RGD optimization method for quantum state tomography #1485

HuberyMing · 2024-10-11T15:14:19Z

Checklist:

Reviewers confirm new code works as expected.
Tests are passing.
Coverage does not decrease.
Documentation is updated.

To include the optimization method of Riemannian Gradient Descent (RGD) algorithm to do the tomography problem. The implementation follows from the arXiv:2210.04717, which was published in Phys. Rev. Lett. 132, 240804

for more information, see https://pre-commit.ci

renatomello · 2024-10-14T05:31:41Z

Not entering the other merits of the PR, this would have to be inside the tomography module instead of being a separate tomography_RGB module.
Moreover, tests should follow how tests in qibo are done. Right now they're in a folder that is being ignored by gitignore.

src/qibo/tomography_RGD/BasicTools.py

src/qibo/tomography_RGD/Run_Tomo.py

src/qibo/tomography_RGD/qibo_states.py

alecandido · 2024-10-16T11:21:19Z

@renatomello to follow up what I told you about remotes during the Qibo meeting, you could do the following:

git remote add hubery-ming git@github.com:HuberyMing/qibo.git
git pull hubery-ming
git switch -c QST -t hubery-ming/QST

in this way you would set up a remote named hubery-ming, tracking @HuberyMing's repo, and then create a local branch in your local repo named QST, tracking the https://github.com/HuberyMing/qibo/tree/QST branch.

Of course, you won't have pushing rights (unless @HuberyMing will grant you), but that's in case you want to test it locally.

Maybe you were already familiar with this, but this may also be useful for everyone else who'd like to have a look (and it could be a reference to deal for PRs from forks in general).

update random_density_matrix usage Co-authored-by: Renato Mello <renato.msf@gmail.com>

for more information, see https://pre-commit.ci

renatomello · 2024-10-17T06:54:41Z

src/qibo/tomography_RGD/qibo_states.py

+                        2 * np.pi * random.random(),
+                        2 * np.pi * random.random(),
+                        2 * np.pi * random.random(),


A tip for the future: This choice of phases does not sample random states uniformly. Please see

qibo/src/qibo/quantum_info/random_ensembles.py

Line 48 in 958b11b

def uniform_sampling_U3(ngates: int, seed=None, backend=None):

and https://pennylane.ai/qml/demos/tutorial_haar_measure

renatomello · 2024-10-17T06:58:01Z

src/qibo/tomography_RGD/qibo_states.py

+from qibo import gates, quantum_info
+
+
+class State:


Correct me if I'm wrong, but it seems like this base class is unnecessary since every method is dependent on a qibo.models.Circuit method. I'd say that if these state classes are needed at all (which I'm not convinced of), you could use Circuit as the base class directly.

renatomello · 2024-10-17T07:03:28Z

src/qibo/tomography_RGD/qibo_states.py

+class HadamardState(State):
+    """
+    Constructor for HadamardState class
+    """
+
+    def __init__(self, n):
+        State.__init__(self, n)
+        self.circuit_name = "Hadamard"
+
+    def create_circuit(self):
+        circuit = qibo.Circuit(self.n)
+
+        for i in range(self.n):
+            circuit.add(gates.H(i))
+
+        self.circuit = circuit


Since this is just a layer of Hadamards, this is a Hadamard transform of the initial state. A more concise way to implement this would be to modify the following function:

qibo/src/qibo/quantum_info/utils.py

Line 118 in 958b11b

def hadamard_transform(array, implementation: str = "fast", backend=None):

For instance, if the input is a circuit, then a layer of Hadamard is added. That would perform the Hadamard transform of whatever state one wants, including the zero state in your case.

renatomello · 2024-10-17T07:05:27Z

src/qibo/tomography_RGD/qibo_states.py

+class GHZState(State):
+    """
+    Constructor for GHZState class
+    """
+
+    def __init__(self, n):
+        State.__init__(self, n)
+        self.circuit_name = "GHZ"
+
+    def create_circuit(self):
+        circuit = qibo.Circuit(self.n)
+
+        circuit.add(gates.H(0))
+        for i in range(1, self.n):
+            circuit.add(gates.CNOT(0, i))
+
+        self.circuit = circuit


This fits as a function in qibo.models.encodings receiving nqubits as input and returning the corresponding circuit as output.

EDIT: I'd recommend doing this on a separate PR. It keeps things concise and it's good training on how to implement Qibo code in a smaller scale.

Just want to clarify. A separate PR you mean that I have to create a new branch, say QST2 or QST_GHZ, and then pre-commit this new branch? Do I understand correctly?

A new branch starting from master, completely unrelated to this one. You can name whatever, but probably ghz is simpler and better. Then in that branch, you only implement the GHZ circuit as a function in qibo.models.encodings and the corresponding tests. Nothing else. After that is merged into master, we change the code here accordingly so it can communicate with the newly existing function.

Thank you for the instruction. I have created a new branch called ghz and push it with sync fork. I have tested some functions inside the directory src/qibo/test. However, I do not understand well the qibo.models.encodings you mentioned. There are some encoders in the qibo.models.encodings but probably we don't need them at the moment. I am not sure if we can add some user-defined properties in the circuit. The way I did it now is to create a new class. Probably some properties do not fit into qibo's standards. I am curious if creating some list as a recorded processed result is allowed. Or do you prefer on-the-fly calculation during processing? Thanks.

I think what @renatomello is that the GHZ state should be added as a function to qibo.models.encodings. It will probably look something like this:

def GHZ_circuit(nqubits): c = QiboCircuit(nqubits) c.add(gates.H(0)) for _i in range(nqubits-1): c.add(gates.CNOT(_i, _i+1)) return c

I don't think it will take so much time to add it in, I can do it after dinner.
Anyway, would qibo.models.utils be a more suitable location than qibo.models.encodings?

EDIT: Upon closer inspection, I see that qibo.models.encodings has a lot of "create circuit" functions. I think qibo.models.encodings will be the suitable location. :)

renatomello · 2024-10-17T07:18:03Z

src/qibo/tomography_RGD/projectors.py

+def build_projector_naive(label, label_format="big_endian"):
+    """to directly generate a Pauli matrix from tensor products
+
+    Args:
+        label (str): label of the projection, e.g.  'XXZYZ'
+        label_format (str, optional): the ordering of the label. Defaults to 'big_endian'.
+
+    Raises:
+        Exception: when the matrix size is too big (i.e. for qubit number > 6)
+
+    Returns:
+        ndarray: a matrix representing the Pauli operator
+    """
+    if label_format == "little_endian":
+        label = label[::-1]
+    if len(label) > 6:
+        raise Exception("Too big matrix to generate!")
+    projector = reduce(
+        lambda acc, item: np.kron(acc, item),
+        [matrix_dict[letter] for letter in label],
+        [1],
+    )
+    return projector
+
+
+# Generate a projector by computing non-zero coordinates and their values in the matrix, aka the "fast" implementation
+def build_projector_fast(label, label_format="big_endian"):
+    """to fastly generate a Pauli projection matrix in sparse matrix format
+
+    Args:
+        label (str): label of the projection, e.g.  'XXZYZ'
+        label_format (str, optional): the ordering of the label. Defaults to 'big_endian'.
+
+    Returns:
+        sparse matrix: sparse matrix of the Pauli operator representing label
+    """
+    if label_format == "little_endian":
+        label = label[::-1]
+
+    n = len(label)
+    d = 2**n
+
+    # map's result NOT subscriptable in py3, just tried map() -> list(map()) for py2 to py3
+    ij = [
+        list(map(binarize, y))
+        for y in [zip(*x) for x in product(*[ij_dict[letter] for letter in label])]
+    ]
+    values = [
+        reduce(lambda z, w: z * w, y)
+        for y in [x for x in product(*[values_dict[letter] for letter in label])]
+    ]
+    ijv = list(map(lambda x: (x[0][0], x[0][1], x[1]), zip(ij, values)))
+
+    i_coords, j_coords, entries = zip(*ijv)
+
+    projector = sparse.coo_matrix(
+        (entries, (i_coords, j_coords)), shape=(d, d), dtype=complex
+    )
+    return projector


@BrunoLiegiBastonLiegi this is a case where it would be convenient to have the quantum_info.basis.pauli_basis function return a generator because then one could use that to directly access a specific element of the basis.

mho291 · 2024-10-17T07:35:36Z

Just curious, would it be helpful to have a call to understand how the algorithm works? It might make it easier to review the code.

update random_density_matrix usage Co-authored-by: Renato Mello <renato.msf@gmail.com>

for more information, see https://pre-commit.ci

update random_density_matrix usage Co-authored-by: Renato Mello <renato.msf@gmail.com>

for more information, see https://pre-commit.ci

mho291 · 2024-10-17T07:57:27Z

@HuberyMing If you're doing the changes locally, you could do pip install pre-commit followed by pre-commit install inside your environment. Each time you do git commit -m "message", the pre-commit will run. If anything fails, you have to do git commit again before git push. Then we we will avoid having extra commits in this PR.

alecandido · 2024-10-17T08:03:14Z

Each time you do git commit -m "message", the pre-commit will run. If anything fails, you have to do git commit again before git push.

Specifically, if it fails the involved hook is "declaring" a failure, that may imply that you have to fix something manually. However, most of the hooks we have registered are also auto-fixing the issues they find (e.g. the formatter fails if the result differs from the original - but if it fails, it also reformats the fails).
The auto-fixes are not staged, so you also need to git add the fixed files.

However, in these cases, I'm sure that @HuberyMing just applied the suggestions from the review from the PR web interface. In which case there is no chance of having pre-commit running anywhere else...

(as you also prepended

@HuberyMing If you're doing the changes locally,

)

renatomello · 2024-10-17T08:05:32Z

Just curious, would it be helpful to have a call to understand how the algorithm works? It might make it easier to review the code.

It might me needed, yes. Let me read the paper first though, so I have a better understanding.

HuberyMing · 2024-10-17T08:15:41Z

Just curious, would it be helpful to have a call to understand how the algorithm works? It might make it easier to review the code.

It might me needed, yes. Let me read the paper first though, so I have a better understanding.

Hi @renatomello Thank you for the review. No problem. How do we discuss?

simplify generating all labels Co-authored-by: Renato Mello <renato.msf@gmail.com>

for more information, see https://pre-commit.ci

renatomello · 2024-10-17T08:30:00Z

Just curious, would it be helpful to have a call to understand how the algorithm works? It might make it easier to review the code.

It might me needed, yes. Let me read the paper first though, so I have a better understanding.

Hi @renatomello Thank you for the review. No problem. How do we discuss?

We can try to arrange a chat via Zoom next week. Right now, I'd like to see a separate PR adding the GHZ state to qibo.models.encodings. I think it will be a good lesson on a small scale on how this PR should move forward in terms of code implementation.
You can follow the style of the functions inside qibo.models.encodings as well as the test style inside tests/test_models_encodings.py.

It's a good first issue.

HuberyMing · 2024-10-17T08:40:26Z

Just curious, would it be helpful to have a call to understand how the algorithm works? It might make it easier to review the code.

It might me needed, yes. Let me read the paper first though, so I have a better understanding.

Hi @renatomello Thank you for the review. No problem. How do we discuss?

We can try to arrange a chat via Zoom next week. Right now, I'd like to see a separate PR adding the GHZ state to qibo.models.encodings. I think it will be a good lesson on a small scale on how this PR should move forward in terms of code implementation. You can follow the style of the functions inside qibo.models.encodings as well as the test style inside tests/test_models_encodings.py.

It's a good first issue.

Thanks. I will do these first.

for more information, see https://pre-commit.ci

renatomello · 2024-10-17T09:57:32Z

src/qibo/tomography_RGD/Run_Tomo.py

+
+    elif circuit_Choice == 2:  # directly generate density matrix via qutip
+        Nr = 1
+        rho = qu.rand_dm_ginibre(2**n, dims=[[2] * n, [2] * n], rank=Nr)


Suggested change

rho = qu.rand_dm_ginibre(2**n, dims=[[2] * n, [2] * n], rank=Nr)

rho = random_density_matrix(2**n, rank=Nr, metric="ginibre")

renatomello · 2024-10-17T09:57:53Z

src/qibo/tomography_RGD/Run_Tomo.py

+import methodsRGD_core
+import numpy as np
+import projectors
+import qutip as qu


Suggested change

import qutip as qu

for more information, see https://pre-commit.ci

MatteoRobbiati · 2025-01-08T22:30:52Z

Do you plan to continue with this PR @HuberyMing? :)

HuberyMing · 2025-01-09T05:56:16Z

Do you plan to continue with this PR @HuberyMing? :)

Sorry for not pushing forward much. I will continue the PR. Hope to finish it soon.

MatteoRobbiati · 2025-01-09T08:21:19Z

Sorry for not pushing forward much. I will continue the PR. Hope to finish it soon.

No problem! Take your time :) I just wanted to check the status!

Hubery Hsu and others added 4 commits October 11, 2024 11:25

QST-RGD

3b78faf

Merge branch 'qiboteam:master' into QST

50a4739

Run_example

13ab8f2

[pre-commit.ci] auto fixes from pre-commit.com hooks

7a54d8f

for more information, see https://pre-commit.ci

renatomello mentioned this pull request Oct 14, 2024

Quantum natural gradient qiboteam/qiboml#35

Open

BrunoLiegiBastonLiegi requested a review from renatomello October 16, 2024 09:20

renatomello reviewed Oct 16, 2024

View reviewed changes

src/qibo/tomography_RGD/BasicTools.py Outdated Show resolved Hide resolved

renatomello reviewed Oct 16, 2024

View reviewed changes

src/qibo/tomography_RGD/Run_Tomo.py Outdated Show resolved Hide resolved

renatomello reviewed Oct 16, 2024

View reviewed changes

src/qibo/tomography_RGD/Run_Tomo.py Outdated Show resolved Hide resolved

renatomello reviewed Oct 16, 2024

View reviewed changes

src/qibo/tomography_RGD/qibo_states.py Outdated Show resolved Hide resolved

renatomello self-requested a review October 16, 2024 10:28

HuberyMing and others added 2 commits October 17, 2024 14:48

Update src/qibo/tomography_RGD/Run_Tomo.py

8e1af29

update random_density_matrix usage Co-authored-by: Renato Mello <renato.msf@gmail.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

cca0e2c

for more information, see https://pre-commit.ci

renatomello reviewed Oct 17, 2024

View reviewed changes

Merge branch 'qiboteam:master' into QST

489f617

HuberyMing and others added 5 commits October 17, 2024 15:39

Update src/qibo/tomography_RGD/Run_Tomo.py

49caa21

update random_density_matrix usage Co-authored-by: Renato Mello <renato.msf@gmail.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

6e3e066

for more information, see https://pre-commit.ci

Update src/qibo/tomography_RGD/qibo_states.py

1ac3eec

update random_density_matrix usage Co-authored-by: Renato Mello <renato.msf@gmail.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

f532834

for more information, see https://pre-commit.ci

update random_density_matrix usage

1e6e7fd

HuberyMing and others added 2 commits October 17, 2024 16:19

Update src/qibo/tomography_RGD/BasicTools.py

16c1670

simplify generating all labels Co-authored-by: Renato Mello <renato.msf@gmail.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

75f1234

for more information, see https://pre-commit.ci

renatomello mentioned this pull request Oct 17, 2024

Add GHZ state to models.encodings #1496

Closed

Hubery Hsu and others added 2 commits October 17, 2024 17:25

update Generate_All_labels fun

a69251c

[pre-commit.ci] auto fixes from pre-commit.com hooks

659dc63

for more information, see https://pre-commit.ci

renatomello reviewed Oct 17, 2024

View reviewed changes

HuberyMing and others added 5 commits November 13, 2024 15:20

Merge branch 'qiboteam:master' into QST

506fde5

shot measure to Pauli coef

b979ca7

[pre-commit.ci] auto fixes from pre-commit.com hooks

2435060

for more information, see https://pre-commit.ci

ghz_state usage

5d95113

[pre-commit.ci] auto fixes from pre-commit.com hooks

b6b2bd5

for more information, see https://pre-commit.ci

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

RGD optimization method for quantum state tomography #1485

RGD optimization method for quantum state tomography #1485

HuberyMing commented Oct 11, 2024

renatomello commented Oct 14, 2024 •

edited

Loading

alecandido commented Oct 16, 2024

renatomello Oct 17, 2024

renatomello Oct 17, 2024

renatomello Oct 17, 2024 •

edited

Loading

renatomello Oct 17, 2024 •

edited

Loading

HuberyMing Oct 17, 2024

renatomello Oct 17, 2024

HuberyMing Oct 22, 2024

mho291 Oct 22, 2024 •

edited

Loading

renatomello Oct 17, 2024

mho291 commented Oct 17, 2024

mho291 commented Oct 17, 2024

alecandido commented Oct 17, 2024

renatomello commented Oct 17, 2024

HuberyMing commented Oct 17, 2024

renatomello commented Oct 17, 2024 •

edited

Loading

HuberyMing commented Oct 17, 2024 •

edited by renatomello

Loading

renatomello Oct 17, 2024 •

edited

Loading

renatomello Oct 17, 2024

MatteoRobbiati commented Jan 8, 2025

HuberyMing commented Jan 9, 2025

MatteoRobbiati commented Jan 9, 2025

	rho = qu.rand_dm_ginibre(2*n, dims=[[2] n, [2] * n], rank=Nr)
	rho = random_density_matrix(2**n, rank=Nr, metric="ginibre")

RGD optimization method for quantum state tomography #1485

Are you sure you want to change the base?

RGD optimization method for quantum state tomography #1485

Conversation

HuberyMing commented Oct 11, 2024

renatomello commented Oct 14, 2024 • edited Loading

alecandido commented Oct 16, 2024

renatomello Oct 17, 2024

Choose a reason for hiding this comment

renatomello Oct 17, 2024

Choose a reason for hiding this comment

renatomello Oct 17, 2024 • edited Loading

Choose a reason for hiding this comment

renatomello Oct 17, 2024 • edited Loading

Choose a reason for hiding this comment

HuberyMing Oct 17, 2024

Choose a reason for hiding this comment

renatomello Oct 17, 2024

Choose a reason for hiding this comment

HuberyMing Oct 22, 2024

Choose a reason for hiding this comment

mho291 Oct 22, 2024 • edited Loading

Choose a reason for hiding this comment

renatomello Oct 17, 2024

Choose a reason for hiding this comment

mho291 commented Oct 17, 2024

mho291 commented Oct 17, 2024

alecandido commented Oct 17, 2024

renatomello commented Oct 17, 2024

HuberyMing commented Oct 17, 2024

renatomello commented Oct 17, 2024 • edited Loading

HuberyMing commented Oct 17, 2024 • edited by renatomello Loading

renatomello Oct 17, 2024 • edited Loading

Choose a reason for hiding this comment

renatomello Oct 17, 2024

Choose a reason for hiding this comment

MatteoRobbiati commented Jan 8, 2025

HuberyMing commented Jan 9, 2025

MatteoRobbiati commented Jan 9, 2025

renatomello commented Oct 14, 2024 •

edited

Loading

renatomello Oct 17, 2024 •

edited

Loading

renatomello Oct 17, 2024 •

edited

Loading

mho291 Oct 22, 2024 •

edited

Loading

renatomello commented Oct 17, 2024 •

edited

Loading

HuberyMing commented Oct 17, 2024 •

edited by renatomello

Loading

renatomello Oct 17, 2024 •

edited

Loading