Scribd
Upload
42 views

Introduction To Quantum Computing

This document is an exercise sheet for a linear algebra primer on quantum computing. It contains 4 exercises on topics such as scalar products, unitary matrices, eigenvalues, diagonalization, trace, and positive semidefinite matrices. The exercises provide definitions, properties, and proof questions relating to these foundational linear algebra concepts for quantum computing.

Uploaded by

Peter Daniel Johannsen
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
42 views

Introduction To Quantum Computing

This document is an exercise sheet for a linear algebra primer on quantum computing. It contains 4 exercises on topics such as scalar products, unitary matrices, eigenvalues, diagonalization, trace, and positive semidefinite matrices. The exercises provide definitions, properties, and proof questions relating to these foundational linear algebra concepts for quantum computing.

Uploaded by

Peter Daniel Johannsen
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Introduction to Quantum

Computing
Exercise sheet 1: linear algebra primer
University of Copenhagen

Exercise 1: (Scalar product and unitary matrices) Recall that given two vectors
C
v, w ∈ d their scalar product hv|wi is defined as:
d
X
hv|wi = v̄i wi .
i=1

Also recall that the norm of the vector v, kvk, is defined as kvk2 = hv|vi. For a matrix
C
A ∈ d1 ×d2 , we define its adjoint matrix A∗ to be given by the d2 × d1 matrix with
C
entries (A∗ )i,j = Āj,i . A matrix U ∈ d×d is called unitary if U ∗ U = I. where I is the
identity matrix.

(a) Let
   
1 2 − 2i
v =  1 + i , w =  (1 − i)−1  .
π
2ei 4 −5

Compute hv|wi.

(b) Show the polarization identity:


1
ku + vk2 − ku − vk2 + iku + ivk2 − iku − ivk2 .

hu, vi =
4

(c) Let A ∈ Cd ×d . Show that for all vectors v ∈ Cd , w ∈ Cd


1 2 1 2 we have:

hv|Awi = hA∗ v|wi.

(d) Show that the matrix:

R, |a|2 + |b|2 = 1
 
a b
U= , ϕ∈
−eiϕ b̄ eiϕ ā

is unitary.

(e) Show the equivalence of the following statements:


I. U is unitary.
II. U preserves the scalar product: hU v|U wi = hv|wi for all v, w ∈ Cd.
III. U preserves the norm: kU vk2 = kvk2 for all v, w ∈ Cd.
Hint: recall that Aij = hei |Aej i for ei the canonical basis vectors. Exercise (b) may
also come in handy.

Exercise 2: (Eigenvalues and diagonalization) Recall that λ ∈ C is called an


C C
eigenvalue of a matrix A ∈ d×d if there exists v ∈ d \{0} such that:

Av = λv.

The vector v is called an eigenvector. The matrix A is called diagonalizable if there exist
a diagonal matrix D and invertible matrix X such that:

A = XDX −1 .

Moreover, the diagonal of D consists of the eigenvalues of A. Also note that if A is


normal (AA∗ = A∗ A), then we can pick X unitary.

(a) Diagonalize the matrix


 
2 1
A= ,
1 2

i.e. find X and D diagonal such that A = XDX −1 . Can you pick X unitary?
Compute A42 .

(b) A is called hermitian if A∗ = A. Show that if A is hermitian and λ one of its


eigenvalues, then λ ∈ . R
(c) Show that if a matrix A has an eigevalue λ, then so does B = XAX −1 . What is the
corresponding eigenvector for B?

(d) Show that unitary matrices are normal.

(e) Show that eigenvalues of unitary matrices are complex number of modulus 1.

Exercise 3: (Trace) Recall that the trace of a matrix A ∈ Cd×d, tr (A), is defined as
d
X
tr (A) = Aii .
i=1

One of the main properties of the trace is its ciclicity:

tr (ABC) = tr (BCA)

for all A, B, C ∈ Cd×d.

2
(a) Let A ∈ Cd×d be Hermitian. Show that
n
X
tr (A) = λi ,
i=1

where λi are the eigenvalues of A.


Hint: diagonalization.
Remark: the statement also holds for nonhermitian matrices.

(b) Show that for all vectors v, w ∈ Cd and A ∈ Cd:


hv|Awi = tr (Avw∗ ) .

(c) Show that the map

(A, B) 7→ tr (A∗ B)

for A, B ∈ Cd×d defines a scalar product on the vector space of matrices.


Exercise 4: (Positive semidefinite matrices) Recall that a hermitian matrix A ∈
Cd×d is called positive semidefinite if for all x ∈ C
d we have:

hx|Axi ≥ 0.

C
A hermitian matrix ρ ∈ d×d is called a density matrix if in addition to it being positive
semidefinite we have tr (ρ) = 1.

(a) Show that A being positive semidefinite is equivalent to A only having positive
eigenvalues.

(b) Let A, B ∈ Cd×d be positive semidefinite. Show that tr (AB) ≥ 0.


(c) Let P1 , . . . , Pm be positive semidefinite operators such that
m
X
Pi = I,
i=1

where I is the identity matrix. Given a density matrix ρ, define the function p :
R
{1, . . . , m} → by p(i) = tr (Pi ρ). Show that p is probability distribution, that is:
m
X
∀i : p(i) ≥ 0 and p(i) = 1.
i=1

You might also like