Section 7D

Isometries, Unitary Operators, and Matrix Factorization

[20/0%]
Show
LaTeX
Problem 7D.1

Suppose dim⁡V≥2\dim V \geq 2 and S∈L(V,W)S \in \mathcal{L}(V, W). Prove that SS is an isometry if and only if Se1,Se2S e_{1}, S e_{2} is an orthonormal list in WW for every orthonormal list e1,e2e_{1}, e_{2} of length two in VV.

?
Problem 7D.2

Suppose T∈L(V,W)T \in \mathcal{L}(V, W). Prove that TT is a scalar multiple of an isometry if and only if TT preserves orthogonality.

?
Remark.
?

The phrase "TT preserves orthogonality" means that ⟨Tu,Tv⟩=0\left\langle T u, T v \right\rangle = 0 for all u,v∈Vu, v \in V such that ⟨u,v⟩=0\left\langle u, v \right\rangle = 0.

Problem 7D.3
?
(a)

Show that the product of two unitary operators on VV is a unitary operator.

(b)

Show that the inverse of a unitary operator on VV is a unitary operator.

Remark.
?

This exercise shows that the set of unitary operators on VV is a group, where the group operation is the usual product of two operators.

Problem 7D.4

Suppose F=C\mathbf{F} = \mathbb {C} and A,B∈L(V)A, B \in \mathcal{L}(V) are self-adjoint. Show that A+iBA+i B is unitary if and only if AB=BAA B = B A and A2+B2=IA^{2}+B^{2} = I.

?
Problem 7D.5

Suppose S∈L(V)S \in \mathcal{L}(V). Prove that the following are equivalent.

  1. SS is a self-adjoint unitary operator.

  2. S=2P−IS = 2 P-I for some orthogonal projection PP on VV.

  3. There exists a subspace UU of VV such that Su=uS u = u for every u∈Uu \in U and Sw=−wS w = -w for every w∈U⊥w \in U^{\perp }.

?
Problem 7D.6

Suppose T1,T2T_{1}, T_{2} are both normal operators on F3\mathbf{F}^{3} with 2,5,72, 5, 7 as eigenvalues. Prove that there exists a unitary operator S∈L(F3)S \in \mathcal{L}\left(\mathbf{F}^{3}\right) such that T1=S∗T2ST_{1} = S^{*} T_{2} S.

?
Problem 7D.7

Give an example of two self-adjoint operators T1,T2∈L(F4)T_{1}, T_{2} \in \mathcal{L}\left(\mathbf{F}^{4}\right) such that the eigenvalues of both operators are 2,5,72, 5, 7 but there does not exist a unitary operator S∈L(F4)S \in \mathcal{L}\left(\mathbf{F}^{4}\right) such that T1=S∗T2ST_{1} = S^{*} T_{2} S. Be sure to explain why there is no unitary operator with the required property.

?
Problem 7D.8

Prove or give a counterexample: If S∈L(V)S \in \mathcal{L}(V) and there exists an orthonormal basis e1,…,ene_{1}, \ldots , e_{n} of VV such that ∥Sek∥=1\left\| S e_{k}\right\| = 1 for each eke_{k}, then SS is a unitary operator.

?
Problem 7D.9

Suppose F=C\mathbf{F} = \mathbb {C} and T∈L(V)T \in \mathcal{L}(V). Suppose every eigenvalue of TT has absolute value 1 and ∥Tv∥≤∥v∥\left\| T v\right\| \leq \left\| v\right\| for every v∈Vv \in V. Prove that TT is a unitary operator.

?
Problem 7D.10

Suppose F=C\mathbf{F} = \mathbb {C} and T∈L(V)T \in \mathcal{L}(V) is a self-adjoint operator such that ∥Tv∥≤∥v∥\left\| T v\right\| \leq \left\| v\right\| for all v∈Vv \in V.

?
(a)

Show that I−T2I-T^{2} is a positive operator.

(b)

Show that T+iI−T2T+i \sqrt{I-T^{2}} is a unitary operator.

Problem 7D.11

Suppose S∈L(V)S \in \mathcal{L}(V). Prove that SS is a unitary operator if and only if

{Sv:v∈V and ∥v∥≤1}={v∈V:∥v∥≤1}. \left\{ S v : v \in V \text{ and } \left\| v\right\| \leq 1\right\} = \left\{ v \in V : \left\| v\right\| \leq 1\right\} .
?
Problem 7D.12

Prove or give a counterexample: If S∈L(V)S \in \mathcal{L}(V) is invertible and ∥S−1v∥=∥Sv∥\left\| S^{-1} v\right\| = \left\| S v\right\| for every v∈Vv \in V, then SS is unitary.

?
Problem 7D.13

Explain why the columns of a square matrix of complex numbers form an orthonormal list in Cn\mathbb {C}^{n} if and only if the rows of the matrix form an orthonormal list in Cn\mathbb {C}^{n}.

?
Problem 7D.14

Suppose v∈Vv \in V with ∥v∥=1\left\| v\right\| = 1 and b∈Fb \in \mathbf{F}. Also suppose dim⁡V≥2\dim V \geq 2. Prove that there exists a unitary operator S∈L(V)S \in \mathcal{L}(V) such that ⟨Sv,v⟩=b\left\langle S v, v \right\rangle = b if and only if ∣b∣≤1\left|b\right| \leq 1.

?
Problem 7D.15

Suppose TT is a unitary operator on VV such that T−IT-I is invertible.

?
(a)

Prove that (T+I)(T−I)−1(T+I)(T-I)^{-1} is a skew operator (meaning that it equals the negative of its adjoint).

(b)

Prove that if F=C\mathbf{F} = \mathbb {C}, then i(T+I)(T−I)−1i(T+I)(T-I)^{-1} is a self-adjoint operator.

Remark.
?

The function z↦i(z+1)(z−1)−1z \mapsto i(z+1)(z-1)^{-1} maps the unit circle in C\mathbb {C} (except for the point 1) to R\mathbf{R}. Thus (b) illustrates the analogy between the unitary operators and the unit circle in C\mathbb {C}, along with the analogy between the self-adjoint operators and R\mathbf{R}.

Problem 7D.16

Suppose F=C\mathbf{F} = \mathbb {C} and T∈L(V)T \in \mathcal{L}(V) is self-adjoint. Prove that (T+iI)(T−iI)−1(T+i I)(T-i I)^{-1} is a unitary operator and 1 is not an eigenvalue of this operator.

?
Problem 7D.17

Explain why the characterizations of unitary matrices given by 7.57 hold.

7.57 (characterizations of unitary matrices): suppose QQ is an nn-by-nn matrix. Then the following are equivalent: (a) QQ is a unitary matrix; (b) the rows of QQ form an orthonormal list in Fn\mathbf{F}^{n}; (c) ∥Qv∥=∥v∥\left\| Q v\right\| = \left\| v\right\| for every v∈Fnv \in \mathbf{F}^{n}; (d) Q∗Q=QQ∗=IQ^{*} Q = Q Q^{*} = I, the nn-by-nn matrix with 1's on the diagonal and 0's elsewhere.

?
Problem 7D.18

A square matrix AA is called symmetric if it equals its transpose. Prove that if AA is a symmetric matrix with real entries, then there exists a unitary matrix QQ with real entries such that Q∗AQQ^{*} A Q is a diagonal matrix.

?
Problem 7D.19

Suppose nn is a positive integer. For this exercise, we adopt the notation that a typical element zz of Cn\mathbb {C}^{n} is denoted by z=(z0,z1,…,zn−1)z = (z_{0}, z_{1}, \ldots , z_{n-1}). Define linear functionals ω0,ω1,…,ωn−1\omega_{0}, \omega_{1}, \ldots , \omega_{n-1} on Cn\mathbb {C}^{n} by

ωj(z0,z1,…,zn−1)=1n∑m=0n−1zme−2πijm/n. \omega _{j}(z_{0}, z_{1}, \ldots , z_{n-1}) = \frac{1}{\sqrt{n}} \sum _{m = 0}^{n-1} z_{m} e^{-2 \pi i j m / n} .

The discrete Fourier transform is the operator F:Cn→Cn\mathcal{F} : \mathbb {C}^{n} \rightarrow \mathbb {C}^{n} defined by

Fz=(ω0(z),ω1(z),…,ωn−1(z)). \mathcal{F} z = (\omega _{0}(z), \omega _{1}(z), \ldots , \omega _{n-1}(z)) .
?
(a)

Show that F\mathcal{F} is a unitary operator on Cn\mathbb {C}^{n}.

(b)

Show that if (z0,…,zn−1)∈Cn(z_{0}, \ldots , z_{n-1}) \in \mathbb {C}^{n} and znz_{n} is defined to equal z0z_{0}, then

F−1(z0,z1,…,zn−1)=F(zn,zn−1,…,z1). \mathcal{F}^{-1}(z_{0}, z_{1}, \ldots , z_{n-1}) = \mathcal{F}(z_{n}, z_{n-1}, \ldots , z_{1}) .
(c)

Show that F4=I\mathcal{F}^{4} = I.

Remark.
?

The discrete Fourier transform has many important applications in data analysis. The usual Fourier transform involves expressions of the form ∫−∞∞f(x)e−2πitxdx\int_{-\infty }^{\infty } f(x) e^{-2 \pi i t x} d x for complex-valued integrable functions ff defined on R\mathbf{R}.

Problem 7D.20

Suppose AA is a square matrix with linearly independent columns. Prove that there exist unique matrices RR and QQ such that RR is lower triangular with only positive numbers on its diagonal, QQ is unitary, and A=RQA = R Q.

?