A convex combination of unitary transforms converts any matrix to identity Announcing the arrival of Valued Associate #679: Cesar Manara Planned maintenance scheduled April 17/18, 2019 at 00:00UTC (8:00pm US/Eastern)Is there always a complete, orthogonal set of unitary matrices?Why does $d_alpha$ divide $#G$ for $alphainhatG$?A general element of U(2)Derivative with Respect to a Variable Pauli MatrixUnitary matrix for matrix representationIs there any interesting relationship between a Hermitian matrix and its corresponding entrywise absolute?Constructing element of the Weyl GroupMeasure to compare a matrix with a given unitary matrixTrace of product of three Pauli matricesImplications of unitary matrix with sub-blocks of matricesFind the maximum of the set of square moduli of eigenvalues of a family of matrices
Why one of virtual NICs called bond0?
Did Kevin spill real chili?
G-Code for resetting to 100% speed
Should I call the interviewer directly, if HR aren't responding?
What causes the vertical darker bands in my photo?
Can a non-EU citizen traveling with me come with me through the EU passport line?
What's the purpose of writing one's academic bio in 3rd person?
Why is "Consequences inflicted." not a sentence?
Does surprise arrest existing movement?
Why don't the Weasley twins use magic outside of school if the Trace can only find the location of spells cast?
"Seemed to had" is it correct?
Gastric acid as a weapon
Is there a service that would inform me whenever a new direct route is scheduled from a given airport?
Is above average number of years spent on PhD considered a red flag in future academia or industry positions?
How discoverable are IPv6 addresses and AAAA names by potential attackers?
What would be the ideal power source for a cybernetic eye?
Models of set theory where not every set can be linearly ordered
If 'B is more likely given A', then 'A is more likely given B'
How to motivate offshore teams and trust them to deliver?
Does accepting a pardon have any bearing on trying that person for the same crime in a sovereign jurisdiction?
When is phishing education going too far?
What is the correct way to use the pinch test for dehydration?
How do I stop a creek from eroding my steep embankment?
What is a Meta algorithm?
A convex combination of unitary transforms converts any matrix to identity
Announcing the arrival of Valued Associate #679: Cesar Manara
Planned maintenance scheduled April 17/18, 2019 at 00:00UTC (8:00pm US/Eastern)Is there always a complete, orthogonal set of unitary matrices?Why does $d_alpha$ divide $#G$ for $alphainhatG$?A general element of U(2)Derivative with Respect to a Variable Pauli MatrixUnitary matrix for matrix representationIs there any interesting relationship between a Hermitian matrix and its corresponding entrywise absolute?Constructing element of the Weyl GroupMeasure to compare a matrix with a given unitary matrixTrace of product of three Pauli matricesImplications of unitary matrix with sub-blocks of matricesFind the maximum of the set of square moduli of eigenvalues of a family of matrices
$begingroup$
Question
Show that there exists a set of unitary matrices $U_i$, and probability $p_i$, such that for any $n times n$ matrix $A$
beginequation
tag1
sum_i p_i U_i A U^dagger_i = texttr(A) fracIn
endequation
Attempts
For $n=2$, it is easy to show
beginequation
frac14 ( sigma^x A sigma^x + sigma^y A sigma^y + sigma^z A sigma^z + I A I ) = texttr(A) I / 2
endequation
where $sigma^x,y,z$ are Pauli sigma matrices. The idea comes from kraus operator sum representation.
We can then generalize to dimension $n = 2^m$, where $U_i$ can be taken as the tensor products of these basis, but not arbitrary dimension.
In indices, Eq.(1) is equivalent to
beginequation
sum_i p_i (U_i)_ab (U_i^*)_dc = delta_bc delta_ad / n
endequation
This looks like the identity from the finite dimensional irreducible unitary representation of finite group, see Peter-Weyl theorem. But again this only works when group $G$ has irreducible representation at dimension $n$, and all $p_i$ in this case are equal.
I feel that "right proof" should not utilize these additional structures.
linear-algebra matrix-decomposition
asked Mar 20 at 17:49
anecdoteanecdote
508313
$endgroup$
add a comment |
$begingroup$
Question
Show that there exists a set of unitary matrices $U_i$, and probability $p_i$, such that for any $n times n$ matrix $A$
beginequation
tag1
sum_i p_i U_i A U^dagger_i = texttr(A) fracIn
endequation
Attempts
For $n=2$, it is easy to show
beginequation
frac14 ( sigma^x A sigma^x + sigma^y A sigma^y + sigma^z A sigma^z + I A I ) = texttr(A) I / 2
endequation
where $sigma^x,y,z$ are Pauli sigma matrices. The idea comes from kraus operator sum representation.
We can then generalize to dimension $n = 2^m$, where $U_i$ can be taken as the tensor products of these basis, but not arbitrary dimension.
In indices, Eq.(1) is equivalent to
beginequation
sum_i p_i (U_i)_ab (U_i^*)_dc = delta_bc delta_ad / n
endequation
This looks like the identity from the finite dimensional irreducible unitary representation of finite group, see Peter-Weyl theorem. But again this only works when group $G$ has irreducible representation at dimension $n$, and all $p_i$ in this case are equal.
I feel that "right proof" should not utilize these additional structures.
linear-algebra matrix-decomposition
asked Mar 20 at 17:49
anecdoteanecdote
508313
$endgroup$
$begingroup$
It may help to observe that the Choi matrix of your operator is given by $$ C_Phi = frac 1n I_n otimes I_n $$ since the desired decomposition is a Kraus decomposition, the matrices $U_i$ must satisfy $$ sum_ip_i operatornamevec(U_i)operatornamevec(U_i)^dagger = C_Phi = frac 1n I_n otimes I_n $$
$endgroup$
– Omnomnomnom
Mar 20 at 19:00
$begingroup$
An educated guess: I think that $U_i = W^i$ should work, where $W$ is the $n times n$ DFT matrix, with all $p_i = frac 1n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:07
$begingroup$
Never mind, my guess fails. The $U_i$ must span the set of all $n times n$ matrices, which my guess fails to do. Notably, the $U_i$ cannot be simultaneously diagonalizable.
$endgroup$
– Omnomnomnom
Mar 20 at 19:10
$begingroup$
Perhaps it helps though to know that you're looking for a set of unitary matrices $U_i$ whose span is $Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:31
$begingroup$
@Omnomnomnom, yes problem solved if you find a basis for matrix.
$endgroup$
– anecdote
Mar 21 at 20:56
add a comment |
$begingroup$
Question
Show that there exists a set of unitary matrices $U_i$, and probability $p_i$, such that for any $n times n$ matrix $A$
beginequation
tag1
sum_i p_i U_i A U^dagger_i = texttr(A) fracIn
endequation
Attempts
For $n=2$, it is easy to show
beginequation
frac14 ( sigma^x A sigma^x + sigma^y A sigma^y + sigma^z A sigma^z + I A I ) = texttr(A) I / 2
endequation
where $sigma^x,y,z$ are Pauli sigma matrices. The idea comes from kraus operator sum representation.
We can then generalize to dimension $n = 2^m$, where $U_i$ can be taken as the tensor products of these basis, but not arbitrary dimension.
In indices, Eq.(1) is equivalent to
beginequation
sum_i p_i (U_i)_ab (U_i^*)_dc = delta_bc delta_ad / n
endequation
This looks like the identity from the finite dimensional irreducible unitary representation of finite group, see Peter-Weyl theorem. But again this only works when group $G$ has irreducible representation at dimension $n$, and all $p_i$ in this case are equal.
I feel that "right proof" should not utilize these additional structures.
linear-algebra matrix-decomposition
asked Mar 20 at 17:49
anecdoteanecdote
508313
$endgroup$
Question
Show that there exists a set of unitary matrices $U_i$, and probability $p_i$, such that for any $n times n$ matrix $A$
beginequation
tag1
sum_i p_i U_i A U^dagger_i = texttr(A) fracIn
endequation
Attempts
For $n=2$, it is easy to show
beginequation
frac14 ( sigma^x A sigma^x + sigma^y A sigma^y + sigma^z A sigma^z + I A I ) = texttr(A) I / 2
endequation
where $sigma^x,y,z$ are Pauli sigma matrices. The idea comes from kraus operator sum representation.
We can then generalize to dimension $n = 2^m$, where $U_i$ can be taken as the tensor products of these basis, but not arbitrary dimension.
In indices, Eq.(1) is equivalent to
beginequation
sum_i p_i (U_i)_ab (U_i^*)_dc = delta_bc delta_ad / n
endequation
This looks like the identity from the finite dimensional irreducible unitary representation of finite group, see Peter-Weyl theorem. But again this only works when group $G$ has irreducible representation at dimension $n$, and all $p_i$ in this case are equal.
I feel that "right proof" should not utilize these additional structures.
linear-algebra matrix-decomposition
linear-algebra matrix-decomposition
asked Mar 20 at 17:49
anecdoteanecdote
508313
asked Mar 20 at 17:49
anecdoteanecdote
508313
asked Mar 20 at 17:49
anecdoteanecdote
508313
asked Mar 20 at 17:49
anecdoteanecdote
508313
asked Mar 20 at 17:49
anecdoteanecdote
508313
508313
$begingroup$
It may help to observe that the Choi matrix of your operator is given by $$ C_Phi = frac 1n I_n otimes I_n $$ since the desired decomposition is a Kraus decomposition, the matrices $U_i$ must satisfy $$ sum_ip_i operatornamevec(U_i)operatornamevec(U_i)^dagger = C_Phi = frac 1n I_n otimes I_n $$
$endgroup$
– Omnomnomnom
Mar 20 at 19:00
$begingroup$
An educated guess: I think that $U_i = W^i$ should work, where $W$ is the $n times n$ DFT matrix, with all $p_i = frac 1n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:07
$begingroup$
Never mind, my guess fails. The $U_i$ must span the set of all $n times n$ matrices, which my guess fails to do. Notably, the $U_i$ cannot be simultaneously diagonalizable.
$endgroup$
– Omnomnomnom
Mar 20 at 19:10
$begingroup$
Perhaps it helps though to know that you're looking for a set of unitary matrices $U_i$ whose span is $Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:31
$begingroup$
@Omnomnomnom, yes problem solved if you find a basis for matrix.
$endgroup$
– anecdote
Mar 21 at 20:56
add a comment |
$begingroup$
It may help to observe that the Choi matrix of your operator is given by $$ C_Phi = frac 1n I_n otimes I_n $$ since the desired decomposition is a Kraus decomposition, the matrices $U_i$ must satisfy $$ sum_ip_i operatornamevec(U_i)operatornamevec(U_i)^dagger = C_Phi = frac 1n I_n otimes I_n $$
$endgroup$
– Omnomnomnom
Mar 20 at 19:00
$begingroup$
An educated guess: I think that $U_i = W^i$ should work, where $W$ is the $n times n$ DFT matrix, with all $p_i = frac 1n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:07
$begingroup$
Never mind, my guess fails. The $U_i$ must span the set of all $n times n$ matrices, which my guess fails to do. Notably, the $U_i$ cannot be simultaneously diagonalizable.
$endgroup$
– Omnomnomnom
Mar 20 at 19:10
$begingroup$
Perhaps it helps though to know that you're looking for a set of unitary matrices $U_i$ whose span is $Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:31
$begingroup$
@Omnomnomnom, yes problem solved if you find a basis for matrix.
$endgroup$
– anecdote
Mar 21 at 20:56
$begingroup$
It may help to observe that the Choi matrix of your operator is given by $$ C_Phi = frac 1n I_n otimes I_n $$ since the desired decomposition is a Kraus decomposition, the matrices $U_i$ must satisfy $$ sum_ip_i operatornamevec(U_i)operatornamevec(U_i)^dagger = C_Phi = frac 1n I_n otimes I_n $$
$endgroup$
– Omnomnomnom
Mar 20 at 19:00
$begingroup$
It may help to observe that the Choi matrix of your operator is given by $$ C_Phi = frac 1n I_n otimes I_n $$ since the desired decomposition is a Kraus decomposition, the matrices $U_i$ must satisfy $$ sum_ip_i operatornamevec(U_i)operatornamevec(U_i)^dagger = C_Phi = frac 1n I_n otimes I_n $$
$endgroup$
– Omnomnomnom
Mar 20 at 19:00
$begingroup$
An educated guess: I think that $U_i = W^i$ should work, where $W$ is the $n times n$ DFT matrix, with all $p_i = frac 1n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:07
$begingroup$
An educated guess: I think that $U_i = W^i$ should work, where $W$ is the $n times n$ DFT matrix, with all $p_i = frac 1n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:07
$begingroup$
Never mind, my guess fails. The $U_i$ must span the set of all $n times n$ matrices, which my guess fails to do. Notably, the $U_i$ cannot be simultaneously diagonalizable.
$endgroup$
– Omnomnomnom
Mar 20 at 19:10
$begingroup$
Never mind, my guess fails. The $U_i$ must span the set of all $n times n$ matrices, which my guess fails to do. Notably, the $U_i$ cannot be simultaneously diagonalizable.
$endgroup$
– Omnomnomnom
Mar 20 at 19:10
$begingroup$
Perhaps it helps though to know that you're looking for a set of unitary matrices $U_i$ whose span is $Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:31
$begingroup$
Perhaps it helps though to know that you're looking for a set of unitary matrices $U_i$ whose span is $Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:31
$begingroup$
@Omnomnomnom, yes problem solved if you find a basis for matrix.
$endgroup$
– anecdote
Mar 21 at 20:56
$begingroup$
@Omnomnomnom, yes problem solved if you find a basis for matrix.
$endgroup$
– anecdote
Mar 21 at 20:56
add a comment |
1 Answer
1
active
oldest
votes
$begingroup$
An attempted proof of existence that doesn't actually construct the spanning set and distribution.
First, we note that the set of unitary matrices spans $Bbb C^n times n$; we could prove this nicely using polar decomposition. From there, we note that there must exist a basis of $Bbb C^n times n$ $U_1,U_2,dots,U_n^2$ consisting of unitary matrices.
It follows that the vectors $operatornamevec(U_1),dots,operatornamevec(U_n^2)$ span $Bbb C^n^2$.
The argument below is incorrect
(Thus, there necessarily exist (positive) $p_k$ such that
$$
frac 1n I_n^2 = sum_i p_i operatornamevec(U_i)operatornamevec(U_i)^dagger
$$
We correspondingly find that these $U_i$ satisfy $sum_i p_i U_iA U_i^dagger = frac 1n operatornametr(A) I$, as desired.)
Some clarification:
First of all, the linear span bit. Let $langle cdot, cdot rangle$ denote the Frobenius (Hilbert-Schmidt) inner product. Suppose that $A$ lies in the orthogonal complement of the span of the unitary matrices. Let $A = UP$ be a polar decomposition. Then we have
$$
0 = langle U, A rangle = operatornametrace(U^dagger A) = operatornametrace(U^dagger UP) = operatornametrace(P)
$$
but $P$ is positive semidefinite, so $operatornametrace(P) = 0$ implies that $P = 0$. Thus, $A$ must be zero.
So, the span of the unitary matrices is all $Bbb C^n times n$.
Another obsrevation:
Let $mathcal C_U$ denote the convex cone generated by the set $uu^* : u = operatornamevec(U) text for some unitary U $. Showing that $sum_i p_i operatornamevec(U_i)operatornamevec(U_i) = I$ can be achieved with non-negative coefficients $p_i$ means that we're trying to show that $I in mathcal C_U$.
One orthogonal basis for $Bbb C^n times n$ consisting of unitary matrices is as follows: let
$$
X = pmatrix0&&&&1\1&0\&1&0\&&ddots\&&&1, Z = pmatrix1\ & omega \ && ddots \ &&& omega^n-1
$$
Then the matrices $Z^j X^k : 0 leq j,k leq n-1$ form our orthogonal basis.
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
$endgroup$
$begingroup$
I am a bit slow here... Can you show how to get the linear span from the polar decomposition?
$endgroup$
– anecdote
Mar 22 at 3:17
$begingroup$
About the resolution identity: let me write $textvec(U_i)$ as $v_i$. Now $ v_i $ is a set of basis, if $I = n sum_i p_i v_i v_i^dagger$, then $G_ij = n G_ik p_k G_kj$ where $G_ij$ is the Gram matrix. Then how do we show that $p_k $ exist? I thought $v_i$ here are not necessarily orthogonal.
$endgroup$
– anecdote
Mar 22 at 3:21
$begingroup$
I clarified the polar decomposition argument. It seems clear that I'm wrong about that second argument though. Nevertheless, I have a feeling that the fact that the unitaries "are a large enough set" will somehow be enough here.
$endgroup$
– Omnomnomnom
Mar 22 at 13:46
$begingroup$
Following the argument in proposition 2.2 of this paper, I can show that there exist real coefficients $a_i$ and unitary matrices $U_i$ such that $$ sum_i a_i operatornamevec(U_i)operatornamevec(U_i)^dagger = frac 1n I $$ That being said, what we would need is a way to guarantee that the coefficients $a_i$ are non-negative, which I haven't been able to come up with.
$endgroup$
– Omnomnomnom
Mar 22 at 14:00
1
$begingroup$
@anecdote Let me rephrase the argument: define $W subseteq Bbb C^n times n$ to be the subspace spanned by the unitary matrices. What I show is that if $A in W^perp$, then $A = 0$. It follows that $W = Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 22 at 16:19
|
show 7 more comments
Your Answer
StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "69"
;
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function()
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled)
StackExchange.using("snippets", function()
createEditor();
);
else
createEditor();
);
function createEditor()
StackExchange.prepareEditor(
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader:
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
,
noCode: true, onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
);
);
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3155780%2fa-convex-combination-of-unitary-transforms-converts-any-matrix-to-identity%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
An attempted proof of existence that doesn't actually construct the spanning set and distribution.
First, we note that the set of unitary matrices spans $Bbb C^n times n$; we could prove this nicely using polar decomposition. From there, we note that there must exist a basis of $Bbb C^n times n$ $U_1,U_2,dots,U_n^2$ consisting of unitary matrices.
It follows that the vectors $operatornamevec(U_1),dots,operatornamevec(U_n^2)$ span $Bbb C^n^2$.
The argument below is incorrect
(Thus, there necessarily exist (positive) $p_k$ such that
$$
frac 1n I_n^2 = sum_i p_i operatornamevec(U_i)operatornamevec(U_i)^dagger
$$
We correspondingly find that these $U_i$ satisfy $sum_i p_i U_iA U_i^dagger = frac 1n operatornametr(A) I$, as desired.)
Some clarification:
First of all, the linear span bit. Let $langle cdot, cdot rangle$ denote the Frobenius (Hilbert-Schmidt) inner product. Suppose that $A$ lies in the orthogonal complement of the span of the unitary matrices. Let $A = UP$ be a polar decomposition. Then we have
$$
0 = langle U, A rangle = operatornametrace(U^dagger A) = operatornametrace(U^dagger UP) = operatornametrace(P)
$$
but $P$ is positive semidefinite, so $operatornametrace(P) = 0$ implies that $P = 0$. Thus, $A$ must be zero.
So, the span of the unitary matrices is all $Bbb C^n times n$.
Another obsrevation:
Let $mathcal C_U$ denote the convex cone generated by the set $uu^* : u = operatornamevec(U) text for some unitary U $. Showing that $sum_i p_i operatornamevec(U_i)operatornamevec(U_i) = I$ can be achieved with non-negative coefficients $p_i$ means that we're trying to show that $I in mathcal C_U$.
One orthogonal basis for $Bbb C^n times n$ consisting of unitary matrices is as follows: let
$$
X = pmatrix0&&&&1\1&0\&1&0\&&ddots\&&&1, Z = pmatrix1\ & omega \ && ddots \ &&& omega^n-1
$$
Then the matrices $Z^j X^k : 0 leq j,k leq n-1$ form our orthogonal basis.
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
$endgroup$
$begingroup$
I am a bit slow here... Can you show how to get the linear span from the polar decomposition?
$endgroup$
– anecdote
Mar 22 at 3:17
$begingroup$
About the resolution identity: let me write $textvec(U_i)$ as $v_i$. Now $ v_i $ is a set of basis, if $I = n sum_i p_i v_i v_i^dagger$, then $G_ij = n G_ik p_k G_kj$ where $G_ij$ is the Gram matrix. Then how do we show that $p_k $ exist? I thought $v_i$ here are not necessarily orthogonal.
$endgroup$
– anecdote
Mar 22 at 3:21
$begingroup$
I clarified the polar decomposition argument. It seems clear that I'm wrong about that second argument though. Nevertheless, I have a feeling that the fact that the unitaries "are a large enough set" will somehow be enough here.
$endgroup$
– Omnomnomnom
Mar 22 at 13:46
$begingroup$
Following the argument in proposition 2.2 of this paper, I can show that there exist real coefficients $a_i$ and unitary matrices $U_i$ such that $$ sum_i a_i operatornamevec(U_i)operatornamevec(U_i)^dagger = frac 1n I $$ That being said, what we would need is a way to guarantee that the coefficients $a_i$ are non-negative, which I haven't been able to come up with.
$endgroup$
– Omnomnomnom
Mar 22 at 14:00
1
$begingroup$
@anecdote Let me rephrase the argument: define $W subseteq Bbb C^n times n$ to be the subspace spanned by the unitary matrices. What I show is that if $A in W^perp$, then $A = 0$. It follows that $W = Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 22 at 16:19
|
show 7 more comments
$begingroup$
An attempted proof of existence that doesn't actually construct the spanning set and distribution.
First, we note that the set of unitary matrices spans $Bbb C^n times n$; we could prove this nicely using polar decomposition. From there, we note that there must exist a basis of $Bbb C^n times n$ $U_1,U_2,dots,U_n^2$ consisting of unitary matrices.
It follows that the vectors $operatornamevec(U_1),dots,operatornamevec(U_n^2)$ span $Bbb C^n^2$.
The argument below is incorrect
(Thus, there necessarily exist (positive) $p_k$ such that
$$
frac 1n I_n^2 = sum_i p_i operatornamevec(U_i)operatornamevec(U_i)^dagger
$$
We correspondingly find that these $U_i$ satisfy $sum_i p_i U_iA U_i^dagger = frac 1n operatornametr(A) I$, as desired.)
Some clarification:
First of all, the linear span bit. Let $langle cdot, cdot rangle$ denote the Frobenius (Hilbert-Schmidt) inner product. Suppose that $A$ lies in the orthogonal complement of the span of the unitary matrices. Let $A = UP$ be a polar decomposition. Then we have
$$
0 = langle U, A rangle = operatornametrace(U^dagger A) = operatornametrace(U^dagger UP) = operatornametrace(P)
$$
but $P$ is positive semidefinite, so $operatornametrace(P) = 0$ implies that $P = 0$. Thus, $A$ must be zero.
So, the span of the unitary matrices is all $Bbb C^n times n$.
Another obsrevation:
Let $mathcal C_U$ denote the convex cone generated by the set $uu^* : u = operatornamevec(U) text for some unitary U $. Showing that $sum_i p_i operatornamevec(U_i)operatornamevec(U_i) = I$ can be achieved with non-negative coefficients $p_i$ means that we're trying to show that $I in mathcal C_U$.
One orthogonal basis for $Bbb C^n times n$ consisting of unitary matrices is as follows: let
$$
X = pmatrix0&&&&1\1&0\&1&0\&&ddots\&&&1, Z = pmatrix1\ & omega \ && ddots \ &&& omega^n-1
$$
Then the matrices $Z^j X^k : 0 leq j,k leq n-1$ form our orthogonal basis.
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
$endgroup$
$begingroup$
I am a bit slow here... Can you show how to get the linear span from the polar decomposition?
$endgroup$
– anecdote
Mar 22 at 3:17
$begingroup$
About the resolution identity: let me write $textvec(U_i)$ as $v_i$. Now $ v_i $ is a set of basis, if $I = n sum_i p_i v_i v_i^dagger$, then $G_ij = n G_ik p_k G_kj$ where $G_ij$ is the Gram matrix. Then how do we show that $p_k $ exist? I thought $v_i$ here are not necessarily orthogonal.
$endgroup$
– anecdote
Mar 22 at 3:21
$begingroup$
I clarified the polar decomposition argument. It seems clear that I'm wrong about that second argument though. Nevertheless, I have a feeling that the fact that the unitaries "are a large enough set" will somehow be enough here.
$endgroup$
– Omnomnomnom
Mar 22 at 13:46
$begingroup$
Following the argument in proposition 2.2 of this paper, I can show that there exist real coefficients $a_i$ and unitary matrices $U_i$ such that $$ sum_i a_i operatornamevec(U_i)operatornamevec(U_i)^dagger = frac 1n I $$ That being said, what we would need is a way to guarantee that the coefficients $a_i$ are non-negative, which I haven't been able to come up with.
$endgroup$
– Omnomnomnom
Mar 22 at 14:00
1
$begingroup$
@anecdote Let me rephrase the argument: define $W subseteq Bbb C^n times n$ to be the subspace spanned by the unitary matrices. What I show is that if $A in W^perp$, then $A = 0$. It follows that $W = Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 22 at 16:19
|
show 7 more comments
$begingroup$
An attempted proof of existence that doesn't actually construct the spanning set and distribution.
First, we note that the set of unitary matrices spans $Bbb C^n times n$; we could prove this nicely using polar decomposition. From there, we note that there must exist a basis of $Bbb C^n times n$ $U_1,U_2,dots,U_n^2$ consisting of unitary matrices.
It follows that the vectors $operatornamevec(U_1),dots,operatornamevec(U_n^2)$ span $Bbb C^n^2$.
The argument below is incorrect
(Thus, there necessarily exist (positive) $p_k$ such that
$$
frac 1n I_n^2 = sum_i p_i operatornamevec(U_i)operatornamevec(U_i)^dagger
$$
We correspondingly find that these $U_i$ satisfy $sum_i p_i U_iA U_i^dagger = frac 1n operatornametr(A) I$, as desired.)
Some clarification:
First of all, the linear span bit. Let $langle cdot, cdot rangle$ denote the Frobenius (Hilbert-Schmidt) inner product. Suppose that $A$ lies in the orthogonal complement of the span of the unitary matrices. Let $A = UP$ be a polar decomposition. Then we have
$$
0 = langle U, A rangle = operatornametrace(U^dagger A) = operatornametrace(U^dagger UP) = operatornametrace(P)
$$
but $P$ is positive semidefinite, so $operatornametrace(P) = 0$ implies that $P = 0$. Thus, $A$ must be zero.
So, the span of the unitary matrices is all $Bbb C^n times n$.
Another obsrevation:
Let $mathcal C_U$ denote the convex cone generated by the set $uu^* : u = operatornamevec(U) text for some unitary U $. Showing that $sum_i p_i operatornamevec(U_i)operatornamevec(U_i) = I$ can be achieved with non-negative coefficients $p_i$ means that we're trying to show that $I in mathcal C_U$.
One orthogonal basis for $Bbb C^n times n$ consisting of unitary matrices is as follows: let
$$
X = pmatrix0&&&&1\1&0\&1&0\&&ddots\&&&1, Z = pmatrix1\ & omega \ && ddots \ &&& omega^n-1
$$
Then the matrices $Z^j X^k : 0 leq j,k leq n-1$ form our orthogonal basis.
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
$endgroup$
An attempted proof of existence that doesn't actually construct the spanning set and distribution.
First, we note that the set of unitary matrices spans $Bbb C^n times n$; we could prove this nicely using polar decomposition. From there, we note that there must exist a basis of $Bbb C^n times n$ $U_1,U_2,dots,U_n^2$ consisting of unitary matrices.
It follows that the vectors $operatornamevec(U_1),dots,operatornamevec(U_n^2)$ span $Bbb C^n^2$.
The argument below is incorrect
(Thus, there necessarily exist (positive) $p_k$ such that
$$
frac 1n I_n^2 = sum_i p_i operatornamevec(U_i)operatornamevec(U_i)^dagger
$$
We correspondingly find that these $U_i$ satisfy $sum_i p_i U_iA U_i^dagger = frac 1n operatornametr(A) I$, as desired.)
Some clarification:
First of all, the linear span bit. Let $langle cdot, cdot rangle$ denote the Frobenius (Hilbert-Schmidt) inner product. Suppose that $A$ lies in the orthogonal complement of the span of the unitary matrices. Let $A = UP$ be a polar decomposition. Then we have
$$
0 = langle U, A rangle = operatornametrace(U^dagger A) = operatornametrace(U^dagger UP) = operatornametrace(P)
$$
but $P$ is positive semidefinite, so $operatornametrace(P) = 0$ implies that $P = 0$. Thus, $A$ must be zero.
So, the span of the unitary matrices is all $Bbb C^n times n$.
Another obsrevation:
Let $mathcal C_U$ denote the convex cone generated by the set $uu^* : u = operatornamevec(U) text for some unitary U $. Showing that $sum_i p_i operatornamevec(U_i)operatornamevec(U_i) = I$ can be achieved with non-negative coefficients $p_i$ means that we're trying to show that $I in mathcal C_U$.
One orthogonal basis for $Bbb C^n times n$ consisting of unitary matrices is as follows: let
$$
X = pmatrix0&&&&1\1&0\&1&0\&&ddots\&&&1, Z = pmatrix1\ & omega \ && ddots \ &&& omega^n-1
$$
Then the matrices $Z^j X^k : 0 leq j,k leq n-1$ form our orthogonal basis.
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
edited Apr 1 at 0:28
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
answered Mar 21 at 22:33
OmnomnomnomOmnomnomnom
129k794188
129k794188
$begingroup$
I am a bit slow here... Can you show how to get the linear span from the polar decomposition?
$endgroup$
– anecdote
Mar 22 at 3:17
$begingroup$
About the resolution identity: let me write $textvec(U_i)$ as $v_i$. Now $ v_i $ is a set of basis, if $I = n sum_i p_i v_i v_i^dagger$, then $G_ij = n G_ik p_k G_kj$ where $G_ij$ is the Gram matrix. Then how do we show that $p_k $ exist? I thought $v_i$ here are not necessarily orthogonal.
$endgroup$
– anecdote
Mar 22 at 3:21
$begingroup$
I clarified the polar decomposition argument. It seems clear that I'm wrong about that second argument though. Nevertheless, I have a feeling that the fact that the unitaries "are a large enough set" will somehow be enough here.
$endgroup$
– Omnomnomnom
Mar 22 at 13:46
$begingroup$
Following the argument in proposition 2.2 of this paper, I can show that there exist real coefficients $a_i$ and unitary matrices $U_i$ such that $$ sum_i a_i operatornamevec(U_i)operatornamevec(U_i)^dagger = frac 1n I $$ That being said, what we would need is a way to guarantee that the coefficients $a_i$ are non-negative, which I haven't been able to come up with.
$endgroup$
– Omnomnomnom
Mar 22 at 14:00
1
$begingroup$
@anecdote Let me rephrase the argument: define $W subseteq Bbb C^n times n$ to be the subspace spanned by the unitary matrices. What I show is that if $A in W^perp$, then $A = 0$. It follows that $W = Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 22 at 16:19
|
show 7 more comments
$begingroup$
I am a bit slow here... Can you show how to get the linear span from the polar decomposition?
$endgroup$
– anecdote
Mar 22 at 3:17
$begingroup$
About the resolution identity: let me write $textvec(U_i)$ as $v_i$. Now $ v_i $ is a set of basis, if $I = n sum_i p_i v_i v_i^dagger$, then $G_ij = n G_ik p_k G_kj$ where $G_ij$ is the Gram matrix. Then how do we show that $p_k $ exist? I thought $v_i$ here are not necessarily orthogonal.
$endgroup$
– anecdote
Mar 22 at 3:21
$begingroup$
I clarified the polar decomposition argument. It seems clear that I'm wrong about that second argument though. Nevertheless, I have a feeling that the fact that the unitaries "are a large enough set" will somehow be enough here.
$endgroup$
– Omnomnomnom
Mar 22 at 13:46
$begingroup$
Following the argument in proposition 2.2 of this paper, I can show that there exist real coefficients $a_i$ and unitary matrices $U_i$ such that $$ sum_i a_i operatornamevec(U_i)operatornamevec(U_i)^dagger = frac 1n I $$ That being said, what we would need is a way to guarantee that the coefficients $a_i$ are non-negative, which I haven't been able to come up with.
$endgroup$
– Omnomnomnom
Mar 22 at 14:00
1
$begingroup$
@anecdote Let me rephrase the argument: define $W subseteq Bbb C^n times n$ to be the subspace spanned by the unitary matrices. What I show is that if $A in W^perp$, then $A = 0$. It follows that $W = Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 22 at 16:19
$begingroup$
I am a bit slow here... Can you show how to get the linear span from the polar decomposition?
$endgroup$
– anecdote
Mar 22 at 3:17
$begingroup$
I am a bit slow here... Can you show how to get the linear span from the polar decomposition?
$endgroup$
– anecdote
Mar 22 at 3:17
$begingroup$
About the resolution identity: let me write $textvec(U_i)$ as $v_i$. Now $ v_i $ is a set of basis, if $I = n sum_i p_i v_i v_i^dagger$, then $G_ij = n G_ik p_k G_kj$ where $G_ij$ is the Gram matrix. Then how do we show that $p_k $ exist? I thought $v_i$ here are not necessarily orthogonal.
$endgroup$
– anecdote
Mar 22 at 3:21
$begingroup$
About the resolution identity: let me write $textvec(U_i)$ as $v_i$. Now $ v_i $ is a set of basis, if $I = n sum_i p_i v_i v_i^dagger$, then $G_ij = n G_ik p_k G_kj$ where $G_ij$ is the Gram matrix. Then how do we show that $p_k $ exist? I thought $v_i$ here are not necessarily orthogonal.
$endgroup$
– anecdote
Mar 22 at 3:21
$begingroup$
I clarified the polar decomposition argument. It seems clear that I'm wrong about that second argument though. Nevertheless, I have a feeling that the fact that the unitaries "are a large enough set" will somehow be enough here.
$endgroup$
– Omnomnomnom
Mar 22 at 13:46
$begingroup$
I clarified the polar decomposition argument. It seems clear that I'm wrong about that second argument though. Nevertheless, I have a feeling that the fact that the unitaries "are a large enough set" will somehow be enough here.
$endgroup$
– Omnomnomnom
Mar 22 at 13:46
$begingroup$
Following the argument in proposition 2.2 of this paper, I can show that there exist real coefficients $a_i$ and unitary matrices $U_i$ such that $$ sum_i a_i operatornamevec(U_i)operatornamevec(U_i)^dagger = frac 1n I $$ That being said, what we would need is a way to guarantee that the coefficients $a_i$ are non-negative, which I haven't been able to come up with.
$endgroup$
– Omnomnomnom
Mar 22 at 14:00
$begingroup$
Following the argument in proposition 2.2 of this paper, I can show that there exist real coefficients $a_i$ and unitary matrices $U_i$ such that $$ sum_i a_i operatornamevec(U_i)operatornamevec(U_i)^dagger = frac 1n I $$ That being said, what we would need is a way to guarantee that the coefficients $a_i$ are non-negative, which I haven't been able to come up with.
$endgroup$
– Omnomnomnom
Mar 22 at 14:00
1
1
$begingroup$
@anecdote Let me rephrase the argument: define $W subseteq Bbb C^n times n$ to be the subspace spanned by the unitary matrices. What I show is that if $A in W^perp$, then $A = 0$. It follows that $W = Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 22 at 16:19
$begingroup$
@anecdote Let me rephrase the argument: define $W subseteq Bbb C^n times n$ to be the subspace spanned by the unitary matrices. What I show is that if $A in W^perp$, then $A = 0$. It follows that $W = Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 22 at 16:19
|
show 7 more comments
Thanks for contributing an answer to Mathematics Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3155780%2fa-convex-combination-of-unitary-transforms-converts-any-matrix-to-identity%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
$begingroup$
It may help to observe that the Choi matrix of your operator is given by $$ C_Phi = frac 1n I_n otimes I_n $$ since the desired decomposition is a Kraus decomposition, the matrices $U_i$ must satisfy $$ sum_ip_i operatornamevec(U_i)operatornamevec(U_i)^dagger = C_Phi = frac 1n I_n otimes I_n $$
$endgroup$
– Omnomnomnom
Mar 20 at 19:00
$begingroup$
An educated guess: I think that $U_i = W^i$ should work, where $W$ is the $n times n$ DFT matrix, with all $p_i = frac 1n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:07
$begingroup$
Never mind, my guess fails. The $U_i$ must span the set of all $n times n$ matrices, which my guess fails to do. Notably, the $U_i$ cannot be simultaneously diagonalizable.
$endgroup$
– Omnomnomnom
Mar 20 at 19:10
$begingroup$
Perhaps it helps though to know that you're looking for a set of unitary matrices $U_i$ whose span is $Bbb C^n times n$
$endgroup$
– Omnomnomnom
Mar 20 at 19:31
$begingroup$
@Omnomnomnom, yes problem solved if you find a basis for matrix.
$endgroup$
– anecdote
Mar 21 at 20:56