Add demo on resourcefulness analysis with generalised Fourier transforms #1462
Conversation
|
👋 Hey, looks like you've updated some demos! 🐘 Don't forget to update the Please hide this comment once the field(s) are updated. Thanks! |
…o resourcefulness_demo
|
Hi @mariaschuld — The metadata checks were failing due to the following categories not passing validation. I think @alvaro-at-xanadu can clarify if these need to be added. |
Your preview is ready 🎉!You can view your changes here |
Hi @mariaschuld These are the allowed categories: 
|
…o resourcefulness_demo
…o resourcefulness_demo
|
@daniela-angulo I added Tarik's pictures, but need to work on this demo a little more to polish and turn code into PL functionality. Will get back to it! |
|
|
||
| ###################################################################### | ||
| # | ||
| # About the authors | ||
| # ----------------- | ||
| # |
There was a problem hiding this comment.
Hey @mariaschuld — we now no longer need these sections at the bottom demos, they are added dynamically when displayed on the website.
| ###################################################################### | |
| # | |
| # About the authors | |
| # ----------------- | |
| # |
mariaschuld
changed the title
…o resourcefulness_demo
|
|
||
| import pennylane as qml | ||
| from scipy.stats import unitary_group | ||
|
|
There was a problem hiding this comment.
@Alex-Preciado @isaacdevlugt , the code shared by the authors was creating these states by hand. I found it much neater to use PL, and I think this is an important use case: to make mathematical computations quicker and more reliable. However, I would love to have a qml.HaarState operator!
|
|
||
| import itertools | ||
| import functools | ||
|
|
There was a problem hiding this comment.
This would be a perfect section for PL's pauli module. I tried this:
from pennylane.pauli import string_to_pauli_word, pauli_word_to_matrix
n=2
all_strs = [''.join(s) for s in itertools.product('IXYZ', repeat=n)]
# sort by number of identities in the Pauli word
basis = sorted(all_strs, key=lambda s: (n - s.count('I'), s))
# collect flattened Pauli basis matrices
B = []
for s in basis:
pauli_word = string_to_pauli_word(s)
M = pauli_word_to_matrix(pauli_word)
print(pauli_word, M)
M_flat = M.reshape(-1) / np.sqrt(2**n)
B.append(M_flat)
B = np.array(B)
Where I gave up for now is that string_to_pauli_word somehow deletes the identities, and I end up with the wrong shape of the matrices. I know I could circumvent this using the wire_map keyword, but didn't have the patience.
|
@daniela-angulo this demo is ready for another technical review and a product review! |
| # `here <https://github.com/PennyLaneAI/qml/demonstrations_v2/tutorial_resourcefulness/utils.py>`__: | ||
| # | ||
|
|
||
| from utils import group_rows_cols_by_sparsity |
There was a problem hiding this comment.
Is this a good way to do it? I could probably find a more elegant way than using this function, but it would take too much time right now.
| # But ``Uvec_diag`` does not look block diagonal. What happened here? | ||
| # Well, it *is* block-diagonal, but we have to reorder the columns and rows of the final matrix to make this visible. | ||
| # This takes a bit of pain, which we outsource to a utility function that can be found | ||
| # `here <https://github.com/PennyLaneAI/qml/demonstrations_v2/tutorial_resourcefulness/utils.py>`__: |
There was a problem hiding this comment.
Todo: check this link when live!
|
I will wait until the technical review is done and then perform the content review. I'll keep an eye on. Thank you so much @mariaschuld |
This is a demo we are creating with the first authors of this paper.