. Exercise. If U and V are ﬁnite dimensional vector spaces, then IU ⊗ IV = IU ⊗V .

CHAPTER 34

Grassmann Algebras
34.1. Background Topics: group algebras, Grassmann algebras, wedge product.

34.1.1. Definition. Let V be an d-dimensional vector space over a ﬁeld F. We say that (V ) is the Grassmann algebra (or the exterior algebra) over V if (1) (V ) is a unital algebra over F (multiplication is denoted by ∧), (2) V is “contained in” (V ), (3) v ∧ v = 0 for every v ∈ V , (4) dim( (V )) = 2d , and (5) (V ) is generated by 1Λ(V ) and V . The multiplication ∧ in a Grassmann algebra is called the wedge product (or the exterior product).

107

108

34. GRASSMANN ALGEBRAS

34.2. Exercises

(Due: Wed. April 15)

34.2.1. Exercise. Near the top of page 48 in AOLT, the author, in the process of deﬁning the group algebra FG, says that, “We regard the elements of [a group] G as linearly independent vectors, . . . .” Explain carefully what this means and why we may in fact do such a thing. 34.2.2. Exercise. In the middle of page 48 of AOLT the author gives a “succinct formula” for the product of two elements in a group algebra. Show that this formula is correct. 34.2.3. Exercise. Check the author’s computation of xy in Example 2.10, page 48, AOLT. Find the product both by using the formula given in the middle of page 48 and by direct multiplication. 34.2.4. Exercise. AOLT, section 2.7, exercise 7. 34.2.5. Exercise. In deﬁnition 34.1.1 why is “contained in” in quotation marks. Give a more precise version of condition (2). 34.2.6. Exercise. In the last condition of deﬁnition 34.1.1 explain more precisely what is meant by saying that (V ) is generated by 1 and V . 34.2.7. Exercise. Show that if 1V(V ) ∈ V . / (V ) is a Grassmann algebra over a vector space V , then

34.2.8. Exercise. Let v and w be elements of a ﬁnite dimensional vector space V . Show that in the Grassmann algebra (V ) generated by V v ∧ w = −w ∧ v . 34.2.9. Exercise. Let V be a d-dimensional vector space with basis E = {e1 , . . . , ed }. For each nonempty subset S = {ei1 , ei2 , . . . , eip } of E let eS = ei1 ∧ ei2 ∧ · · · ∧ eip . Also let e∅ = 1V(V ) . Show that {eS : S ⊆ E} is a basis for the Grassmann algebra (V ).

CHAPTER 35

35.1. Background Topics: graded algebras, homogeneous elements, decomposable elements.

35.1.1. Definition. An algebra A is a Z+ -graded algebra if it is a direct sum A =
k≥0

Ak

of vector subspaces Ak and its multiplication ∧ takes elements in Aj × Ak to elements in Aj+k for all j, k ∈ Z+ . Elements in Ak are said to be homogeneous of degree k. The deﬁnitions of Z-graded algebras, N-graded algebras and Z2 -graded algebras are similar. (In the case of a Z2 -graded algebra the indices are 0 and 1 and A1 ∧ A1 ⊆ A0 .) Usually the unmodiﬁed expression “graded algebra” refers to an Z+ -graded algebra. Exercise 35.2.1 says that every Grassmann algebra (V ) over a vector space V is a graded algebra. The set of elements homogeneous of degree k is denoted by k (V ). An element of k (V ) which can be written in the form v1 ∧ v2 ∧ · · · ∧ vk (where v1 , . . . , vk all belong to V ) is a decomposable element of degree k.

109

110

35.2. Exercises
k

(Due: Fri. April 17)

35.2.1. Exercise. Show that every Grassmann algebra is an Z+ -graded algebra. We denote by (V ) the subspace of all homogeneous elements of degree k in (V ). In particular, 0 (V ) = R and 1 (V ) = V . If the dimension of V is d, take k (V ) = {0} for all k > d. (And if you wish to regard (V ) as a Z-graded algebra also take k (V ) = {0} whenever k < 0.) 35.2.2. Exercise. If the dimension of a vector space V is 3 or less, then every homogeneous element of the corresponding Grassmann algebra is decomposable. 35.2.3. Exercise. If the dimension of a (ﬁnite dimensional) vector space V is at least four, then there exist homogeneous elements in the corresponding Grassmann algebra which are not decomposable. Hint. Let e1 , e2 , e3 , and e4 be distinct basis elements of V and consider (e1 ∧ e2 ) + (e3 ∧ e4 ). 35.2.4. Exercise. The elements v1 , v2 , . . . , vp in a vector space V are linearly independent if and only if v1 ∧ v2 ∧ · · · ∧ vp = 0 in the corresponding Grassmann algebra (V ). 35.2.5. Exercise. Let T : V → W be a linear map between ﬁnite dimensional vector spaces (over the same ﬁeld). Then there exists a unique extension of T to a unital algebra homomorphism (T ) : (V ) → (W ). This extension maps k (V ) into k (W ) for each k ∈ N. 35.2.6. Exercise. The pair of maps V → (V ) and T → (T ) is a covariant functor from the category of vector spaces and linear maps to the category of unital algebras and unital algebra homomorphisms. 35.2.7. Exercise. If V is a vector space of dimension d, then dim 35.2.8. Exercise. If V is a ﬁnite dimensional vector space, ω ∈ ω ∧ µ = (−1) µ ∧ ω.
pq p p

(V ) =

d p

for 0 ≤ p ≤ d.
q

(V ), and µ ∈

(V ), then

CHAPTER 36

Existence of Grassmann Algebras
36.1. Background Topics: tensor algebras, shuﬄe permutations, (A good reference for much of this material—and the following material on diﬀerential forms—is chapter 4 of [3].)

36.1.1. Definition. Let V0 , V1 , V2 , . . . be vector spaces (over the same ﬁeld). Then their

(external) direct sum, which is denoted by
k=0

Vk , is deﬁned to be the set of all functions

v : Z+ → ∞ Vk with ﬁnite support such that v(k) = vk ∈ Vk for each k ∈ Z+ . The usual k=0 pointwise addition and scalar multiplication make this set into a vector space. 36.1.2. Definition. Let V be a vector space over a ﬁeld F. Deﬁne T 0 (V ) = F, T 1 (V ) = V , ) = V ⊗ V , T 3 (V ) = V ⊗ V ⊗ V , . . . , T k (V ) = V ⊗ · · · ⊗ V (k factors), . . . . Then let

T

2 (V

T (V ) =
k=0

T k (V ). Deﬁne multiplication on T (V ) by using the obvious isomorphism T k (V ) ⊗ T m (V ) ∼ T k+m (V ) =

and extending by linearity to all of T (V ). The resulting algebra is the tensor algebra of V (or generated by V ). 36.1.3. Notation. If V is a vector space over F and k ∈ N we denote by Altk (V ) the set of all alternating k-linear maps from V k into F. (The space Alt1 (V ) is just V ∗ .) Additionally, take Alt0 (V ) = F. 36.1.4. Definition. Let p, q ∈ N. We say that a permutation σ ∈ Sp+q is a (p, q)-shuffle if σ(1) < · · · < σ(p) and σ(p + 1) < · · · < σ(p + q). The set of all such permutations is denoted by S(p, q). 36.1.5. Definition. Let V be a vector space. For p, q ∈ N deﬁne ∧ : Altp (V ) × Altq (V ) → Altp+q (V ) : (ω, µ) → ω ∧ µ where (ω ∧ µ)(v 1 , . . . , v p+q ) =
σ∈S(p,q)

(sgn σ)ω(v σ(1) , . . . , v σ(p) )µ(v σ(p+1) , . . . , v σ(p+q) ).

111

112

36. EXISTENCE OF GRASSMANN ALGEBRAS

36.2. Exercises

(Due: Mon. April 20)

36.2.1. Exercise. Show that T (V ) as deﬁned in 36.1.2 is in fact a unital algebra. 36.2.2. Exercise. Let V be a ﬁnite dimensional vector space and J be the ideal in the tensor algebra T (V ) generated by the set of all elements of the form v ⊗ v where v ∈ V . Show that the quotient algebra T (V )/J is the Grassmann algebra over V ∗ (or, equivalently, over V ). 36.2.3. Exercise. Show that if V is a ﬁnite dimensional vector space and k > dim V , then Altk (V ) = {0}. 36.2.4. Exercise. Give an example of a (4, 5)-shuﬄe permutation σ of the set N9 = {1, . . . , 9} such that σ(7) = 4. 36.2.5. Exercise. Show that deﬁnition 36.1.5 is not overly optimistic by verifying that if ω ∈ Altp (V ) and µ ∈ Altq (V ), then ω ∧ µ ∈ Altp+q (V ). 36.2.6. Exercise. Show that the multiplication deﬁned in 36.1.5 is associative. That is if ω ∈ Altp (V ), µ ∈ Altq (V ), and ν ∈ Altr (V ), then ω ∧ (µ ∧ ν) = (ω ∧ µ) ∧ ν. 36.2.7. Exercise. Let V be a ﬁnite dimensional vector space. Explain in detail how to make Altk (V ) (or, if you prefer, Altk (V ∗ ) ) into a vector space for each k ∈ Z and how to make the collection of these into a Z-graded algebra. Show that this algebra is the Grassmann algebra generated by V . Hint. Take Altk (V ) = {0} for each k < 0 and extend the deﬁnition of the wedge product so that if α ∈ Alt0 (V ) = F and ω ∈ Altp (V ), then α ∧ ω = αω. 36.2.8. Exercise. Let ω1 , . . . , ωp be members of Alt1 (V ) (that is, linear functionals on V ). Then p (ω1 ∧ · · · ∧ ωp )(v 1 , . . . , v p ) = det ωj (v k ) j,k=1 for all v 1 , . . . , v p ∈ V . 36.2.9. Exercise. If {e1 , . . . , en } is a basis for an n-dimensional vector space V , then {e∗ ∧ · · · ∧ e∗ σ(1) σ(p) : σ ∈ S(p, n − p)} is a basis for Altp (V ).

CHAPTER 37

The Hodge ∗-operator
37.1. Background Topics: right-handed basis, orientation, opposite orientation, Hodge star operator.

37.1.1. Definition. Let E be a basis for an n-dimensional vector space V . Then the n-tuple (e1 , . . . , en ) is an ordered basis for V if e1 , . . . , en are distinct elements of E. 37.1.2. Definition. Let E = (e1 , . . . , en ) be an ordered basis for Rn . We say that the basis E is right-handed if det[e1 , . . . , en ] > 0 and left-handed otherwise. 37.1.3. Definition. Let V be a real n-dimensional vector space and T : Rn → V be an isomorphism. Then the set of all n-tuples of the form (T (e1 ), . . . , T (en )) where (e1 , . . . , en ) is a right-handed basis in Rn is an orientation of V . Another orientation consists of the set of ntuples (T (e1 ), . . . , T (en )) where (e1 , . . . , en ) is a left-handed basis in Rn . Each of these orientations is the opposite (or reverse) of the other. A vector space together with one of these orientations is an oriented vector space.

113

114

37. THE HODGE ∗-OPERATOR

37.2. Exercises

(Due: Wed. April 22)

37.2.1. Exercise. For T : V → W a linear map between vector spaces deﬁne Altp (T ) : Altp (W ) → Altp (V ) : ω → Altp (T )(ω) where Altp (T )(ω) (v 1 , . . . , v p ) = ω(T v 1 , . . . , T v p ) for all v 1 , . . . , v p ∈ V . Then Altp is a contravariant functor from the category of vector space and linear maps into itself. 37.2.2. Exercise. Let V be an n-dimensional vector space and T ∈ L(V ). If T is diagonalizable, then
n

cT (λ) =
k=0

(−1)k [Altn−k (T )]λk .

37.2.3. Exercise. Let V be an n-dimensional real inner product space. In exercise 9.2.13 we established an isomorphism Φ : v → v ∗ between V and its dual space V ∗ . Show how this isomorphism can be used to induce an inner product on V ∗ . Then show how this may be used to create an inner product on Altp (V ) for 2 ≤ p ≤ n. Hint. For v, w ∈ V let v ∗ , w∗ = v, w . Then for ω1 , . . . , ωp , µ1 , . . . , µp ∈ Alt1 (V ) let ω1 ∧ · · · ∧ ωp , µ1 ∧ · · · ∧ µp = det[ ωj , µk ]. 37.2.4. Exercise. Let V be an d-dimensional oriented real inner product space. Fix a unit vector vol ∈ Altd (V ). This vector is called a volume element. (In the case where V = Rd , we will always choose vol = e∗ ∧ · · · ∧ e∗ where (e1 , . . . , ed ) is the usual ordered basis for Rd .) 1 d Let ω ∈ Altp (V ) and q = d − p. Show that there exists a vector ∗ ω ∈ Altq (V ) such that ∗ ω, µ vol = ω ∧ µ for each µ ∈ Alt (V ). Show that the map ω → ∗ ω from Altp (V ) into Altq (V ) is a vector space isomorphism. This map is the Hodge star operator. 37.2.5. Exercise. Let V be a ﬁnite dimensional oriented real inner product space of dimension n. Suppose that p + q = n. Show that ∗ ∗ ω = (−1)pq ω for every ω ∈ Altp (V ).
q

CHAPTER 38

Diﬀerential Forms
38.1. Background Topics: tangents, tangent space, cotangent space, partial derivatives on a manifold, vector ﬁelds, diﬀerential forms, the diﬀerential of a map between manifolds. (One good source for much of this material is chapters 1 and 4 of [3].

In everything that follows all vector spaces are assumed to be real, ﬁnite dimensional, and oriented; and all manifolds are smooth oriented differentiable manifolds.
∞ 38.1.1. Notation. Let m be a point on a manifold M . A real valued function f belongs to Cm if it is deﬁned in a neighborhood of m and is smooth. (Recall that a function is smooth if it has derivatives of all orders).

38.1.2. Definition. Let m be a point on a manifold M . A tangent at m is a linear map ∞ t : Cm → R such that t(f g) = f (m)t(g) + f (t)g(m)
∞ for all f , g ∈ Cm . The set Tm of all tangents at m is called the tangent space at m. The dual ∗ Tm of the tangent space at m is called the cotangent space at m.

38.1.3. Definition. Let U ⊆ M where M is a manifold. A function X deﬁned on U is a vector field on U if Xm (also written as X(m)) belongs to the tangent space Tm for each point m ∈ U. ∞ For each f ∈ Cm deﬁne the function Xf by (Xf )(m) = Xm (f )
∞ for all m in the intersection of the domains of X and f . A vector ﬁeld X is smooth if Xf ∈ Cm ∞ . Thus a smooth vector ﬁeld is often regarded as a mapping from C ∞ into itself. whenever f ∈ Cm m In everything that follows we will assume that all vector ﬁelds are smooth.

38.1.4. Definition. Let U ⊆ M where M is a manifold. A function ω deﬁned on U is a ∗ differential p -form on U if ωm (also written as ω(m)) belongs to p (Tm ) for all m ∈ U . 1 , . . . , X p are vector ﬁelds on U the function A diﬀerential p -form is smooth if whenever X ω(X 1 , . . . , X p ) deﬁned on U by
1 p ω(X 1 , . . . , X p )(m) = ωm (Xm , . . . , Xm )

is smooth. The vector space of all smooth diﬀerential p-forms on U is denoted by Ωp (U ). If M is a manifold of dimension d, we agree (as in exercise 35.2.1 that Ωp (U ) = 0 whenever p < 0 or p > d. The space ∗ of all diﬀerential forms on U is denoted by Ω(U ). Since 0 (Tm ) = R, a member of Ω0 (U ) is just a 0 (U ) = C ∞ (U ). We denote the Z-graded algebra generated smooth real valued function on U , so Ω
115

116

38. DIFFERENTIAL FORMS

by the vector spaces Ωp (U ) and the wedge product by Ω(U ). In everything that follows when we refer to a p-form or a diﬀerential form we will assume that it is smooth. 38.1.5. Definition. Let φ = (x1 , . . . , xd ) be a chart (or coordinate system) at a point m on a d-dimensional manifold M . We deﬁne the partial derivative at m with respect to xk , denoted by Dxk (m), to be the tangent given by the formula Dxk (m)(f ) = ∂(f ◦ φ−1 ) (φ(m)) ∂uk

where ∂/∂uk is the usual partial derivative with respect to the k th coordinate on Rd . Depending on which variable we are interested in we often write Dxk (f )(m) for Dxk (m)(f ). It is common ∂ practice to write ∂xk for Dxk . In Rn we often write Dk for Duk , where, as above, u1 , . . . , un are the standard coordinates on Rn . Also, if f is a smooth real valued function on a manifold, we may write fxk (or just fk ) for Dxk (f ). 38.1.6. Theorem. Let φ = (x1 , . . . , xd ) be a chart (or coordinate system) at a point m on a d-dimensional manifold M . If t is a tangent at m, then
d

t=
k=1

t(xk ) Dxk

For a proof of this theorem consult [3] (section 1.3, theorem 1) or [25] (chapter 1, proposition 3.1). 38.1.7. Definition. Let F : M → N be a smooth function between manifolds and m be a point in M . Deﬁne dFm : Tm → TF (m) , the differential of F at m, by dFm (t)(g) = t(g ◦ F ) for all ∞ t ∈ Tm and all g ∈ CF (m) . 38.1.8. Notation. In exercise 37.2.4 we deﬁned a volume element vol in Rd . We will adopt the same notation vol for the diﬀerential form dx1 ∧· · ·∧dxd on a d-manifold M (where φ = (x1 , . . . , xd ) is a chart on M ). This is called the volume form.

38.2. EXERCISES

(DUE: FRI. APRIL 24)

117

38.2. Exercises

(Due: Fri. April 24)

38.2.1. Exercise. Let m be a point on a d-dimensional manifold M . Show that the tangent space Tm is a vector space of dimension d. 38.2.2. Exercise. Let F : M → N be a smooth function between manifolds and m be a point in M . Show that dFm (t) ∈ TF (m) whenever t ∈ Tm and that dFm ∈ L(Tm , TF (m) ). 38.2.3. Exercise (Chain Rule). Let F : M → N and G : N → P be smooth functions between manifolds. Prove that d(G ◦ F )m = dGF (m) ◦ dFm . 38.2.4. Exercise. Help future students. Design some reasonably elementary exercises which you think would help clarify points in the background material for this assignment that you found confusing at ﬁrst reading.

CHAPTER 39

The Exterior Diﬀerentiation Operator
39.1. Background Topics: exterior derivative.

39.1.1. Definition. Let U be an open subset of a manifold M . The exterior differentiation operator (or exterior derivative) d is a mapping which takes k-forms on U to (k + 1)-forms on U . That is, d : Ωk (U ) → Ωk+1 (U ). It is deﬁned to have the following properties: (i) If f is a 0-form on U , then d(f ) is the usual diﬀerential df of f (as deﬁned in exercise 38.1.7; (ii) d is linear; (iii) d2 = 0 (that is, d(dω) = 0 for every k-form ω); and (iv) if ω is a k-form and µ is any diﬀerential form, then d(ω ∧ µ) = (dω) ∧ µ + (−1)k ω ∧ dµ. Proofs of the existence and uniqueness of such a function can be found in [20] (theorem 12.14), [25] (chapter 1, theorem 11.1), and [3] (section 4.6).

119

120

39. THE EXTERIOR DIFFERENTIATION OPERATOR

39.2. Exercises

(Due: Mon. April 27)

39.2.1. Exercise. Theorem 38.1.6 tells us that at each point m of a d-dimensional manifold ∂ ∂ the vectors (m). . . . , (m) constitute a basis for the tangent space at m. Explain why the ∂x1 ∂xd vectors d x1 m , . . . , d xd m make up its dual basis. 39.2.2. Exercise. If f is a smooth real valued function on an open subset U of Rd , then
d

df =
k=1

fk dxk .

39.2.3. Exercise. Let φ = (x, y, z) be a chart at a point m in a 3-manifold M and let U be the domain of φ. Explain why Ω0 (U ), Ω1 (U ), Ω2 (U ), Ω3 (U ), and k∈Z Ωk (U ) are vector spaces and exhibit a basis for each. 39.2.4. Exercise. In beginning calculus texts some curious arguments are given for replacing the expression dx dy in the integral R f dx dy by r dr dθ when we change from rectangular to polar coordinates in the plane. Show that if we interpret dx dy as the diﬀerential form dx ∧ dy, then this is a correct substitution. (Assume additionally that R is a region in the open ﬁrst quadrant and that the integral of f over R exists.) 39.2.5. Exercise. Give an explanation similar to the one in the preceding exercise of the change in triple integrals from rectangular to spherical coordinates. 39.2.6. Exercise. Generalize the two preceding exercises. 39.2.7. Exercise. Let φ = (x1 , . . . , xd ) be a chart at a point m on a d-dimensional manifold. If ω is a diﬀerential form deﬁned on U then it can be written in the form ωJ dxJ where J ranges J = dx ∧ · · · ∧ dx J over all subsets J = {j1 , . . . , jk } of Nd with j1 < · · · < jk , dx j1 jk (letting dx = 1 when J = ∅), and ωJ is a smooth real valued function on U . Show that dωm = dωJ (m) ∧ dxJ (m).

CHAPTER 40

Diﬀerential Calculus on R3
40.1. Background Topics: association of vector ﬁelds with 1-forms, gradient, curl, divergence.

40.1.1. Notation. If, on some manifold, f is a 0-form and ω is a diﬀerential form (of arbitrary degree), we write f ω instead of f ∧ ω for the product of the forms. 40.1.2. Definition. Let P , Q, and R be scalar ﬁelds on an open subset U of R3 . In 38.1.3 we deﬁned the term vector ﬁeld. Recall that this term is used in a somewhat diﬀerent fashion in beginning calculus. There it refers simply to a mapping from R3 to R3 . So in this more elementary context a vector ﬁeld F can be written in the form P i+Q j+R k. We say that ω = P dx+Q dy+R dz is the diﬀerential 1-form associated with the vector ﬁeld F = P i + Q j + R k. (And, of course, F is the vector ﬁeld associated with ω.) Notice that a vector ﬁeld is identically zero if and only if its associated 1-form is.

121

122

40. DIFFERENTIAL CALCULUS ON R3

40.2. Exercises

(Due: Wed. April 29)

Recall that in beginning calculus the cross product is deﬁned only for vectors in R3 . The following exercise makes clear the sense in which the wedge product is a generalization of the cross product. 40.2.1. Exercise. Show that there exists an isomorphism u → u# from R3 to Alt2 (R3 ) such that v ∗ ∧ w∗ = (v × w)# for every v, w ∈ R3 . 40.2.2. Exercise. Let M be a 3-manifold, φ = (x, y, z) : U → R3 be a chart on M , and f : U → R be a 0-form on U . Compute d(f dy). 40.2.3. Exercise. Let M be a 3-manifold and φ = (x, y, z) : U → R3 be a chart on M . Compute d(x dy ∧ dz + y dz ∧ dx + z dx ∧ dy). 40.2.4. Exercise. Let M be a 3-manifold and φ = (x, y, z) : U → R3 be a chart on M . Compute d[(3xz dx + xy 2 dy) ∧ (x2 y dx − 6xy dz)]. 40.2.5. Exercise. Let M be a d-manifold, φ = (φ1 , . . . , φd ) : U → Rd be a chart on M , and f : U → R be a 0-form on U . Show that ∗ (f dφk ) = f ∗ dφk for 1 ≤ k ≤ d. 40.2.6. Exercise. Let M be a 3-manifold, φ = (x, y, z) : U → R3 be a chart on M , and a, b, c : U → R be 0-forms on U . Compute the following: (1) ∗ a (2) ∗ (a dx + b dy + c dz) (3) ∗ (a dx ∧ dy + b dx ∧ dz + c dy ∧ dz) and (4) ∗ (a dx ∧ dy ∧ dz) 40.2.7. Exercise. Explain how the Hodge star operator is extended to diﬀerential forms on Rn . For diﬀerential forms on an open subset U of R3 prove that ∗ dx = dy ∧ dz, ∗ dy = dz ∧ dx, and ∗ dz = dx ∧ dy. 40.2.8. Exercise. Let f be a scalar ﬁeld on an open subset U of R3 . Show that grad f (the gradient of f ) is the vector ﬁeld associated with the 1-form df . 40.2.9. Exercise. Let F and G be vector ﬁelds on a region in R3 . Show that if ω and µ are, respectively, their associated 1-forms, then ∗ (ω ∧ µ) is the 1-form associated with F × G. 40.2.10. Exercise. Let ω = x2 yz dx + yz dy + xy 2 dz. Find d ∗ ω and ∗ dω. 40.2.11. Exercise. Let f be a scalar ﬁeld. Write the left side of Laplace’s equation fxx + fyy + fzz = 0 in terms of d, ∗, and f only. 40.2.12. Exercise. Suppose that F is a smooth vector ﬁeld in R3 and that ω is its associated 1-form. Show that ∗ dω is the 1-form associated with curl F. 40.2.13. Exercise. Let F be a vector ﬁeld on R3 and ω be its associated 1-form. Show that ∗ d ∗ ω = div F. (Here div F is the divergence of F .) 40.2.14. Exercise. Let F be a vector ﬁeld on an open subset of R3 . Use diﬀerential forms (but not partial derivatives) to show that div curl F = 0.

40.2. EXERCISES

(DUE: WED. APRIL 29)

123

40.2.15. Exercise. Let f be a smooth scalar ﬁeld (that is, a 0-form) in R3 . Use diﬀerential forms (but not partial derivatives) to show that curl grad f = 0.

CHAPTER 41

Closed and Exact Forms
41.1. Background Topics: closed diﬀerential forms, exact diﬀerential forms,

41.1.1. Definition. Let U be an open subset of a manifold and ···
/ Ωp−1 (U )
dp−1

/ Ωp (U )

dp

/ Ωp+1 (U )

/ ···

where dp−1 and dp are exterior diﬀerentiation operators. Elements of ker dp are called closed p-forms and elements of ran dp−1 are exact p-forms.

125

126

41. CLOSED AND EXACT FORMS

41.2. Exercises

(Due: Mon. May 4)

41.2.1. Exercise. Writers of elementary texts in calculus and physics in their otherwise laudable eﬀorts to express complicated ideas in a simple fashion will occasionally lose all semblance of self control and write things like dr = dx i + dy j + dz k and dS = dy ∧ dz i + dz ∧ dx j + dx ∧ dy k. In the depths of their depravity they may even go so far as to claim for a vector ﬁeld F = (F 1 , F 2 , F 3 ) in R3 that the derivative of F · dr is curl F · dS. Is it possible to make any sense of such a statement? 41.2.2. Exercise. With the shameless notations suggested in exercise 41.2.1 try to make sense of the claim that for a vector ﬁeld F in R3 d(F · dS) = (div F) dV where dV is the “volume element” dx ∧ dy ∧ dz. 41.2.3. Exercise. Show that every exact diﬀerential form is closed. 41.2.4. Exercise. Show that if ω and µ are closed diﬀerential forms, then so is ω ∧ µ. 41.2.5. Exercise. Show that if ω is an exact diﬀerential form and µ is a closed form, then ω ∧ µ is exact. 41.2.6. Exercise. Let φ = (x, y, z) : U → R3 be a chart on a manifold and ω = a dx+b dy +c dz ∂c ∂b ∂a ∂c ∂b ∂a be a 1-form on U . Show that if ω is exact, then = , = , and = . ∂y ∂z ∂z ∂x ∂x ∂y 41.2.7. Exercise. Explain why solving the initial value problem ex cos y + 2x − ex (sin y)y = 0, is essentially the same thing as showing that the 1-form Do it. y(0) = π/3 (ex cos y + 2x) dx − ex (sin y) dy is exact.

CHAPTER 42

The de Rham Cohomology Group
42.1. Background Topics: cocycles, coboundaries, the de Rham cohomology group. 42.1.1. Definition. Let M be a d-manifold. For 0 ≤ p ≤ d we denote by Z p (M ) (or just Z p ) the vector space of all closed p -forms on M . For reason which will become apparent shortly, members of Z p are sometimes called (de Rham)p -cocycles. Also let B p (M ) (or just B p ) denote the vector space of all exact p-forms on M (sometimes called p -coboundaries). Since there are no diﬀerential forms of degree less than 0, we take B 0 = B 0 (M ) = {0}. For convenience we also set Z p = Z p (M ) = {0} and B p = B p (M ) = {0} whenever p < 0 or p > d. It is a trivial consequence of exercise 41.2.3 that B p (M ) is a subspace of the vector space of Z p (M ). Thus it makes sense to deﬁne Z p (M ) . H p = H p (M ) = p B (M ) This is the pth de Rham cohomology group of M . (Clearly this cohomology “group” is actually a vector space. It is usually called a “group” because most other cohomology theories produce only groups.) The dimension of the vector space H p (M ) is the pth Betti number of M . Another (obviously equivalent) way of phrasing the deﬁnition of the pth de Rham cohomology group is in terms of the maps ···
/ Ωp−1 (M )
dp−1

/ Ωp (M )

dp

/ Ωp+1 (M )

/ ···

where dp−1 and dp are exterior diﬀerentiation operators. Deﬁne ker dp H p (M ) := ran dp−1 for all p. 42.1.2. Definition. Let F : M → N be a smooth function between smooth manifolds. For p ≥ 1 deﬁne Ωp F : Ωp (N ) → Ωp (M ) : ω → (Ωp F )(ω) where (Ωp F )(ω) m (v 1 , . . . , v p ) = ωF (m) dFm (v 1 ), . . . , dFm (v p ) (42.1) for every m ∈ M and v 1 , . . . v p ∈ Tm . Also deﬁne (Ω0 F )(ω) m = ωF (m) . We simplify the notation in 42.1 slightly (Ωp F )(ω)(v 1 , . . . , v p ) = ω(dF (v 1 ), . . . , dF (v p )). (42.2) ∗ the map induced by the maps Ωp F which takes the Z-graded algebra Ω(N ) to the Denote by F Z-graded algebra Ω(M ). 42.1.3. Definition. Let F : M → N be a smooth function between smooth manifolds. For each integer p deﬁne H p (F ) : H p (N ) → H p (M ) : [ω] → [Ωp (F )(ω)]. Denote by H ∗ (F ) the induced map which takes the Z-graded algebra H ∗ (N ) into H ∗ (M ).
127

128

42. THE DE RHAM COHOMOLOGY GROUP

42.2. Exercises

(Due: Wed. May 6)

42.2.1. Exercise. Show that Ωp (as deﬁned in 38.1.4 and 42.1.2) is a contravariant functor from the category of smooth manifolds and smooth maps to the category of vector spaces and linear maps. 42.2.2. Exercise. If F : M → N is a smooth function between smooth manifolds, ω ∈ Ωp (N ), and µ ∈ Ωq (N ), then (Ωp+q F )(ω ∧ µ) = (Ωp F )(ω) ∧ (Ωq F )(µ). 42.2.3. Exercise. In exercise 42.2.1 you showed that Ωp was a functor for each p. What about Ω itself? Is it a functor? Explain. 42.2.4. Exercise. Let F : M → N be a smooth function between smooth manifolds. Prove that d ◦ F ∗ = F ∗◦ d . 42.2.5. Exercise. Calculate the cohomology group H 0 (R). 42.2.6. Exercise. For U ⊆ Rn give a very clear description of H 0 (U ) and explain why its dimension is the number of connected components of U . Hint. A function is said to be locally constant if it is constant in some neighborhood of each point in its domain. 42.2.7. Exercise. Let V = {0} be the 0-dimensional Euclidean space. Compute the pth de Rham cohomology group H p (V ) for all p ∈ Z. 42.2.8. Exercise. Compute H p (R) for all p ∈ Z. 42.2.9. Exercise. Let U be the union of m disjoint open intervals in R. Compute H p (U ) for all p ∈ Z. 42.2.10. Exercise. Let U ⊆ Rn . For [ω] ∈ H p (U ) and [µ] ∈ H q (U ) deﬁne [ω][µ] = [ω ∧ µ] ∈ H p+q (U ). Explain why exercise 41.2.4 is necessary for this deﬁnition to make sense. Prove also that this definition does not depend on the representatives chosen from the equivalence classes. Show that this deﬁnition makes H ∗ (U ) = H p (U ) into a Z-graded algebra. This is the de Rham cohomology p∈Z algebra of U . 42.2.11. Exercise. Prove that with deﬁnition 42.1.3 H ∗ becomes a contravariant functor from the category of open subsets of Rn and smooth maps to the category of Z-graded algebras and their homomorphisms.
◦ ◦

CHAPTER 43

Cochain Complexes
43.1. Background Topics: cochain complexes, cochain maps.

43.1.1. Definition. A sequence ···
/ Vp−1
dp−1

/ Vp

dp

/ Vp+1

/ ···

of vector spaces and linear maps is a cochain complex if dp ◦ dp−1 = 0 for all p ∈ Z. Such a sequence may be denoted by (V ∗ , d) or just by V ∗ . 43.1.2. Definition. We generalize deﬁnition 42.1.1 in the obvious fashion. If V ∗ is a cochain complex, then the pth cohomology group H p (V ∗ ) is deﬁned to be ker dp / ran dp−1 . (As before, this “group” is actually a vector space.) In this context the elements of Vp are often called pcochains, elements of ker dp are p-cocycles, elements of ran dp−1 are p-coboundaries, and d is the coboundary operator. W∗ 43.1.3. Definition. Let (V ∗ , d) and (W ∗ , δ) be cochain complexes. A cochain map G : V ∗ → is a sequence of linear maps Gp : Vp → Wp satisfying δp ◦ Gp = Gp+1 ◦ dp for every p ∈ Z. That is, the diagram ...
/ Vp 
Gp dp

/ Vp+1 
Gp+1

/ ...

... commutes. 43.1.4. Definition. A sequence 0

/ Wp

δp

/ Wp+1

/ ...

/ U∗

F

/V∗

G

/ W∗

/0

of cochain complexes and cochain maps is (short) exact if for every p ∈ Z the sequence 0
/ Up
Fp

/ Vp

Gp

/ Wp

/0

of vector spaces and linear maps is (short) exact.

129

130

43. COCHAIN COMPLEXES

43.2. Exercises

(Due: Fri. May 8)

43.2.1. Exercise. Let G : V ∗ → W ∗ be a cochain map between cochain complexes. For each p ∈ Z deﬁne G∗ : H p (V ∗ ) → H p (W ∗ ) : [v] → [Gp (v)] p whenever v is a cocycle in Vp . Show that the maps G∗ are well deﬁned and linear. Hint. To prove p that G∗ is well-deﬁned we need to show two things: that Gp (v) is a cocycle in Wp and that the p deﬁnition does not depend on the choice of representative v. 43.2.2. Exercise. If 0 complexes, then
/ U∗
F

/V∗
∗ Fp

G

/ W∗
G∗ p

/ 0 is a short exact sequence of cochain / H p (W ∗ )

H p (U ∗ ) is exact at H p (V ∗ ) for every p ∈ Z.

/ H p (V ∗ )

43.2.3. Exercise. A short exact sequence 0
ηp−1

/ U∗

F

/V∗

G

/ W∗
G∗ p

/0

of cochain complexes induces a long exact sequence
/ H p−1 (W ∗ ) / H p (U ∗ )
∗ Fp

/ H p (V ∗ )

/ H p (W ∗ )

ηp

/ H p+1 (U ∗ ) /

Hint. If w is a cocycle in Wp , then, since Gp is surjective, there exists v ∈ Vp such that w = Gp (v). It follows that dv ∈ ker Gp+1 = ran Fp+1 so that dv = Fp+1 (u) for some u ∈ Up+1 . Let ηp ([w]) = [u].

CHAPTER 44

Simplicial Homology
44.1. Background Topics: convex, convex combinations, convex hull, convex independent, closed simplex, open simplex, oriented simplex, face of a simplex, simplicial complex, r -skeleton of a complex, polyhedron of a complex, chains, boundary maps, cycles, boundaries, Betti number, Euler characteristic.

44.1.1. Definition. Let V be a vector space. Recall that a linear combination of a ﬁnite n set {x1 , . . . , xn } of vectors in V is a vector of the form k=1 αk xk where α1 , . . . , αn ∈ R. If α1 = α2 = · · · = αn = 0, then the linear combination is trivial ; if at least one αk is diﬀerent from zero, the linear combination is nontrivial. A linear combination n αk xk of the vectors k=1 x1 , . . . , xn is a convex combination if αk ≥ 0 for each k (1 ≤ k ≤ n) and if n αk = 1. k=1 44.1.2. Definition. If a and b are vectors in the vector space V , then the closed segment between a and b, denoted by [a, b], is {(1 − t)a + tb : 0 ≤ t ≤ 1}. 44.1.3. CAUTION. Notice that there is a slight conﬂict between this notation, when applied to the vector space R of real numbers, and the usual notation for closed intervals on the real line. In R the closed segment [a, b] is the same as the closed interval [a, b] provided that a ≤ b. If a > b, however, the closed segment [a, b] is the same as the segment [b, a], it contains all numbers c such that b ≤ c ≤ a, whereas the closed interval [a, b] is empty. 44.1.4. Definition. A subset C of a vector space V is convex if the closed segment [a, b] is contained in C whenever a, b ∈ C. 44.1.5. Definition. Let A be a subset of a vector space V . The convex hull of A is the smallest convex subset of V which contain A. 44.1.6. Definition. A set S = {v0 , v1 , . . . , vp } of p + 1 vectors in a vector space V is convex independent if the set {v1 − v0 , v2 − v0 , . . . , vp − v0 } is linearly independent in V . 44.1.7. Definition. An affine subspace of a vector space V is any translate of a linear subspace of V . 44.1.8. Definition. Let p ∈ Z+ . The closed convex hull of a convex independent set S = {v0 , . . . , vp } of p + 1 vectors in some vector space is a closed p -simplex. It is denoted by [s] or by [v0 . . . . , vp ]. The integer p is the dimension of the simplex. The open p -simplex determined by the set S is the set of all convex combinations p αk vk of elements of S where each αk > 0. k=0 The open simplex will be denoted by (s) or by (v0 , . . . , vp ). We make the special convention that a single vector {v} is both a closed and an open 0 -simplex. If [s] is a simplex in Rn then the plane of [s] is the aﬃne subspace of Rn having the least dimension which contains [s]. It turns out that the open simplex (s) is the interior of [s] in the plane of [s].
131

132

44. SIMPLICIAL HOMOLOGY

44.1.9. Definition. Let [s] = [v0 , . . . , vp ] be a closed p -simplex in Rn and {j0 , . . . , jq } be a nonempty subset of {0, 1, . . . , p}. Then the closed q -simplex [t] = [vj0 , . . . , vjq ] is a closed q-face of [s]. The corresponding open simplex (t) is an open q-face of [s]. The 0 -faces of a simplex are called the vertices of the simplex. Note that distinct open faces of a closed simplex [s] are disjoint and that the union of all the open faces of [s] is [s] itself. 44.1.10. Definition. Let [s] = [v0 , . . . , vp ] be a closed p -simplex in Rn . We say that two orderings (vi0 , . . . , vip ) and (vj0 , . . . , vjp ) of the vertices are equivalent if (j0 , . . . , jp ) is an even permutation of (i0 , . . . , ip ). (This is an equivalence relation.) For p ≥ 1 there are exactly two equivalence classes; these are the orientations of [s]. An oriented simplex is a simplex together with one of these orientations. The oriented simplex determined by the ordering (v0 , . . . , vp ) will be denoted by v0 , . . . , vp . If, as above, [s] is written as [v0 , . . . , vp ], then we may shorten v0 , . . . , vp to s . Of course, none of the preceding makes sense for 0 -simplexes. We arbitrarily assign them two orientations, which we denote by + and −. Thus s and − s have opposite orientations. 44.1.11. Definition. A ﬁnite collection K of open simplexes in Rn is a simplicial complex if the following conditions are satisﬁed: (1) if (s) ∈ K and (t) is an open face of [s], then (t) ∈ K; and (2) if (s), (t) ∈ K and (s) = (t), then (s) ∩ (t) = ∅. The dimension of a simplicial complex K, denoted by dim K, is the maximum dimension of the simplexes constituting K. If r ≤ dim K, then the r-skeleton of K, denoted by K r , is the set of all open simplexes in K whose dimensions are no greater than r. The polyhedron, |K|, of the complex K is the union of all the simplexes in K. 44.1.12. Definition. Let K be a simplicial complex in Rn . For 0 ≤ p ≤ dim K let Ap (K) (or just Ap ) denote the free vector space generated by the set of all oriented p -simplexes belonging to K. For 1 ≤ p ≤ dim K let Wp (K) (or just Wp ) be the subspace of Ap generated by all elements of the form v0 , v1 , v2 , . . . , vp + v1 , v0 , v2 , . . . , vp and let Cp (K) (or just Cp ) be the resulting quotient space Ap /Wp .For p = 0 let Cp = Ap and for p < 0 or p > dim K let Cp = {0}. The elements of Cp are the p -chains of K. Notice that for any p we have [ v0 , v1 , v2 , . . . , vp ] = −[ v1 , v0 , v2 , . . . , vp ] . To avoid cumbersome notation we will not distinguish between the p -chain [ v0 , v1 , v2 , . . . , vp ] and its representative v0 , v1 , v2 , . . . , vp . 44.1.13. Definition. Let s = v0 , v1 , . . . , vp+1 be an oriented (p + 1) -simplex. We deﬁne the boundary of s , denoted by ∂ s , by
p+1

∂ s =
k=0

(−1)k v0 , . . . , vk , . . . , vp+1 .

(The caret above the vk indicates that that term is missing; so the boundary of a (p + 1) -simplex is an alternating sum of p -simplexes.) 44.1.14. Definition. Let K be a simplicial complex in Rn . For 1 ≤ p ≤ dim K deﬁne ∂p = ∂ : Cp+1 (K) → Cp (K) : as follows. If a(s) s is a p -chain in K, let ∂ a(s) s = a(s) s .

44.1. BACKGROUND

133

For all other p let ∂p be the zero map. The maps ∂p are called boundary maps. Notice that each ∂p is a linear map. 44.1.15. Definition. Let K be a simplicial complex in Rn and 0 ≤ p ≤ dim K. Deﬁne Zp (K) = Zp to be the kernel of ∂p : Cp → Cp−1 and Bp (K) = Bp to be the range of ∂p+1 : Cp+1 → Cp . The members of Zp are p -cycles and the members of Bp are p -boundaries. It is clear from exercise 44.2.2 that Bp is a subspace of the vector space Zp . Thus we may deﬁne Hp (K) = Hp to be Zp /Bp . It is the pth simplicial homology group of K. (And, of course, Zp , Bp , and Hp are the trivial vector space whenever p < 0 or p > dim K.) 44.1.16. Definition. Let K be a simplicial complex. The number βp := dim Hp (K) is the pth Betti number of the complex K. And χ(K) := dim K (−1)p βp is the Euler characteristic p=0 of K.

134

44. SIMPLICIAL HOMOLOGY

44.2. Exercises

(Due: Wed. May 13)

44.2.1. Exercise. Show that deﬁnition 44.1.5 makes sense by showing that the intersection of a family of convex subsets of a vector space is itself convex. Then show that a “constructive characterization” is equivalent; that is, prove that the convex hull of A is the set of all convex combinations of elements of A. 44.2.2. Exercise. Let K be a simplicial complex in Rn . Show that ∂ 2 : Cp+1 (K) → Cp−1 (K) is identically zero. Hint. It suﬃces to prove this for generators v0 , . . . , vp+1 . 44.2.3. Exercise. Let K be the topological boundary (that is, the 1 -skeleton) of an oriented 2 -simplex in R2 . Compute Cp (K), Zp (K), Bp (K), and Hp (K) for each p. 44.2.4. Exercise. What changes in exercise 44.2.3 if K is taken to be the oriented 2 -simplex itself. 44.2.5. Exercise. Let K be the simplicial complex in R2 comprising two triangular regions similarly oriented with a side in common. For all p compute Cp (K), Zp (K), Bp (K), and Hp (K). 44.2.6. Exercise. Let K be a simplicial complex. For 0 ≤ p ≤ dim K let αp be the number of p -simplexes in K. That is, αp = dim Cp (K). Show that
dim K

χ(K) =
p=0

(−1)p αp .

CHAPTER 45

Simplicial Cohomology
45.1. Background Topics: simplicial cochain, simplicial cocycles, simplicial coboundaries, simplicial cohomology group, smooth submanifold, smoothly triangulated manifold, de Rham’s theorem, pullback of differential forms.

45.1.1. Definition. Let K be a simplicial complex. For each p ∈ Z let C p (K) = Cp (K) . The elements of C p (K) are (simplicial) p -cochains. Then the adjoint ∂p ∗ of the boundary map ∂p : Cp+1 (K) → Cp (K) is the linear map ∂p ∗ = ∂ ∗ : C p (K) → C p+1 (K) . (Notice that ∂ ∗ ◦ ∂ ∗ = 0.) Also deﬁne (1) Z p (K) := ker ∂p ∗ ; (2) B p (K) := ran ∂p−1 ∗ ; and (3) H p (K) := Z p (K)/B p (K). Elements of Z p (K) are (simplicial) p -cocycles and elements of B p (K) are (simplicial) p coboundaries. The vector space H p (K) is the pth simplicial cohomology group of K. 45.1.2. Definition. Let F : N → M be a smooth injection between smooth manifolds. The pair (N, F ) is a smooth submanifold of M if dFn is injective for every n ∈ N . 45.1.3. Definition. Let M be a smooth manifold, K be a simplicial complex in Rn , and h : [K] → M be a homeomorphism. The triple (M, K, h) is a smoothly triangulated manifold if for every open simplex (s) in K the map h : [s] → M has an extension hs : U → M to a
[s]

neighborhood U of [s] lying in the plane of [s] such that (U, hs ) is a smooth submanifold of M . 45.1.4. Theorem. A smooth manifold can be triangulated if and only if it is compact. The proof of this theorem is tedious enough that very few textbook authors choose to include it in their texts. You can ﬁnd a “simpliﬁed” proof in [6]. 45.1.5. Theorem (de Rham’s theorem). If (M, K, φ) is a smoothly triangulated manifold, then H p (M ) ∼ H p (K) = for every p ∈ Z. You can ﬁnd proofs of de Rham’s theorem in [16], chapter IV, theorem 3.1; [20], theorem 16.12; [24], pages 167–173; and [26], theorem 4.17. 45.1.6. Definition (pullbacks of diﬀerential forms). Let F : M → N be a smooth mapping between smooth manifolds. Then there exists an algebra homomorphism F ∗ : Ω(N ) → Ω(M ), called the pullback associated with F which satisﬁes the following conditions:
135

136

45. SIMPLICIAL COHOMOLOGY

(1) F ∗ maps Ωp (N ) into Ωp (M ) for each p; (2) F ∗ (g) = g ◦ F for each 0-form g on N ; and (3) (F ∗ µ)m (v) = µF (m) (dFm (v)) for every 1-form µ on N , every m ∈ M , and every v ∈ Tm .

45.2. EXERCISES

(DUE: FRI. MAY 15)

137

45.2. Exercises

(Due: Fri. May 15)

45.2.1. Exercise. Show that if K is a simplicial complex in Rn , then H p (K) ∼ Hp (K) = every integer p. 45.2.2. Exercise. Prove the assertion made in deﬁnition 45.1.6.

for

45.2.3. Exercise. Show that if F : M → N is a smooth map between n-manifolds, then F ∗ is a cochain map from the cochain complex (Ω∗ (N ), d ) to the cochain complex (Ω∗ (M ), d ). That is, show that the diagram Ωp (N )
d

/ Ωp+1 (N )

F∗

F∗

Ωp (M ) commutes for every p ∈ Z. 

d 

/ Ωp+1 (M )

CHAPTER 46

Integration of Diﬀerential Forms
46.1. Background Topics: integral of a diﬀerential form over a simplicial complex, integral of a diﬀerential form over a manifold, manifolds with boundary.

46.1.1. Definition. Let s be an oriented p -simplex in Rn (where 1 ≤ p ≤ n) and µ be a p form deﬁned on a set U which is open in the plane of s and which contains [s]. If s = v0 , . . . , vp take (v1 − v0 , . . . , vp − v0 to be an ordered basis for the plane of s and let x1 , . . . , xp be the coordinate projection functions relative to this ordered basis; that is, if a = p ak (vk − v0 ) ∈ U , k=1 then xj (a) = ak for 1 ≤ j ≤ p. Then φ = (x1 , . . . , xp ) : U → Rp is a chart on U ; so there exists a smooth function g on U such that µ = g dx1 ∧ · · · ∧ dxp . Deﬁne µ=
s [s]

g dx1 . . . dxp

where the right hand side is an ordinary Riemann integral. If v0 is a 0 -simplex, we make a special deﬁnition f = f (v0 )
v0

for every 0 -form f . Extend the preceding deﬁnition to p -chains by requiring the integral to be linear as a function of simplexes; that is, if c = as s is a p -chain (in some simplicial complex) and µ is a p -form, deﬁne µ= a(s) µ .
c s

46.1.2. Definition. For a smoothly triangulated manifold (M, K, h) we deﬁne a map : Ωp (M ) → C p (K)
p

as follows. If ω is a p -form on M , then p ω is to be a linear functional on Cp (K); that is, a member ∗ of C p (K) = Cp (K) . In order to deﬁne a linear functional on Cp (K) it suﬃces to specify its values on the basis vectors of Cp (K); that is, on the oriented p -simplexes s which constitute Cp (K). Let hs : U → M be an extension of h to an open set U in the plane of s . Then hs ∗ pulls back p -forms on M to p -forms on U so that hs ∗ (ω) ∈ Ωp (U ). Deﬁne ω
p [s]

s :=
s

hs ∗ (ω) .

139

140

46. INTEGRATION OF DIFFERENTIAL FORMS

46.1.3. Notation. Let Hn = {x ∈ Rn : xn ≥ 0}. This is the upper half-space of Rn . 46.1.4. Definition. A n -manifold with boundary is deﬁned in the same way as an n manifold except that the range of a chart is assumed to be an open subset of Hn . The interior of Hn , denoted by int Hn , is deﬁned to be {x ∈ Rn : xn > 0}. (Notice that this is the interior of Hn regarded as a subset of Rn —not of Hn .) The boundary of Hn , denoted by ∂Hn , is deﬁned to be {x ∈ Rn : xn = 0}. If M is an n -manifold with boundary, a point m ∈ M belongs to the interior of M (denoted by int M ) if φ(m) ∈ int Hn for some chart φ. And it belongs to the boundary of M (denoted by ∂M ) if φ(m) ∈ ∂Hn for some chart φ. 46.1.5. Theorem. Let M and N be a smooth n -manifolds with boundary and F : M → N be a smooth diﬀeomorphism. Then both int M and ∂M are smooth manifolds (without boundary). The interior of M has dimension n and the boundary of M has dimension n − 1. The mapping F induces smooth diﬀeomorphisms int F : int M → int N and ∂F : ∂M → ∂N . For a proof of this theorem consult the marvelous text [1], proposition 7.2.6.

46.2. EXERCISES

(DUE: MON. MAY 18)

141

46.2. Exercises

(Due: Mon. May 18)

46.2.1. Exercise. Let V be an open subset of Rn , F : V → R, and c : [t0 , t1 ] → V be a smooth curve in V . Let C = ran c. It is conventional to deﬁne the “integral of the tangential component of F over C ”, often denoted by C FT , by the formula
t1 t1

FT =
C t0

F ◦ c, Dc =
t0

F (c(t)), c (t) dt

(46.1)

The “tangential component of F ”, written FT may be regarded as the 1 -form n F k dxk . k=1 Make sense of the preceding deﬁnition in terms of the deﬁnition of the integral of 1 -forms over a smoothly triangulated manifold. For simplicity take n = 2. Hint. Suppose we have the following: (1) t0 , t1 (with t0 < t1 ) is an oriented 1 -simplex in R; (2) V is an open subset of R2 ; (3) c : J → V is an injective smooth curve in V , where J is an open interval containing [t0 , t1 ]; and (4) ω = a dx + b dy is a smooth 1 -form on V . First show that c∗ (dx) (t) = Dc1 (t) for t0 ≤ t ≤ t1 . (We drop the notational distinction between c and its extension cs to J. Since the tangent space Tt is one-dimensional for every t, we identify Tt with R. Choose v (in (3) of deﬁnition 45.1.6) to be the usual basis vector in R, the number 1.) Show in a similar fashion that c∗ (dy) (t) = Dc2 (t) . Then write an expression for c∗ (ω) (t). Finally conclude that t1 t0 (a, b) ◦ c, Dc as claimed in 46.1.

( t0 , t1 ) is indeed equal to

46.2.2. Exercise. Let S1 be the unit circle in R2 oriented counterclockwise and let F be the vector ﬁeld deﬁned by F(x, y) = (2x3 − y 3 ) i + (x3 + y 3 ) j. Use your work in exercise 46.2.1 to calculate S1 FT . Hint. You may use without proof two facts: (1) the integral does not depend on the parametrization (triangulation) of the curve, and (2) the results of exercise 46.2.1 hold also for simple closed curves in R2 ; that is, for curves c : [t0 , t1 ] → R2 which are injective on the open interval (t0 , t1 ) but which satisfy c(t0 ) = c(t1 ). 46.2.3. Exercise. Let V be an open subset of R3 , F : V → R3 be a smooth vector ﬁeld, and (S, K, h) be a smoothly triangulated 2 -manifold such that S ⊆ V . It is conventional to deﬁne the “normal component of F over S ”, often denoted by S FN , by the formula FN =
S K

F ◦ h, n

where n = h1 × h2 . (Notation: hk is the partial derivative of h.) Make sense of the preceding deﬁnition in terms of the deﬁnition of the integral of 2 -forms over a smoothly triangulated manifold (with or without boundary). In particular, suppose that F = a i+b j+c k (where a, b, and c are smooth functions) and let ω = a dy ∧dz +b dz ∧dx+c dx∧dy. This 2 -form is conventionally called the “normal component of F” and is denoted by FN . Notice that FN is just ∗µ where µ is the 1 -form associated with the vector ﬁeld F. Hint. Proceed as follows. (a) Show that the vector n(u, v) is perpendicular to the surface S at h(u, v) for each (u, v) in [K] by showing that it is perpendicular to D(h ◦ c)(0) whenever c is a smooth curve in [K] such that c(0) = (u, v).

k th

142

46. INTEGRATION OF DIFFERENTIAL FORMS

(b) Let u and v (in that order) be the coordinates in the plane of [K] and x, y, and z (in that order) be the coordinates in R3 . Show that h∗ (dx) = h1 du + h1 dv. Also compute h∗ (dy) 1 2 and h∗ (dz). Remark. If at each point in [K] we identify the tangent plane to R2 with R2 itself and if we use conventional notation, the “v” which appears in (3) of deﬁnition 45.1.6 is just not written. One keeps in mind that the components of h and all the diﬀerential forms are functions on (a neighborhood of) [K]. (c) Now ﬁnd h∗ (ω). (Recall that ω = FN is deﬁned above.) (d) Show for each simplex (s) in K that ω ( s )=
2 [s] n k=1

F ◦ h, n . sk ,

(e) Finally show that if s1 , . . . , sn are the oriented 2 -simplexes of K and c = then ω (c) =
2 [K]

F ◦ h, n .

46.2.4. Exercise. Let F(x, y, z) = xz i + yz j and H be the hemisphere of x2 + y 2 + z 2 = 4 for FN . which z ≥ 0. Use exercise 46.2.3 to ﬁnd
H

CHAPTER 47

Stokes’ Theorem
47.1. Background Topics: Stokes’ theorem.

47.1.1. Theorem (Stokes’ theorem). Suppose that (M, K, h) is an oriented smoothly triangulated manifold with boundary. Then the integration operator = p p∈Z is a cochain map from the cochain complex (Ω∗ (M ), d ) to the cochain complex (C ∗ (K), ∂ ∗ ). This is an important and standard theorem, which appears in many versions and with many diﬀerent proofs. See, for example, [1], theorem 7.2.6; [19], chapter XVII, theorem 2.1; [21], theorem 10.8; or [26], theorems 4.7 and 4.9. Recall that when we say in Stokes’ theorem that the integration operator is a cochain map, we are saying that the following diagram commutes.
d

/ Ωp (M )
R

d

/ Ωp+1 (M )
R

d

/

∂∗ 

/ C p (K)

∂∗ 

/ C p+1 (K)

∂∗

/

Thus if ω is a p -form on M and s is an oriented (p + 1) -simplex belonging to K, then we must have dω ( s ) =
p+1

∂∗
p

ω

( s ).

(47.1)

This last equation (47.1) can be written in terms of integration over oriented simplexes: d hs ∗ ω =
s ∂ s

hs ∗ ω .

(47.2)

In more conventional notation all mention of the triangulating simplicial complex K and of the map h is suppressed. This is justiﬁed by the fact that it can be shown that the value of the integral is independent of the particular triangulation used. Then when the equations of the form (47.2) are added over all the (p + 1) -simplexes comprising K we arrive at a particularly simple formulation of (the conclusion of) Stokes’ theorem dω =
M ∂M

ω.

(47.3)

One particularly important topic that has been glossed over in the preceding is a discussion of orientable manifolds (those which possess nowhere vanishing volume forms), their orientations, and the manner in which an orientation of a manifold with boundary induces an orientation on its
143

144

47. STOKES’ THEOREM

boundary. One of many places where you can ﬁnd a careful development of this material is in sections 6.5 and 7.2 of [1]. 47.1.2. Theorem. Let ω be a 1 -form on a connected open subset U of R2 . Then ω is exact on U if and only if C ω = 0 for every simple closed curve in U . For a proof of this result see [10], chapter 2, proposition 1.

47.2. EXERCISES

(DUE: WED. MAY 20)

145

47.2. Exercises 47.2.1. Exercise. Let ω = − 1-form ω is closed but not exact.

(Due: Wed. May 20)

y dx x dy + 2 . Show that on the region R2 \ {(0, 0)} the x2 + y 2 x + y2

47.2.2. Exercise. What classical theorem do we get (for smooth functions) from the version of Stokes’ theorem given by equation (47.3) in the special case that ω is a 0 -form and M is a 1- manifold (with boundary) in R. 47.2.3. Exercise. What classical theorem do we get (for smooth functions) from the version of Stokes’ theorem given by equation (47.3) in the special case that ω is a 0 -form and M is a 1- manifold (with boundary) in R3 . 47.2.4. Exercise. What classical theorem do we get (for smooth functions) from the version of Stokes’ theorem given by equation (47.3) in the special case that ω is a 1 -form and M is a 2- manifold (with boundary) in R2 . 47.2.5. Exercise. Use exercise 47.2.4to compute the unit circle oriented counterclockwise).
3 S1 (2x

− y 3 ) dx + (x3 + y 3 ) dy (where S1 is

47.2.6. Exercise. What classical theorem do we get (for smooth functions) from the version of Stokes’ theorem given by equation (47.3) in the special case that ω is a 1 -form and M is a 2- manifold (with boundary) in R3 . 47.2.7. Exercise. What classical theorem do we get (for smooth functions) from the version of Stokes’ theorem given by equation (47.3) in the special case that ω is a 2 -form and M is a 3- manifold (with boundary) in R3 . 47.2.8. Exercise. Your good friend Fred R. Dimm calls you on his cell phone seeking help with a math problem. He says that he wants to evaluate the integral of the normal component of the vector ﬁeld on R3 whose coordinate functions are x, y, and z (in that order) over the surface of a cube whose edges have length 4. Fred is concerned that he’s not sure of the coordinates of the vertices of the cube. How would you explain to Fred (over the phone) that it doesn’t matter where the cube is located and that it is entirely obvious that the value of the surface integral he is interested in is 192?

CHAPTER 48

48.1.1. Definition. Let V be a ﬁnite dimensional real vector space. A function Q : V → R is a quadratic form if (i) Q(v) = Q(−v) for all v, and (ii) the map B : V × V → R : (u, v) → Q(u + v) − Q(u) − Q(v) is a bilinear form. In this case B is the bilinear form associated with the quadratic form Q. It is obviously sym1 metric. Note: In many texts B(u, v) is deﬁned to be 2 [Q(u + v) − Q(u) − Q(v)].

147

148

48.2. Exercises

(Due: Fri. May 22)

48.2.1. Exercise. Let B be a symmetric bilinear form on a real ﬁnite dimensional vector space V . Deﬁne Q : V → R by Q(v) = B(v, v). Show that Q is a quadratic form on V . 48.2.2. Exercise. Let Q be a quadratic form on a real ﬁnite dimensional vector space V . Prove that Q(u + v + w) − Q(u + v) − Q(u + w) − Q(v + w) + Q(u) + Q(v) + Q(w) = 0 for all u, v, w ∈ V . 48.2.3. Exercise. Let Q be a quadratic form on a real ﬁnite dimensional vector space V . Prove that Q(αv) = α2 Q(v) for all α ∈ R and v ∈ V . Hint. First use exercise 48.2.2 to show that Q(2v) = 4Q(v) for every v ∈ V .

CHAPTER 49

Deﬁnition of Cliﬀord Algebra
49.1. Background Topics: an algebra universal over a vector space, Cliﬀord map; Cliﬀord algebra.

49.1.1. Definition. Let V be a vector space. A pair (U, ι), where U is a unital algebra and ι : V → U is a linear map, is universal over V if for every unital algebra A and every linear map f : V → A there exists a unique unital algebra homomorphism f : U → A such that f ◦ ι = f . 49.1.2. Definition. Let V be a real ﬁnite dimensional vector space with a quadratic form Q and A be a real unital algebra. A map f : V → A is a Clifford map if (i) f is linear, and 2 (ii) f (v) = Q(v)1A for every v ∈ V . 49.1.3. Definition. Let V be a real ﬁnite dimensional vector space with a quadratic form Q. The Clifford algebra over V is a real unital algebra Cl(V, Q), together with a Cliﬀord map j : V → Cl(V, Q), which satisﬁes the following universal condition: for every real unital algebra A and every Cliﬀord map f : V → A, there exists a unique unital algebra homomorphism f : Cl(V, Q) → A such that f ◦ j = f .

149

150

49. DEFINITION OF CLIFFORD ALGEBRA

49.2. Exercises

(Due: Wed. May 27)

49.2.1. Exercise. Let V be a vector space. Prove that if U and U are unital algebras universal over V , then they are isomorphic. 49.2.2. Exercise. Let V be a vector space. Prove that there exists a unital algebra which is universal over V . 49.2.3. Exercise. Show that condition (ii) in deﬁnition 49.1.2 is equivalent to (ii ) f (u)f (v) + f (v)f (u) = B(u, v)1A for all u, v ∈ V , where B is the bilinear form associated with Q. 49.2.4. Exercise. Let V be a real ﬁnite dimensional vector space with a quadratic form Q. Prove that if the Cliﬀord algebra Cl(V, Q) exists, then it is unique up to isomorphism. 49.2.5. Exercise. Let V be a real ﬁnite dimensional vector space with a quadratic form Q. Prove that the Cliﬀord algebra Cl(V, Q) exists. Hint. Recall the deﬁnition of the tensor algebra T (V ) in 36.1.2. Try T (V )/J where J is the ideal in T (V ) generated by elements of the form v ⊗ v − Q(v)1T (V ) where v ∈ V .

CHAPTER 50

Orthogonality with Respect to Bilinear Forms
50.1. Background Topics: orthogonality with respect to bilinear forms, the kernel of a bilinear form, nondegenerate bilinear forms.

50.1.1. Definition. Let B be a symmetric bilinear form on a real vector space V . Vectors v and w in V are orthogonal, in which case we write v ⊥ w, if B(v, w) = 0. The kernel of B is the set of all k ∈ V such that k ⊥ v for every v ∈ V . The bilinear form is nondegenerate if its kernel is {0}. 50.1.2. Definition. Let V be a ﬁnite dimensional real vector space and B be a symmetric bilinear form on V . An ordered basis E = (e1 , . . . , en ) for V is B-orthonormal if (a) B(ei , ej ) = 0 whenever i = j and (b) for each i ∈ Nn the number B(ei , ei ) is −1 or +1 or 0. 50.1.3. Theorem. If V is a ﬁnite dimensional real vector space and B is a symmetric bilinear form on V , then V has a B-orthonormal basis. Proof. See [5], chapter 1, theorem 7.6. 50.1.4. Convention. Let V be a ﬁnite dimensional real vector space, let B be a symmetric bilinear form on V , and let Q be the quadratic form associated with B. Let us agree that whenever E = (e1 , . . . , en ) is an ordered B-orthonormal basis for V , we order the basis elements in such a way that for some positive integers p and q  if 1 ≤ i ≤ p;  1, −1, if p + 1 ≤ i ≤ p + q; Q(ei ) =  0, if p + q + 1 ≤ i ≤ n. 50.1.5. Theorem. Let V be a ﬁnite dimensional real vector space, let B be a symmetric bilinear form on V , and let Q be the quadratic form associated with B. Then there exist p, q ∈ Z+ such that if E = (e1 , . . . , en ) is a B-orthonormal basis for V and v = vk ek , then
p p+q

Q(v) =
k=1

vk −
k=p+1

2

vk 2 .

Proof. See [5], chapter 1, theorem 7.11.

151

152

50. ORTHOGONALITY WITH RESPECT TO BILINEAR FORMS

50.2. Exercises

(Due: Fri. May 29)

50.2.1. Exercise. One often sees the claim that “the classiﬁcation of Cliﬀord algebras amounts to classifying vector spaces with quadratic forms”. Explain precisely what is meant by this assertion. 50.2.2. Exercise. Let B be a symmetric bilinear form on a real ﬁnite dimensional vector space V . Suppose that V is an orthogonal direct sum V1 ⊕ · · · ⊕ Vn of subspaces. Then B is nondegenerate if and only if the restriction of B to Vk is nondegenerate for each k. In fact, if Vk is the kernel of the restriction of B to Vk , then the kernel of B is the orthogonal direct sum V1 ⊕ · · · ⊕ Vn . 50.2.3. Exercise. Let B be a nondegenerate symmetric bilinear form on a real ﬁnite dimensional vector space V . If W is a subspace of V , then the restriction of B to W is nondegenerate if and only if its restriction to W ⊥ is nondegenerate. Equivalently, B is nondegenerate on W if and only if V = W + W ⊥ . 50.2.4. Exercise. Let Q be a quadratic form on a real ﬁnite dimensional vector space V and let {e1 , . . . , en } be a basis for V which is orthogonal with respect to the bilinear form B associated with Q. Show that if Q(ek ) is nonzero for 1 ≤ k ≤ p and Q(ek ) = 0 for p < k ≤ n, then the kernel of B is the span of {ep+1 , . . . , en }. 50.2.5. Exercise. Let Q be a quadratic form on a real ﬁnite dimensional vector space V . Show that if dim V = n, then dim Cl(V, Q) = 2n . 50.2.6. Exercise. Let V be a real ﬁnite dimensional vector space and let Q be the quadratic form which is identically zero on V . Identify Cl(V, Q).

CHAPTER 51

Examples of Cliﬀord Algebras
51.1. Background Topics: No additional topics. 51.1.1. Notation. If (V, Q) is a ﬁnite dimensional real vector space V with a nondegenerate quadratic form Q, we often denote the Cliﬀord algebra C(V, Q) by C(p, q) where p and q are as in theorem 50.1.5.

153

154

51. EXAMPLES OF CLIFFORD ALGEBRAS

51.2. Exercises

(Due: Mon. June 1)

51.2.1. Exercise. Let f : (V, Q) → (W, R) be a linear map between ﬁnite dimensional real vector spaces with quadratic forms. If R(f (v)) = Q(v) for every v ∈ V we say that f is an isometry. (a) Show that if f is such a linear isometry, then there exists a unique unital algebra homomorphism Cl(f ) : Cl(V, Q) → Cl(W, R) such that Cl(f )(v) = f (v) for every v ∈ V . (b) Show that Cl is a covariant functor from the category of vector spaces with quadratic forms and linear isometries to the category of Cliﬀord algebras and unital algebra homomorphisms. (c) Show that if the linear isometry f is an isomorphism, then Cl(f ) is an algebra isomorphism. 51.2.2. Exercise. Let V be a real ﬁnite dimensional vector space, Q be a quadratic form on V , and A = Cl(V, Q) be the associated Cliﬀord algebra. (a) Show that the map f : V → V : v → −v is a linear isometry. (b) Let ω = Cl(f ). Show that ω 2 = id. (c) Let A0 = {a ∈ A : ω(a) = a} and A1 = {a ∈ A : ω(a) = −a}. Show that A = A0 ⊕ A1 . 1 Hint. If a ∈ A, let a0 = 2 (a + ω(a)). (d) Show that Ai Aj ⊆ Ai+j where i, j ∈ {0, 1} and i + j is addition modulo 2. This says that a Cliﬀord algebra is a Z2 -graded (or Z/2Z -graded) algebra. 51.2.3. Exercise. Let V = R and Q(v) = v 2 for every v ∈ V . Show that Cliﬀord algebra Cl(1, 0) associated with (R, Q) is isomorphic to R ⊕ R. Hint. Consider the map u1 + ve → (u − v, u + v). 51.2.4. Exercise. Let V = R and Q(v) = −v 2 for every v ∈ V . The Cliﬀord algebra Cl(V, Q) is often denoted by Cl(0, 1). (a) Show that Cl(0, 1) ∼ C. = (b) Show that Cl(0, 1) can be represented as a subalgebra of M2 (R). Hint. AOLT, page 50, Example 2.12.
2 2 51.2.5. Exercise. Let V = R2 and Q(v) = v1 + v2 for every v ∈ V . The Cliﬀord algebra 1 0 Cl(V, Q) is often denoted by Cl(2, 0). Show that Cl(2, 0) ∼ M2 (R). Hint. Let ε1 := , = 0 −1 0 1 ε2 := , and ε12 := ε1 ε2 . 1 0 2 2 51.2.6. Exercise. Let V = R2 and Q(v) = −v1 − v2 for every v ∈ V . The Cliﬀord algebra Cl(V, Q) is often denoted by Cl(0, 2). (a) Show that Cl(0, 2) ∼ H. = (b) Show that Cl(0, 2) can be represented as a subalgebra of M4 (R). Hint. AOLT, pages 50–51, Example 2.13.

51.2.7. Exercise. Take a look at the web page written by Pertti Lounesto: http://users.tkk.ﬁ/ ppuska/mirror/Lounesto/counterexamples.htm 51.2.8. Exercise. Show that the Cliﬀord algebra Cl(3, 1) (Minkowski space-time algebra) is isomorphic to M4 (R). Hint. Exercise 51.2.7.

Bibliography
1. Ralph Abraham, Jerrold E. Marsden, and Tudor Ratiu, Manifolds, Tensor Analysis, and Applications, AddisonWesley, Reading, MA, 1983. 140, 143, 144 2. Robert A. Beezer, A First Course in Linear Algebra, 2004, http://linear.ups.edu/download/fcla-electric-2.00.pdf. vii 3. Richard L. Bishop and Richard J. Crittenden, Geometry of Manifolds, Academic Press, New York, 1964. 111, 115, 116, 119 4. Przemyslaw Bogacki, Linear Algebra Toolkit, 2005, http://www.math.odu.edu/∼bogacki/lat. vii 5. William C. Brown, A second Course in Linear Algebra, John Wiley, New York, 1988. vii, 151 6. Stewart S. Cairns, A simple triangulation method for smooth manifolds, Bull. Amer. Math. Soc. 67 (1961), 389–390. 135 7. P. M. Cohn, Basic Algebra: Groups, Rings and Fields, Springer, London, 2003. 7 8. Charles W. Curtis, Linear Algebra: An Introductory Approach, Springer, New York, 1984. vii 9. Paul Dawkins, Linear Algebra, 2007, http://tutorial.math.lamar.edu/pdf/LinAlg/LinAlg Complete.pdf. vii 10. Manfredo P. do Carmo, Diﬀerential Forms and Applications, Springer-Verlag, Berlin, 1994. 144 11. John M. Erdman, A Companion to Real Analysis, 2007, http://www.mth.pdx.edu/∼erdman/CTRA/CRAlicensepage.html. 33 12. , Operator Algebras: K-Theory of C ∗ -Algebras in Context, 2008, http://www.mth.pdx.edu/∼erdman/614/operator algebras pdf.pdf. 69, 77 13. Paul R. Halmos, Finite-Dimensional Vector Spaces, D. Van Nostrand, Princeton, 1958. vii 14. Jim Heﬀeron, Linear Algebra, 2006, http://joshua.smcvt.edu/linearalgebra. vii 15. Kenneth Hoﬀman and Ray Kunze, Linear Algebra, second ed., Prentice Hall, Englewood Cliﬀs,N.J., 1971. vii 16. S. T. Hu, Diﬀerentiable Manifolds, Holt, Rinehart, and Winston, New York, 1969. 135 17. Thomas W. Hungerford, Algebra, Springer-Verlag, New York, 1974. 7 18. Saunders Mac Lane and Garrett Birkhoﬀ, Algebra, Macmillan, New York, 1967. 7 19. Serge Lang, Fundamentals of Diﬀerential Geometry, Springer Verlag, New York, 1999. 143 20. John M. Lee, Introduction to Smooth Manifolds, Springer, New York, 2003. 119, 135 21. Ib Madsen and Jørgen Tornehave, From Calculus to Cohomology: de Rham cohomology and characteristic classes, Cambridge University Press, Cambridge, 1997. 143 22. Theodore W. Palmer, Banach Algebras and the General Theory of ∗ -Algebras I–II, Cambridge University Press, Cambridge, 1994/2001. 7 23. Steven Roman, Advanced Linear Algebra, second ed., Springer-Verlag, New York, 2005. vii, 101 24. I. M. Singer and J. A. Thorpe, Lecture Notes on Elementary Topology and Geometry, Springer Verlag, New York, 1967. 135 25. Gerard Walschap, Metric Structures in Diﬀerential Geometry, Springer-Verlag, New York, 2004. 116, 119 26. Frank W. Warner, Foundations of Diﬀerentiable Manifolds and Lie Groups, Springer Verlag, New York, 1983. 135, 143 27. Eric W. Weisstein, MathWorld, A Wolfram Web Resource, http://mathworld.wolfram.com. vii 28. Wikipedia, Wikipedia, The Free Encyclopedia, http://en.wikipedia.org. vii

155

Index

α−1 (inverse of a morphism α), 33  (V ) (Grassmann or exterior algebra), 107 k (V ) (homogeneous elements of degree k), 110 ∂ s (boundary of a simplex), 132 ∂p (boundary map), 133 ∗ -algebra (algebra with involution), 81 ∗ -homomorphism (star homomorphism), 81 x, v (alternative notation for v ∗ (x)), 28 (f, g) (function into a product), 37 M ⊕ N (direct sum), 37 S ⊗ T (tensor products of linear maps), 105 U ⊗ V (tensor product), 101 V /M (quotient of V by M ), 41 ω ∧ µ (wedge product), 107 m ⊕ n (direct sum of vectors), 37 v∞ V (direct sum), 111 vk=0 k n k=1 Mk (direct sum), 37 A + B (sum of sets of vectors), 9 f × g (product function), 3 U V (U is a subspace of V ), 9 [a, b] (closed segment in a vector space), 131 [x] (equivalence class containing x), 41 |K| (polyhedron of a complex), 132 A[x] (algebra of polynomials), 59 ¢ £ A [x] (algebra of formal power series), 59 F[x] (polynomial algebra with coeﬃcients in F), 59 1 (standard one-element set), 4 1V , idV , IV (the identity map on V ), 21 T ← (U ) (inverse image of U under T ), 21 T → (U ) (direct image of U under T ), 21 (s) = (v0 , . . . , vp ) (open p -simplex), 131 [s] = [v0 , . . . , vp ] (closed p -simplex), 131 s = v0 , . . . , vp (closed p -simplex), 132 x (= Γ(x)), 32  V ∼ W (isomorphism of vector spaces), 21 = F⊥ (pre-annihilator of F ), 31 xs (alternative notation for x(s)), 27 M ⊥ (annihilator of M ), 31 T ∗ (adjoint of T ), 81 T ∗ (vector space adjoint of T ), 34 T t (transpose of T ), 81 f ← (B) (preimage of B under f ), 34 f → (A) (image of A under f ), 34 v ∗ , 28 (V ∗ , d), V ∗ (cochain complex), 129 ∂M (boundary of a manifold), 140 ∂Hn (boundary of a half-space), 140
157

int M (interior of a manifold with boundary), 140 int Hn (interior of a half-space), 140 (the forgetful functor), 33 Abelian group, ix adjoint, 81 vector space, 34 aﬃne subspace, 131 algebra Z+ -graded, 109 Cliﬀord, 149 de Rham cohomology, 128 exterior, 107 Grassmann, 107 quotient, 52 tensor, 111 alternating, 94 Altk (V ) (set of alternating k-linear maps), 111 annihilating polynomial, 60 annihilator, 8, 31 antisymmetric, 17 axiom of choice, 18 basis, 17 dual, 29 ordered, 113 right-handed, 113 βp (Betti number), 133 Betti number, 127, 133 bilinear, 94 form associated with a quadratic form, 147 nondegenerate, 151 B-orthonormal, 151 bound lower, 17 upper, 17 boundaries, 133 boundary maps, 133 of a half-space, 140 of a manifold, 140 of a simplex, 132 bounded real valued function, 9 B p (M ) (space of de Rham p˙coboundaries), 127 B p (K) (space of simplicial p -coboundaries), 135 Bp (K) (space of simplicial p -boundaries), 133

158

INDEX

calculus polynomial functional, 60 cancellation property, 7 category, 33 concrete, 33 chain, 18 chains, 132 characteristic Euler, 133 polynomial, 71 χ(K) (Euler characteristic), 133 choice axiom of, 18 ∞ Cm (smooth functions at m), 115 Cliﬀord algebra, 149 map, 149 closed diﬀerential form, 125 face of a simplex, 132 segment, 131 simplex, 131 Cl(V, Q) (Cliﬀord algebra), 149 coboundary p-, 129 de Rham, 127 operator, 129 simplicial, 135 cochain, 135 p-, 129 complex, 129 map, 129 cocycle p-, 129 de Rham, 127 simplicial, 135 codimension, 46 coeﬃcient leading, 60 cohomology de Rham, 128 group, 129 de Rham, 127 simplicial, 135 combination convex, 131 formal linear, 100 commutative diagram, 3 ring, ix comparable, 18 complementary subspace, 37 complete order, 34 complex cochain, 129 simplicial, 132 dimension of a, 132 components, 37 composition

of morphisms, 33 concrete category, 33 constant locally, 128 polynomial, 60 continuous diﬀerentiability, 22 continuously diﬀerentiable, 44 contravariant, 33 conventions all categories are concrete, 33 all diﬀerential forms are smooth, 116 all manifolds are smooth and oriented, 115 all vector ﬁeld are smooth, 115 all vector spaces are real, ﬁnite dimensional, and oriented, 115 on Cartesian products, 3 on ordering B-orthonormal bases, 151 on the standard basis for P([a, b]), 25 write f ω for f ∧ ω when f is a 0-form, 121 convex combination, 131 hull, 131 independent, 131 set, 131 convolution, 59 coordinate projections, 38 coproduct in a category, 41 uniqueness of, 43 cotangent space, 115 covariant, 33 Cp (K) (p -chains), 132 C p (K) (p -cochains), 135 cycle, 93 length of a, 94 cycles, 133 d (exterior derivative), 119 de Rham coboundary, 127 cocycle, 127 cohomology group, 127 de Rham cohomology algebra, 128 decomposable element of a Grassmann algebra, 109 tensor, 102 degree of a decomposable element, 109 of a homogeneous element, 109 of a polynomial, 59 δ (diagonal mapping), 4 ∆fa , 93 derivative exterior, 119 partial, 116 total, 93 determinant, 98 function, 97

INDEX

159

dfa (diﬀerential of f at a), 93 diagonal mapping, 4 matrix, 69 diag(α1 , . . . , αn ) (diagonal matrix), 69 diagonalizable, 69 plus nilpotent decomposition, 73 unitarily, 89 diagram commutative, 3 diﬀerentiable, 93 continuously, 22, 44 diﬀerential, 93 p-form, 115 smooth, 115 form closed, 125 exact, 125 of a function on a manifold, 116 diﬀerential form associated with a vector ﬁeld, 121 dimension of a simplex, 131 of a simplicial complex, 132 of a vector space, 19 dim K (dimension of a simplicial complex K), 132 dim V (dimension of a vector space V ), 19 direct image, 21 sum as a product, 41 external, 37, 111 internal, 37 projections associated with, 49 disjoint permutations, 94 divides, 64 division ring, ix divisor greatest common, 65 of zero, 7 dual basis, 29 duality functor for vector spaces, 35 Dxk (m) (partial derivative on a manifold), 116 eigenspaces generalized, 73 elementary tensor, 102 EN M (projection along N onto M ), 49 εS , ε (function from S into 1), 4 equivalent orderings of vertices, 132 unitarily, 89 essential uniqueness, 43 Euler characteristic, 133 even permutation, 94

even function, 40 exact diﬀerential form, 125 exact sequence of cochain complexes, 129 of vector spaces, 42 short, 42 exterior algebra, 107 derivative, 119 product, 107 external direct sum, 37, 111 F (S) (real valued functions on S), 4 F (S, T ) (functions from S to T ), 4 face, 132 factorization, 64 ﬁeld, ix vector, 115, 121 ﬁnal projection, 89 ﬁnite support, 27 ﬁrst isomorphism theorem, 43 forgetful functor, 34 form closed , 125 diﬀerential p-, 115 exact, 125 multilinear, 94 quadratic, 147 formal linear combination, 100 formal power series, 59 free object, 99 vector space, 99 function bounded, 9 diagonal, 4 integrable, 9 interchange, 4 product, 3 switching, 4 functional calculus polynomial, 60 multilinear, 94 functor contravariant, 33 covariant, 33 forgetful, 34 power set, 35 vector space duality, 35 fundamental quotient theorem, 43 Γ (the map from V to V ∗∗ taking x to x), 32  generalized eigenspaces, 73 graded algebra, 109 Grassmann algebra, 107 greatest common divisor, 65 group, ix

160

INDEX

Abelian, ix cohomology, 129 de Rham cohomology, 127 simplicial cohomology, 135 simplicial homology, 133 symmetric, 94 H(A) (self-adjoint elements of A), 81 half-space boundary of a, 140 interior of a, 140 half-space, upper, 140 Hamel basis, 17 Hermitian, 81 Hn (upper half-space), 140 Hodge star operator, 114 Hom(G, H), Hom(G) (group homomorphisms), 1 homogeneous elements of a graded algebra, 109 tensor, 102 homology, 133 homomorphism of Abelian groups, 1 ring, 7 unital ring, 7 H p (M ) (the pth de Rham cohomology group), 127 H p (K) (the pth simplicial cohomology group), 135 Hp (K) (the pth simplicial homology group), 133 hull, convex, 131 idempotent, 49, 56 idS (identity function on S), 3 identity function on a set, 3 resolution of the, 69 image, 21, 34 direct, 21 inverse, 21 independent convex, 131 indeterminant, 59 initial projection, 89 inner product space spectral theorem for, 89 integrable real valued function, 9 integral of a p -form over a p -chain, 139 interchange operation, 4 interior of a half-space, 140 of a manifold with boundary, 140 internal direct sum, 37 interpolation Lagrange, 64 inverse image, 21 left, 33 of a morphism, 33

right, 33 invertible, 33 element of an algebra, 52 left, 21 linear map, 21 right, 21 involution, 81 irreducible, 63 isometry, 154 partial, 89 isomorphism, 21 in a category, 33 kernel, 21 of a bilinear form, 151 ker T (the kernel of T ), 21 l(S, F), l(S) (functions on S), 27 l2 (R), l2 (C) (square summable sequences), 78 Lagrange interpolation formula, 64 largest, 17 lc (S, F), lc (S) (functions on S with ﬁnite support), 27 lc (S, A) (functions from S into A with ﬁnite support), 59 leading coeﬃcient, 60 left -handed basis, 113 inverse of a morphism, 33 invertible linear map, 21 length of a cycle, 94 of a vector, 77 linear, 21 combination formal, 100 ordering, 18 transformation invertible, 21 transformations, 21 tensor products of, 105 L(V, W ), L(V ) (family of linear functions), 21 Ln (V1 , . . . , Vn ) (family of n-linear functions), 94 little-oh functions, 93 locally constant, 128 lower bound, 17 l(S, A) (functions from S into A), 59 manifold boundary of a, 140 smoothly triangulated, 135 with boundary, 140 interior of a, 140 matrix diagonal, 69 similarity, 69 symmetric, 19 maximal, 17 minimal, 17 polynomial, 60 existence of, 61

INDEX

161

Mn (A) (n × n matrices of members of A), 97 monic polynomial, 60 monoid, ix morphism map, 33 Mor(S, T ) (morphisms from S to T ), 33 morphisms, 33 composition of, 33 mT (minimal polynomial for T ), 60 multilinear, 94 form, 94 functional, 94 N (A) (normal elements of A), 81 n-linear function, 94 nondegenerate bilinear form, 151 norm, 77 normal, 81 nullity, 21 nullspace, 21 o (the family of little-oh functions, 93 object map, 33 objects, 33 odd permutation, 94 odd function, 40 Ωp (U ) (space of p-forms on U ), 115 Ωp F , 127 Ω(U ) (the algebra of diﬀerential forms on U ), 116 open face of a simplex, 132 simplex, 131 operator, 21 diagonalizable, 69 projection, 49 similarity, 69 unilateral shift, 56 opposite orientation, 113 order complete, 34 ordered basis, 113 by inclusion, 17 linearly, 18 partially, 17 orientation of a simplex, 132 of a vector space, 113 oriented simplex, 132 vector space, 113 orthogonal direct sum projections associated with, 89 orthogonal projections, 85 projection, 85 assocuated with an orthogonal direct sum decomposition, 89 resolution of the identity, 89

with respect to a bilinear form, 151 orthonormal, B-, 151 partial derivative, 116 isometry, 89 ordering, 17 p -boundaries, 133 p -chains, 132 p -coboundary, 129 simplicial, 135 p -cochain, 129, 135 p -cocycle, 129 simplicial, 135 p -cycles, 133 permutation, 93 cyclic, 93 even, 94 odd, 94 sign of a, 94 p-form, 115 plane of a simplex, 131 Pn ([a, b]) (polynomials of degree strictly less than n), 25 polarization identity, 77 polyhedron, 132 polynomial, 59 annihilating, 60 characteristic, 71 constant, 60 degree of a, 59 function, 60 functional calculus, 60 irreducible, 63 minimal, 60 existence of, 61 monic, 60 prime, 63 reducible, 63 power series formal, 59 power set, 34 functor, 35 (p, q)-shuﬄe, 111 pre-annihilator, 31 preimage, 34 primary decomposition theorem, 73 prime polynomial, 63 relatively, 65 product exterior, 107 in a category, 41 tensor, 101 uniqueness of, 43 wedge, 107 projection along one subspace onto another, 49 assocuated with direct sum decomposition, 49 ﬁnal, 89

162

INDEX

in a ∗ -algebra, 85 initial, 89 operator, 49 orthogonal, 85 projections coordinate, 38 P(S) (power set of S), 34 p -simplex, 131 pullback of diﬀerential forms, 135 quadratic form, 147 associated with a bilinear form, 147 quotient algebra, 52 map, 41, 45, 46 object, 45 space, 41 R∞ (sequences of real numbers), 11 range, 21 ran T (the range of T ), 21 rank, 21 rank-plus-nullity theorem, 47 reducible, 63 reducing subspaces, 67 reﬂexive, 17 relation, 17 relatively prime, 65 resolution of the identity in vector spaces, 69 orthogonal, 89 retraction, 21 reverse orientation, 113 Riesz-Fr´chet theorem e for vector spaces, 29 right -handed basis, 113 inverse of a morphism, 33 invertible linear map, 21 ring, ix commutative, ix division, ix homomorphism, 7 unital, 7 with identity, ix r -skeleton, 132 section, 21 segment, closed, 131 self-adjoint, 81 semigroup, ix sequence exact, 42 series formal power, 59 shift operator, 56 short exact sequence, 42, 129 shuﬄe, 111 σ (interchange operation), 4 sign

of a permutation, 94 similar, 69 simplex closed, 131 dimension of a, 131 face of a, 132 open, 131 oriented, 132 plane of a, 131 vertex of a, 132 simplicial coboundary, 135 cocycle, 135 cohomology group, 135 complex, 132 dimension of a, 132 simplicial homology group, 133 skeleton of a complex, 132 skew-symmetric, 94 smallest, 17 smooth diﬀerential form, 115 function, 115 submanifold, 135 triangulation, 135 vector ﬁeld, 115 span, 13 span(A) (the span of A), 13 spectral theorem for complex inner product spaces, 89 for vector spaces, 69 spectrum, 55 S(p, q) (shuﬄe permutations), 111 square summable sequence, 78 star operator, 114 Stokes’ theorem, 143 submanifold, 135 subprojection, 89 subspace, 9 aﬃne, 131 complementary, 37 reducing, 67 sum direct, 37, 111 of sets of vectors, 9 supp F (the support of f ), 27 support, 27 switching operation, 4 symmetric group, 94 matrix, 19 skew-, 94 tangent, 93 at a point, 115 space, 115 tensor algebra, 111 decomposable, 102 elementary, 102 homogeneous, 102

INDEX

163

product, 101 of linear maps, 105 ∗ Tm (cotangent space at m), 115 Tm (tangent space at m), 115 total derivative, 93 transitive, 17 transpose, 81, 98 transposition, 94 triangulated manifold, 135 trilinear, 94 T (V ) (the tensor algebra of V ), 111 U (A) (unitary elements of A), 81 unilateral shift operator, 56 unique factorization theorem, 64 uniqueness essential, 43 of products, coproducts, 43 unit vector, 77 unital ring, ix ring homomorphism, 7 unitarily diagonalizable, 89 equivalent, 89 unitary, 81 unitization, 51 universal, 149 upper bound, 17 half-space, 140 vector ﬁeld, 115, 121 associated with a diﬀerential form, 121 smooth, 115 space, 8 adjoint, 34 dimension of a, 19 free, 99 orientation of a, 113 oriented, 113 spectral theorem for, 69 unit, 77 vertex of a simplex, 132 vol (volume element or form), 114, 116 volume element, 114 form, 116 wedge product, 107 zero divisor, 7 Z+ -graded algebra, 109 Zorn’s lemma, 18 Z p (M ) (space of de Rham p -cocycles), 127 Z p (K) (space of simplicial p -cocycles), 135 Zp (K) (space of simplicial p -cycles), 133