Fix the custom_gradient issue with hermite_renormalized

maliasadi commented 1 month ago

User description

Before submitting

Please complete the following checklist when submitting a PR:

[x] All new features must include a unit test. If you've fixed a bug or added code that should be tested, add a test to the test directory!
[x] All new functions and code must be clearly commented and documented. If you do make documentation changes, make sure that the docs build and render correctly by running make docs.
[X] Ensure that the test suite passes, by running make test.
[x] Ensure that code and tests are properly formatted, by running make format or `black -l 100
` on any relevant files. You will need to have the Black code format installed: `pip install black`.
[ ] Add a new entry to the .github/CHANGELOG.md file, summarizing the change, and including a link back to the PR.
[x] The Mr Mustard source code conforms to PEP8 standards except line length. We check all of our code against Pylint. To lint modified files, simply pip install pylint, and then run pylint mrmustard/path/to/file.py.

When all the above are checked, delete everything above the dashed line and fill in the pull request template.

Context: This PR fixes the following MM issue with TensorFlow >= 2.15:

import tensorflow as tf
print(tf.__version__)

import numpy as np
import mrmustard.lab_dev as mrml
from mrmustard import math
from mrmustard.training import Optimizer
math.change_backend('tensorflow')

S = mrml.SqueezedVacuum([0], r=1.0, phi=0, r_trainable=True, phi_trainable=True)

# made up, means nothing
def cost():
    ket = S.fock(shape=[3])
    return - math.real(ket[2])

circuit = mrml.Circuit([S])

# we used to do this:
opt = Optimizer()
opt.minimize(cost, by_optimizing=[circuit], max_steps=20)

ValueError: custom_gradient function expected to return 4 gradients, but returned 3 instead.

This issue arises in custom operations where a gradient function is manually defined, and the expected number of gradients doesn't match the number returned by the function. To fix this issue, we updated all custom_gradient use cases in MM to return correct number of gradients. :heavy_check_mark:

Python formatting could also be improved with supporting different python version targets in black (to avoid formatting inconsistencies when using different py versions) and sorting imports using isort.

PR Type

enhancement, tests

Description

Updated TensorFlow version to 2.17.0 in pyproject.toml to ensure compatibility with the latest features and improvements.
Enhanced the Makefile to include isort for sorting imports and added conditional checks for formatting.
Modified the gradient function in backend_tensorflow.py to return an additional None value, likely for improved error handling or compatibility.
Reformatted import statements across multiple test files for consistency and readability.
Added isort as a dependency and configured it alongside black for code formatting.

[sc-75755]

Changes walkthrough 📝

Relevant files

Enhancement

backend_tensorflow.py

Update gradient function to return extra None value

mrmustard/math/backend_tensorflow.py - Modified the `grad` function to return an additional `None` value.

+1/-1

Dependencies

pyproject.toml

Update TensorFlow version and add isort configuration

pyproject.toml

Updated TensorFlow version to 2.17.0.

Added isort as a dependency.

Configured isort and black settings.

+11/-6

Configuration changes

Makefile

Enhance Makefile with isort and conditional formatting

Makefile

Added isort to the format command.

Introduced conditional checks for formatting.

+11/-2

Tests

conftest.py

Add hypothesis settings import

tests/conftest.py - Added import for `hypothesis.settings`.

+3/-1

Formatting

test_circuit.py

Reformat import statements for consistency

tests/test_lab/test_circuit.py - Reformatted import statements for consistency.

+1/-8

💡 PR-Agent usage: Comment /help "your question" on any pull request to receive relevant information

codiumai-pr-agent-pro[bot] commented 1 month ago

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 3 🔵🔵🔵⚪⚪

🏅 Score: 85

🧪 No relevant tests

🔒 No security concerns identified

⚡ Recommended focus areas for review

Possible Bug
The `grad` function in `hermite_renormalized` now returns an additional `None` value. This change might affect the behavior of the function and its callers. Code Smell
Multiple import statements have been reordered and grouped. This change might affect the readability and organization of the code.

codiumai-pr-agent-pro[bot] commented 1 month ago

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Score
Best practice	Specify exact versions for dependencies to improve reproducibility ___ Consider specifying exact versions for dependencies to ensure reproducibility, especially for critical packages like TensorFlow. [pyproject.toml [52-53]](https://github.com/XanaduAI/MrMustard/pull/506/files#diff-50c86b7ed8ac2cf95bd48334961bf0530cdc77b5a56f852c5c61b89d735fd711R52-R53) ```diff tensorflow = {version = "2.17.0" } -tensorflow-intel = { version = ">2.15.0", platform = "win32" } +tensorflow-intel = { version = "2.17.0", platform = "win32" } ``` - [ ] Apply this suggestion Suggestion importance[1-10]: 7 Why: Specifying exact versions for dependencies can enhance reproducibility by ensuring that the same versions are used across different environments. This is particularly important for critical packages like TensorFlow.	7
Enhancement	Enhance user guidance by providing an example for the new check option ___ Consider adding a help message for the new `check` option in the format target to improve user guidance. [Makefile [21]](https://github.com/XanaduAI/MrMustard/pull/506/files#diff-76ed074a9305c04054cdebb9e9aad2d818052b07091de1f20cad0bbac34ffb52R21-R21) ```diff @echo " format [check=1] to run isort and black formatting; use with 'check=1' to check instead of modify" +@echo " Example: make format check=1" ``` - [ ] Apply this suggestion Suggestion importance[1-10]: 5 Why: Adding an example for the new `check` option in the help message can improve user guidance and make it easier for users to understand how to use the new feature. This is a minor enhancement but can be helpful for clarity.	5

💡 Need additional feedback ? start a PR chat

codecov[bot] commented 1 month ago

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 89.41%. Comparing base (5fe4a5f) to head (60d106c). Report is 1 commits behind head on develop.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## develop #506 +/- ## ======================================== Coverage 89.41% 89.41% ======================================== Files 89 89 Lines 6026 6026 ======================================== Hits 5388 5388 Misses 638 638 ``` | [Files with missing lines](https://app.codecov.io/gh/XanaduAI/MrMustard/pull/506?dropdown=coverage&src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=XanaduAI) | Coverage Δ | | |---|---|---| | [mrmustard/lab\_dev/circuit\_components.py](https://app.codecov.io/gh/XanaduAI/MrMustard/pull/506?src=pr&el=tree&filepath=mrmustard%2Flab_dev%2Fcircuit_components.py&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=XanaduAI#diff-bXJtdXN0YXJkL2xhYl9kZXYvY2lyY3VpdF9jb21wb25lbnRzLnB5) | `99.09% <100.00%> (ø)` | | ------ [Continue to review full report in Codecov by Sentry](https://app.codecov.io/gh/XanaduAI/MrMustard/pull/506?dropdown=coverage&src=pr&el=continue&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=XanaduAI). > **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=XanaduAI) > `Δ = absolute (impact)`, `ø = not affected`, `? = missing data` > Powered by [Codecov](https://app.codecov.io/gh/XanaduAI/MrMustard/pull/506?dropdown=coverage&src=pr&el=footer&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=XanaduAI). Last update [5fe4a5f...60d106c](https://app.codecov.io/gh/XanaduAI/MrMustard/pull/506?dropdown=coverage&src=pr&el=lastupdated&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=XanaduAI). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=XanaduAI).

XanaduAI / MrMustard