PROBLEMS IN LINEAR ALGEBRA

Hojoo Lee

Version 0.1 [2005/06/06]

1
2 PROBLEMS IN LINEAR ALGEBRA

1. Introduction

The heart of Mathematics is its problems. Paul Halmos

The purpose of this book is to present a collection of interesting questions

in linear algebra. The book is available at

http://my.netian.com/∼ideahitme/eng.html

This is a baby edition of the book. I would greatly like to hear about
interesting problems in linear algebra. I also would appreciate hearing about
any errors in the book, even minor ones. You can send all comments to the
author at [email protected].

2. Symbol Index

V a vector space
V∗ the dual space of V = the set of linear functionals on V
L(V ) the set of linear operators in the vector space V
Mn (k) the set of n × n matrices with entries in k
(aij )m×n the matrix of size m × n
(aij )n the matrix of size n × n
(, ) an inner product
detX the determinant of the matrix X
I the identity matrix or the identity operator
T∗ the adjoint operator of T
AT the transpose of the matrix A
(aij )m×n = (aij )m×n , where aij ∈ C
¡ ¢T
A∗ = A
P (F) the set of all polynomials with coefficients in F
Pm (F) = {p(x) ∈ F[x]|deg p(x) < m}
 PROBLEMS IN LINEAR ALGEBRA 3

Contents
1. Vector Spaces 4
Vector Spaces 4
Inner Product Spaces 5
Linear Functionals and Dual Spaces 7
2. Theory of Matrices 8
Matrices 8
Traces and Determinants 10
3. Linear Maps 13
Linear Operators 13
Polynomials 15
Eigenvalues and Eigenvectors 16
Nilpotence 16
Bilinear and Quadratic Forms 17
Isometries 17
Projections 18
Contraction Operators 19
Adjoint Operators 19
Self-adjoint Operators 19
Normal Operators 20
Positive Operators 20
Square Roots and Cube Roots 20
4. References 22
4 PROBLEMS IN LINEAR ALGEBRA

1. Vector Spaces
1.1. Vector Spaces.
1. [G015] Let E be a real vector space. Consider the set E × E of ordered
pairs (x, y) with x, y ∈ E. Show that the set E × E becomes a complex vector
space under the operations:
(x1 , y1 ) + (x2 , y2 ) = (x1 + x2 , y1 + y2 ), (α + iβ)(x, y) = (αx − βy, αy + βx),
where x, y ∈ E and α, β ∈ R.
2. [I177] If V is a vector space over an infinite field F , prove that V cannot
be written as the set-theoretic union of a finite number of proper subspaces.
3. [I175] Let V be the set of all sequences (a1 , a2 , · · · ) (ai ∈ R), where
equality, addition and scalar multiplication are defined componentwise. We
define W and U by
W = {(a1 , a2 , · · · ) ∈ V | lim an = 0},
n→∞
∞
X
U = {(a1 , a2 , · · · ) ∈ V | an 2 is finite}.
n=1
(1) Prove that V is a vector space over F .
(2) Prove that W is a subspace of V .
(3) Prove that U is a subspace of V and is contained in W .
4. [P048] Let e1 , · · · , em be vectors in an n-dimensional vector space V with
m ≥ n + 2. Prove that there exist scalars α1 , · · · , αm not all of them equal
to 0 such that
Xm Xm
αi ei = 0 and αi = 0.
i=1 i=1

5. [P046] Let T ∈ L(V ). Assume that the vectors

ξ, T ξ, · · · , T n ξ
are linearly dependent for any ξ ∈ V. Show that the operators
I, T, · · · , T n
are linearly dependent.
6. [N025] Show that if W1 and W2 are subspaces, then W1 ∪ W2 is not a
subspace unless one is a subspace of the other.
7. [A020] Prove or give a counterexample : if U1 , U2 , W are subspaces of a
vector space V such that
U1 + W = U2 + W,
then U1 = U2 .
 PROBLEMS IN LINEAR ALGEBRA 5

8. [N025] Let U1 , U2 , W are subspaces of a finite-dimensional vector space

V for which
U1 + W = U2 + W, U1 ∩ W = U2 ∩ W, U1 ⊂ U2 .
Show that U1 = U2 .
9. [A020] Prove or give a counterexample : if U1 , U2 , W are subspaces of a
vector space V such that
U1 ⊕ W = U2 ⊕ W,
then U1 = U2 .
10. [N025] If V = W1 ⊕ W2 and W is any subspace of V such that W1 ⊂ W
or W2 ⊂ W , show that
W = (W ∩ W1 ) ⊕ (W ∩ W2 ).
Show by an example that the condition W1 ⊂ W or W2 ⊂ W is necessary.
11. [A034] Suppose that U1 , · · · , Um are subspaces of a finite dimensional
vector space V such that
V = U1 + · · · + Um
and
dimV = dimU1 + · · · + dimUm
Show that V = U1 ⊕ · · · ⊕ Um .
12. [A036] Suppose that p0 , p1 , · · · , pm are polynomials in Pm (F) such that
p0 (2) = · · · = pm (2) = 0 for each j. Prove that (p0 , p1 , · · · , pm ) is linearly
dependent in Pm (F).
13. [A061] Prove that if V is a finite dimensional vector space with dimV >
1, then the set of noninvertible operators on V is not a subspace of L(V ).
14. [I267] Let V be a finite dimensional vector space over F. Prove that T ∈
L(V ) is invertible if and only if whenever v1 , · · · , vn are linearly independent,
then T (v1 ), · · · , T (vn ) are also linearly independent
1.2. Inner Product Spaces.
15. [A122] Let V be a real inner product space. Let u, v ∈ V . Prove that
(u, v) = 0
if and only if, for all a ∈ R,
ku + avk ≥ kuk.
16. [A122] Suppose that e1 , · · · , em is an orthonormal list of vectors in a
complex inner product space V . Let v ∈ V . Prove that
kvk = |(v, e1 )|2 + · · · + |(v, em )|2
if and only if
v ∈ span(e1 , · · · , em ).
6 PROBLEMS IN LINEAR ALGEBRA

17. [P048] Given vectors e1 , · · · , en+1 in an n-dimensional Euclidean space

such that
(ei , ej ) < 0
for all i 6= j, prove that any n of these vectors form a basis.
18. [N181] Let X = {e1 , · · · , em } be a set of vectors in the n-dimensional
vector space V . Consider the matrix G = (gij ) where
gij = (ei , ej ).
Show that if X is linearly dependent, then the columns of G are also linearly
dependent. Show that if X is linearly independent, then the columns of G
are also linearly independent. det G is known as the Gramian of the set
X. Show that X is linearly dependent if and only if det G 6= 0. Choose an
orthonormal basis in V and represent the vectors in X with respect to that
basis. Show that G can be represented as the product of an m × n matrix
and n × m matrix. Show that Det G ≥ 0.
19. [G220] Consider two inner product space E and F . Prove that an inner
product is defined in the space L(E, F ) by
(φ, ψ) = tr (ψ ∗ φ)
for all φ, ψ ∈ L(E, F ). Derive the inequality
(tr (ψ ∗ φ))2 ≤ tr (ψ ∗ ψ)tr (φ∗ φ)
and show that the equality holds if and only if ψ = λφ, where λ ∈ R.
20. [G191] Let E be an inner product space. Prove the Ptolemy-inequality
k(z − x)yk + k(y − z)xk ≥ k(x − y)zk
for all x, y, z ∈ E. Assume that x, y, z are all nonzero vectors in E. Prove
that the equation
k(z − x)yk + k(y − z)xk = k(x − y)zk
holds if and only if the four points x, y, z, 0 are contained on a circle such
that the pairs x, y and z, 0 separate each other.
21. [G191] Consider the vector space S of all infinite sequence x = (x1 , x2 , · · · )
such that
X∞
xi 2
n=1
is finite. Show that
∞
X
xi yi
n=1
is finite whenever x, y ∈ S. Prove that
X∞
(x, y) = xi yi
n=1
is an inner product.
 PROBLEMS IN LINEAR ALGEBRA 7

1.3. Linear Functionals and Dual Spaces.

22. [I191] Let f, g ∈ V ∗ . Suppose that, for any v ∈ V , f (v) = 0 implies that
g(v) = 0. Prove that f = λg for some scalar λ.
23. [G129] Let f be a linear functional on L(E, E) Show that f can be
written as
f (φ) = tr (φα)
where α is a fixed linear operator on E. Prove that α is uniquely determined
by f .
8 PROBLEMS IN LINEAR ALGEBRA

2. Theory of Matrices
2.1. Matrices.
24. [BSp89 ] Let B = (bij ) be a real 20 × 20 matrix such that
bii = 0 for all i = 1, · · · , 20, bij ∈ {1, −1} for all i 6= j
Show that B is invertible.
25. [R011] Show that every square matrix can be written as the sum of a
symmetric matrix and a skew-symmetric matrix
26. [P068] Prove that any square matrix A ∈ Mn (C) is the sum of two
invertible matrices.
27. [N160] Show by example that, if A and B are symmetric, it is not
necessarily true that AB are symmetric. What can be concluded if A and B
are symmetric and AB = BA ?
28. [N160] Show the following : If A is skew-symmetric, then A2 are sym-
metric. If A is skew-symmetric and B is symmetric, then AB − BA is
symmetric. If A is skew-symmetric and B is symmetric, then AB is skew-
symmetric if and only if AB = BA.
29. [N170] Show that A is a real non-singular matrix, then AT A is positive
definite.
30. [N170] Show that if A is real symmetric positive definite matrix, then
there exists a real non-singular matrix P such that A = P T P .
31. [N170] If A1 , · · · , Ar are real symmetric matrices, show that
A1 2 + · · · + Ar 2 = O
implies A1 = · · · = Ar = O.
32. [N174] If A1 , · · · , Ar are Hermitian matrices, show that
A1 2 + · · · + Ar 2 = O
implies A1 = · · · = Ar = O.
33. [N ] Show that if A is complex, then A∗ A is Hermitian non-negative
semi-definite.
34. [N174] Show that A is complex and A∗ A = 0, then A = 0.
35. [N174] Show that A is Hermitian and A2 = 0, then A = 0.
36. [P175, I283] Suppose that X ∈ Mn (C) commutes with all matrices.
Show that
X = λI
for some λ ∈ C.
 PROBLEMS IN LINEAR ALGEBRA 9

37. [P042] Let A ∈ Mn (C). Suppose that

tr (AX) = 0
for any X ∈ Mn (C with tr X = 0. Show that
A = λI
for some λ ∈ C.
38. [P068, A222, I284] Are there real matrices A and B such that AB −
BA = I?
39. [I284] If F is of characteristic of 2, prove that in M2 (F) it is possible
to find matrices such that AB − BA = I.
40. [P068] Let A and B be complex matrices with AB − BA = A. Show
that detA = 0.
41. [P068, BF a81,Su81,Sp84,F a87,F a95 ] Let A and B be two real matrices and
A = P −1 BP , where P is a complex matrix. Show that there exists a real
matrix Q such that A = Q−1 BQ.
42. [P048,BF a95 ] Let A = (aij )n ∈ Mn (C). Assume that
X
|aii | > |aki |
k6=i

for all i ∈ {1, · · · , n}. Show that A is invertible.

43. [BSp93 ] Let A = (aij )n ∈ Mn (C) such that
n
X
|aij |
j=1

for all i ∈ {1, · · · , n}. Show that I − A is invertible.

44. [P103] Suppose that A is a real skew-symmetric matrix. Show that I +A
is invertible.
45. [P176] Let tr A = 0. Show that there exist matrices X and Y such that
X is Hermitian, tr Y = 0, and A = XY − Y X.
46. [P084] Prove that for any nonzero square matrix A there exists a matrix
X such that the matrices X and A + X have no common eigenvalues.
47. [P105] Let A ∈ Mn (C) be a normal matrix. Prove that there exists a
normal matrix B such that A = B 2 .
48. [P105] The complex matrices A and B are normal. Prove that BA is
also normal.
49. [P108] Let A and B be two matrices of order n. Prove that A + λB is
a nilpotent matrix for n + 1 distinct values of λ, then A and B are nilpotent
matrices.
10 PROBLEMS IN LINEAR ALGEBRA

50. [P151] Let A and B be Hermitian projections, i.e., A2 = A and B 2 = B.

Show that the eigenvalues of AB are real and belong to the segment [0, 1].
51. [BSp96 ] Prove or disprove : for any 2 × 2 matrix A over C, there is a
2 × 2 matrix B such that A = B 2 .
52. [BSu85 ] Let ² > 0. Show that there is no real 2 × 2 matrix A such that
· ¸
2 −1 0
A = .
0 −1 − ²
53. [BSp92 ] Find a square root of the matrix
 
−1 3 3
 0 4 5 .
0 0 9
How many square roots does this matrix have?
54. [BSu83 ] Let A be an n × n Hermitian matrix satisfying the condition
A5 + A3 + A = 3I.
Show that A = I
55. [BSp92 ] Find a real matrix B such that
 
2 0 0
B 4 = 0 2 0 .
0 −1 1
56. [R140] Prove that a complex 2 × 2 matrix is not diagonalizable if and
only if it’s similar to a matrix of the form
· ¸
a b
,
0 a
where b 6= 0.
57. [N045] Show be example that it is possible for two matrices A and B to
have the same rank while A2 and B 2 have different ranks.
58. [N202] Show that if M = A+Bi where A and B are real and symmetric,
then M is normal if and only if A and B are commute.
59. [G098] Two n × m matrices C and D are called equivalent if there exists
a regular n×n matrix A and a regular m×m matrix B such that D = ACB.
Prove that two matrices are equivalent if and only if they have the same rank.
2.2. Traces and Determinants.
60. [N212] Show that if A and B are real symmetric matrices, and A is
positive definite, then the roots of det (B − xA) = 0 are all real.
 PROBLEMS IN LINEAR ALGEBRA 11

61. [P180] Let f be a linear functional on Mn . Prove that if

f (XY ) = f (Y X)
for any matrices X and Y , then f (X) = λ tr X.
62. [I320] Let V be a n-dimensional vector space over a field F which has
an infinite number of elements. A function f , defined on V and having its
values in F, satisfies
(1) f (A + B) = f (A) + f (B) ; (2) f (λA) = λf (A) ; (3) f (BAB −1 ) =
f (A)4 ; for every A ∈ V , λ ∈ F, and invertible elements B ∈ V . Prove that
f (A) = α0 trA for a particular α0 ∈ F and all A ∈ V .
63. [P050] Let A be a square matrix such that rank A = 1. Prove that
det(A + I) = 1 + trA.
64. [P079] Let A and B be complex matrices of order n such that
tr Am = tr B m for all m ∈ {1, · · · , n}.
Show that the eigenvalues of A and B coincide.
65. [S243] Let A = (aij )n be a real matrix. Suppose that
|aij | ≤ M
n
for some M > 0. Prove the inequality detA ≤ n 2 M n . Show that this esti-
mate cannot be improved for n = 2m .
66. [A245] Let T be a linear operator on a complex inner product space V .
Suppose that (e1 , · · · , en ) is an orthonormal basis of V . Show that
tr (T ∗ T ) = kT e1 k2 + · · · + kT en k2 .
Conclude that the right side of the equation above is independent of which
orthonormal basis (e1 , · · · , en ) is chosen for V .
67. [A244] Suppose that V is a complex inner product space. Prove that if
T ∈ L(V ) is a positive semidefinite operator and tr T = 0, then T = 0.
68. [A244] Let V be a complex inner product space and let T ∈ L(V ). Show
that
tr T ∗ = T .
69. [A244] Let V be a complex inner product space. Prove that if P ∈ L(V )
satisfies P 2 = P , then traceT is a nonnegative integer.
70. [A244] Let V be a complex inner product space. Let v, w ∈ V . Define
T ∈ L(V ) by T v = (u, v)w. Find a formula for tr T .
71. [A244] Let V be a real inner product space. Suppose that V has a basis
consisting of eigenvectors of T ∈ L(V ). Prove that
tr (T 2 ) ≥ 0.
12 PROBLEMS IN LINEAR ALGEBRA

72. [G112] Given an n × n-matrix A = (aij ) define the matrix B = (bij ) by

bij = (−1)i+j aij
for all i, j ∈ {1, · · · , n}. Prove that det B = det A.
73. [G112] Let α1 , · · · , αn be complex numbers. Prove that
 
α1 α2 · · · αn−1 αn
 α2 α3 ··· αn α1  n(n−1)
  = (−1) 2 β1 · · · βn ,
 ··· 
αn α1 · · · αn−2 αn−1
where β1 , · · · , βn are defined by
X n µ ¶
2πk 2πk j
βk = cos + i sin αj (k = 1, · · · , n).
n n
j=1

(Hint: Multiply the above matrix by the matrix

 
²1 ²2 ··· ²n−1 ²n
 ²1 2 ²2 2 ··· ²n−1 2 ²n 2 
 ,
 ··· 
²1 n ²2 n ··· ²n−1 n ²n n

where ²k = cos 2πk 2πk

n + i sin n .)
 PROBLEMS IN LINEAR ALGEBRA 13

3. Linear Maps
3.1. Linear Operators.
74. [S116] In the space of all polynomials in the variable t, consider the
operators A and B defined by relations
A(a0 + a1 t + · · · + an tn ) = a1 + a2 t + · · · + an−1 tn ,
B(a0 + a1 t + · · · + an tn ) = a0 t + a1 t2 + · · · + an tn+1 .
(1) Show that A and B are linear operators such that AB = I and BA 6= I.
(2) Does the operator A have an inverse ?
(3) Show that the operator B has infinitely many left inverses.
75. [N036] In the space of all real polynomials in the variable x, consider
the operators σ and τ defined by relations
Xn n
X
σ( ai xi ) = iai xi−1 ,
i=0 i=0
n
X n
X 1
τ( ai xi ) = ai xi+1 .
i+1
i=0 i=0
Show that στ = I, but that τ σ 6= I.
76. [N118] Show that any 2 × 2 complex matrix T such that T 2 + I = O is
similar to the matrix · ¸
0 −1
.
1 0
77. [P140] Let f : Mn −→ Mn be an invertible map and
f (XY ) = f (X)f (Y )
for any matrices X and Y . Prove that f (X) = AXA−1 , where A is a fixed
matrix.
78. [I184] Define a linear operator T on Pm (F) by
T (a0 + a1 x + · · · + an−1 xn−1 ) = a0 + a1 (x + 1) + · · · + an−1 (x + 1)n−1 .
Prove that T is an isomorphism.
79. [I269] If T ∈ L(V ) is nonzero and noninvertible, prove that there is an
element Z ∈ L(V ) such that T Z = 0 but ZT 6= 0.
80. [I273, H1 069] Let V be a finite dimensional vector space over F. Let
S, T ∈ L(V ). Prove that (ST − T S)2 commutes with all elements in L(V ).
81. [H1 061] Let V be a complex vector space over F. Let A, B ∈ L(V ).
Suppose that AB − BA commutes with A. Show that
Ak B − BAk = kAk−1 (AB − BA)
for every positive integer k.
14 PROBLEMS IN LINEAR ALGEBRA

82. [H1 069] Let T be the linear operator on C2 defined by

T (x, y) = (x + y, y)
for all x, y ∈ C. Prove that if a linear operator U on C2 commutes with T ,
then there exists a polynomial p such that U = p(T ).
83. [I269] Let V be a finite dimensional vector space. Prove that every linear
operator on V can be written as a sum of invertible operators.
84. [A095] Suppose that T ∈ L(V ) is such that every subspace of V with
dimension dimV − 1 is invariant under T . Prove that T is a scalar multiple
of the identity of the identity operator.
85. [A159] Give an example of a real inner product space V , T (V ), and
p(x) = x2 + αx + β ∈ R[x] such that p(x) has no real root and that p(T ) is
not invertible.
86. [H2 236] Suppose that two linear operators A and B on a vector space
are such that range A ⊂ range B. Show that A is left divisible by B, i.e.,
there exists a linear operator T such that A = BT .
87. [H2 236] Suppose that two linear operators A and B on a vector space
are such that Ker B ⊂ Ker A. Show that A is right divisible by B, i.e.,
there exists a linear operator T such that A = SB.
88. [G049] Prove that if φ : E −→ F and ψ : E −→ G are linear maps,
then the relation Ker φ ⊂ Ker µ is necessary for the existence of a linear
map ξ : F −→ G such that φ = ξψ.
89. [G054] Let φ : E −→ F be a linear map. Consider the space M l (φ) of
linear maps ψ : F −→ E such that ψφ = 0. Prove that if φ is surjective
then M l (φ) = 0.
90. [G054] Let φ : E −→ F be a linear map. Consider the space M r (φ) of
linear maps ψ : F −→ E such that φψ = 0. Prove that if φ is injective then
M r (φ) = 0.
91. [G054] Prove Schur’s lemma : Let E and F be vector spaces and α :
E −→ F be a linear map. Assume SE and SF are two sets of linear maps
of E and F such that
αSE = SF α,
i.e., to every linear map φ ∈ SE there exists a linear map ψ ∈ SF such that
αφ = ψα and conversely. Prove that α = 0 or α is a linear isomorphism of
E onto F .
92. [H2 263] Let A and B be linear operators on a vector spaces. Prove that
if at least one of A and B is invertible, then AB and BA are similar.
93. [BSp91 ] Let A be a linear operator on an n-dimensional vector space over
C with characteristic polynomial (x − 1)n . Prove that A is similar to A−1 .
 PROBLEMS IN LINEAR ALGEBRA 15

94. [G108] Let α : E −→ F be a linear isomorphism and φ be a linear

operator on E. Prove that det αφα−1 = det φ.
95. [G109] Let J be a linear operator on a finite dimensional vector space
E such that J 2 = −I. Show that the dimension of E must be even. Prove
that det J = 1.
96. [G129] Let φ and ψ be two linear operators on E. Consider the sum
X
δ(x1 , · · · , φ(xi ), · · · , ψ(xj ), · · · , xn )
i6=j

where δ is a determinant function in E. This sum is again a determinant

function and hence it can be written as
X X
δ(x1 , · · · , φ(xi ), · · · , ψ(xj ), · · · , xn ) = B(φ, ψ) δ(x1 , · · · , xn ).
i6=j i6=j

By the above relation a bilinear function B is defined in the space L(E, F ).

Prove : (1) B(φ, ψ) = tr φ tr ψ − tr (φψ). (2) 21 B(φ, φ) = (−1)n α2 ,
where α2 is the coefficient of xn−2 in the characteristic polynomial of φ. (3)
n ¡ ¢
α2 = (−1)
2 (tr φ)2 − tr (φ2 ) .
97. [G130] Let A : L(E) −→ L(E) be a linear map such that
A(φψ) = A(φ)A(ψ) φ, ψ ∈ L(E)
and
A(I) = I.
Prove that tr A(φ) = tr φ.
3.2. Polynomials.
98. [I264] Let V be a finite dimensional vector space over F. Show that
T ∈ L(V ) is invertible if and only if the constant term of the minimal
polynomial for T is not 0.
99. [A189, I265, H1 064] Let V be a finite dimensional vector space over F.
Suppose that T ∈ L(V ) is invertible. Prove that there exists a polynomial
p(x) ∈ F[x] such that
T −1 = p(T ).
(Hint. Find a nonzero polynomial q of least degree such that q(T ) = 0 and
prove its constant term cannot be 0.)
100. [H1 102] Suppose that A and B are linear operators on a finite finite
dimensional vector space. (1) Prove that A is a projection, then AB and
BA have the same characteristic polynomial. (Hint: Choose a basis that the
matrix of A as simple as possible and then compute directly with matrices.)
(2) Prove that, in all cases, AB and BA have the same characteristic poly-
nomial. (Hint: Find an invertible P such that P A is a projection and apply
(1) to P A and BP −1 .
16 PROBLEMS IN LINEAR ALGEBRA

3.3. Eigenvalues and Eigenvectors.

101. [S117, A095] Prove that every vector of the space V is an eigenvalue
of the operator T , then T = λI for some scalar λ.
102. [A094] Suppose that T ∈ L(V ) and dim (range T ) = k. Prove that T
has at most k + 1 distinct eigenvalues.
103. [A094] Suppose that T ∈ L(V ) is invertible and λ ∈ F − {0}. Prove
that λ is an eigenvalue of T if and only if λ1 is an eigenvalue of T −1 .
104. [A095] Let S, T ∈ L(V ). Prove that ST and T S have the same eigen-
values.
105. [A095] Let S, T ∈ L(V ). Suppose that T has dimV distinct eigenvalues
and that S has the same eigenvectors as T (not necessarily with the same
eigenvalues). Prove that ST = T S.
106. [A096] Suppose that V is a real vector space and that T ∈ L(V ) has
no eigenvalues. Prove that every subspace of V invariant under T has even
dimension.
3.4. Nilpotence.
107. [I303] Let S, T ∈ L(V ) be nilpotent operators with ST = T S. Prove
that ST and S + T are nilpotent.
108. [A189] Let V be a complex vector space. Suppose that 0 is the only
eigenvalue of N ∈ L(V ). Prove that N is nilpotent. Give an example to
show that this is not necessarily true on a real vector space.
109. [I315, BF a99 ] Let V be a finite dimensional vector space over a field
F of characteristic 0. Suppose that tr T i = 0 for all i ≥ 1. Show that T is
nilpotent. (However, if F is a field of characteristic 2, the identity operator
has trace 0.)
110. [I315] Let V be a finite dimensional vector space over a field F of
characteristic 0. Suppose that S, T ∈ L(V ) are such that ST −T S commutes
with S. Show that ST − T S is nilpotent.
111. [I265] Let V be a finite dimensional vector space over F. Suppose that
T ∈ L(V ) is not invertible. Show that there exists a Z ∈ L(V ) such that
T Z = ZT = 0.
112. [I268] If T ∈ L(V ) is nilpotent, prove that
α0 I + α1 T + · · · + αk T k
is invertible, provided that α0 6= 0.
113. [I285] If N ∈ L(V ) is nilpotent, prove that I + N is invertible and find
its inverse as a polynomial in N .
114. [A189] Suppose that V is a complex inner-product space and that N ∈
L(V ) is nilpotent. Prove that there exists an orthonormal basis of V with
respect to which N has an upper triangular matrix.
 PROBLEMS IN LINEAR ALGEBRA 17

3.5. Bilinear and Quadratic Forms.

115. [BSu80,F a92 ] Let E be a finite dimensional vector space over a field F.
Suppose that B : E × E −→ F is a linear map (not necessarily symmetric).
Define subspaces
E1 = {x ∈ E|B(x, y) = 0 for all y ∈ E},
E2 = {y ∈ E|B(x, y) = 0 for all x ∈ E}
Prove that dim E1 = dim E2 .
116. [S212] Prove that every nonnegative bilinear form of rank r can in the
space Rn can be represented as a sum of r nonnegative bilinear forms of
rank 1.
117. [S212] Prove that every bilinear form B of rank 1 in a n-dimensional
space K is of the form
B(x, y) = f (x)g(y),
where f and g are linear functionals on K.
118. [S310] Two quadratic forms A(x, y) and B(x, y) in Rn are said to be
comparable if the inequality
A(x, x) ≥ B(x, x)
holds for all x ∈ Rn . Let λ1 ≥ · ≥ λn be the canonical coefficients of the
form A(x, x), and let µ1 ≥ · ≥ µn be the canonical coefficients of the form
B(x, x). Show that the inequality
λk ≥ µk
holds for all i = 1, · · · , n. (This is obvious in the case when A(x, x) and
B(x, x) have a common canonical basis.)
3.6. Isometries.
119. [G237] Let φ be a map of the inner product space E into itself such
that φ(0) = 0 and
kφ(x) − φ(y)k = kx − yk.
Prove that φ is linear.
120. [G236] Let φ be a regular linear operator of a Euclidean space E such
that (φ(x), φ(y)) = 0 whenever (x, y) = 0. Prove that φ is of the form
φ = λτ , where λ 6= 0 and τ is a rotation.
121. [S244] Let X = {x1 , · · · , xk } and Y = {y1 , · · · , yk } be two finite sys-
tems of vectors in a Euclidean space Rn . Show that a necessary and suf-
ficient condition for the existence of an isometric operator T taking every
vector xi into the corresponding vector yi (i = 1, · · · , k) is that the following
relations hold.
(xi , xj ) = (yi , yj ) for all i, j ∈ {1, · · · , k}
18 PROBLEMS IN LINEAR ALGEBRA

122. [A160] Prove that if S ∈ L(R3 ) is an isometry, then there exists a

nonzero vector x ∈ R3 such that S 2 x = x.
123. [A160] Prove or give a counterexample : if S ∈ L(V ) and there exists
an orthonormal basis (e1 , · · · , en ) of a complex inner product space V such
that kS(ej )k = 1 for each ej , then S is an isometry.
124. [S244] Let T be a linear operator in a Euclidean space Rn with n ≥ 3.
Suppose that T does not change the area of any parallelogram, so that
V [x, y] = V [T x, T y],
for all x, y ∈ Rn . Show T is an isometry.
125. [S244] (M. A. Krasnosyelski) Let T be a linear operator in a Euclidean
space Rn . Assume that T does not change the volume of any k-dimensional
hyper-parallelepiped with k < n. Show that T is an isometry.
126. [P103] Suppose that the absolute values of all eigenvalues of an operator
A are equal to 1 and |Ax| ≤ |x| for all x. Prove that A is a unitary operator.
127. [S246] Show that if the polynomial [P (x)]k is an annihilating polyno-
mial of an isometry T , then so is the polynomial P (x).
128. [G237] If φ is a rotation of an n-dimensional Euclidean space, show
that
|tr φ| ≤ n.
129. [G237] Prove that the characteristic polynomial of a proper rotation
satisfies the relation µ ¶
n 1
f (x) = (−x) f .
x
3.7. Projections.
130. [N118] Show that any two projections of the same rank are similar.
131. [A124] Let V be a complex inner product space. Prove that if P ∈ L(V )
is such that P 2 = P and
kP vk ≤ kvk
for all v ∈ V , then P is an orthogonal projection.
132. [H1 078] Let E, F and G be projections on a vector space over a field
whose characteristic is not equal to 2. Prove that if E + F + G = I, then
EF = F E = EG = GE = F G = GF = 0.
Does the proof work for four projections instead of three?
133. [H1 109] Suppose that the scalar field has characteristic zero. Prove that
if E1 , · · · , Ek and E1 + · · · + Ek are projections, then Ei Ej = 0 whenever
i 6= j. (Hint: From the fact that tr (E1 + · · · + Ek ) = tr (E1 ) + · · · + tr (Ek )
conclude that the range of E1 + · · · + Ek is the direct sum of the ranges of
E1 , · · · , Ek .)
 PROBLEMS IN LINEAR ALGEBRA 19

3.8. Contraction Operators.

134. [P085] Let A : Rn −→ Rn be a contraction operator, i. e., |Ax| ≤ |x|
for all x ∈ Rn . The space Rn can be considered as a subspace of R2n .
Prove that A is the restriction to Rn of the composition of an orthogonal
transformation of R2n and the projection on Rn .
135. [S272] A linear operator A is said to be a contraction operator on
Cn if |Ax| ≤ |x| for all x ∈ Cn . Prove that every linear operator T can
be represented as the product of a self-adjoint operator and a contraction
operator on Cn .
3.9. Adjoint Operators.
136. [A125] Let V be a complex inner product space. Suppose that T ∈ L(V )
and that U is a subspace of V . Prove that U is invariant under T if and
only if U ⊥ is invariant under T ∗
137. [A125] Let V be a complex inner product space. Let T ∈ L(V, W ).
Prove that
(1) T is injective if and only if T ∗ is surjective.
(2) T is surjective if and only if T ∗ is injective.
3.10. Self-adjoint Operators.
138. [A129] Let V be a complex inner product space and let T ∈ L(V ).
Show that T is self-adjoint if and only if (T v, v) ∈ R for all v ∈ V .
139. [A136] (Real Spectral Theorem) Suppose that V is a real inner prod-
uct space and T ∈ L(V ). Then V has an orthonormal basis consisting of
eigenvectors of T if and only if T is self-adjoint.
140. [A158] Prove or give a counterexample : the product of any two self-
adjoint operators on a finite-dimensional inner product space is self-adjoint.
141. [A158] Show that if V is a real inner product space, then the set of
self-adjoint operators on V is a subspace of L(V ).
142. [A158] Show that if V is a complex inner product space, then the set
of self-adjoint operators on V is not a subspace of L(V ).
143. [A158] Let V be a complex inner product space. Suppose that P ∈ L(V )
is such that P 2 = P . Prove that P is an orthogonal projection if and only
if P is self-adjoint.
144. [A134] Suppose that T ∈ L(V ) is self-adjoint on a complex inner prod-
uct space V and that the polynomial p(x) = x2 + αx + β ∈ R[x] has no real
root. Show that p(T ) is invertible.
145. [I322] Let V be a finite dimensional complex vector space. Let A∗ be
the Hermitian adjoint of A ∈ L(V ). If A commutes with AA∗ − A∗ A, prove
that AA∗ = A∗ A.
146. [I322] Let V be a n-dimensional complex vector space. Let A ∈ L(V )
be a Hermitian operator. If BAk = 0, prove that BA = 0.
20 PROBLEMS IN LINEAR ALGEBRA

3.11. Normal Operators.

147. [A246] Let V be a complex inner product space and let T ∈ L(V ).
Prove that if
kT ∗ vk ≤ kT vk
for all v ∈ V , then T is normal.
148. [A133] (Complex Spectral Theorem) Suppose that V is a complex inner
product space and T ∈ L(V ). Then V has an orthonormal basis consisting
of eigenvectors of T if and only if T is normal.
149. [A159] Prove that a normal operator on a complex inner-product space
is self-adjoint if and only if all its eigenvalues are real.
150. [A159] Suppose that V is a complex inner-product space and T ∈ L(V )
is a normal operator such that T 9 = T 8 . Prove that T is self-adjoint and
T2 = T.
151. [A158] Let V be a complex inner product space with dimV ≥ 2. Show
that the set of normal operators on V is not a subspace of L(V ).
152. [G225] Prove that the product of two self-adjoint operators φ and ψ is
self-adjoint if and only if φψ = ψφ.
3.12. Positive Operators.
153. [A146] Let V be a complex inner product space. An operator T ∈ L(V )
is called positive semidefinite operator if T is self-adjoint and (T v, v) ≥ 0
for all v ∈ V . Show that every positive semidefinite operator on V has a
unique square root which is positive semidefinite.
154. [A160] Suppose that T is a positive semidefinite operator on a complex
inner product space V . Prove that T is invertible if and only if (T v, v) > 0
for all v ∈ V − {0}.
3.13. Square Roots and Cube Roots.
155. [A177] Suppose that V is a complex vector space. Suppose that N ∈
L(V ) is nilpotent. Show that I + N has a square root.
156. [A178] Suppose that V is a complex vector space. Show that every
invertible T ∈ L(V ) has a square root.
157. [A159] Suppose that V is a complex inner-product space. Prove that
every normal operator on V has a square root. (An operator S ∈ L(V ) is
called a square root of T ∈ L(V ) if T = S 2 .)
158. [A160] Prove or disprove that the identity operator on C 2 has infinitely
many self-adjoint square roots.
159. [H2 314] Show that every positive operator on a finite dimensional inner
product space has a unique positive square root.
 PROBLEMS IN LINEAR ALGEBRA 21

160. [H2 154] Let A and B be positive operators√on a √

finite dimensional
inner product space such that B ≥ A. Prove that B ≥ A.
161. [A189] Prove that if V is a complex vector space, then every invertible
operator on V has a cube root.
22 PROBLEMS IN LINEAR ALGEBRA

4. References

[A] Sheldon Axler, Linear Algebra Done Right, Springer

[B] Paulo Ney de Souza, Jorge-Nuno Silva, Berkeley Problems in Math-
ematics(2nd edition), Springer
[G] Werner Greub, Linear Algebra(4th edition), Springer
[H1 ] Paul R. Halmos, Finite Dimensional Vector Space, Springer
[H2 ] Paul R. Halmos, Linear Algebra Problem Book, MAA
[I] I. N. Herstein, Topics in Algebra(2nd edition), John Wiley and Sons,
Inc.
[N] Evar D. Nering, Linear Algebra and Matrix Theory(2nd edition),
John Wiley and Sons, Inc
[P] V. V. Prasolov, Problems and Theorems in Linear Algebra, AMS
[R] Derek Robinson, A Course in Linear Algebra with Applications, So-
lutions to the Exercises, World Scientific
[S] G. E. Shilov, Translated and Edited by R. A. Silverman, Linear
Algebra, Dover