Numpy2

[lang-m]

User description

• [x] rerun tests after merging ubermag/ubermagutil#53

---

PR Type

Enhancement, Bug fix

---

Description

• Updated datetime usage to handle deprecation of utcnow() in Python 3.12.
• Added .item() to convert xarray and numpy values to Python built-in types.
• Converted mesh.n, region.pmin, and region.pmax to lists using tolist() method in Jinja2 templates.

---

Changes walkthrough 📝

	Relevant files
Bug fix	hdf5.py Update datetime usage to handle deprecation in Python 3.12 discretisedfield/io/hdf5.py Updated utcnow to now with datetime.UTC to handle deprecation in Python 3.12. +1/-1
Enhancement	hv.py Ensure compatibility with xarray in resampling function    discretisedfield/plotting/hv.py Added .item() to convert xarray to Python built-in type before using linspace. +4/-1      tools.py Convert numpy values to Python types in count_bps function discretisedfield/tools/tools.py - Added `.item()` to convert numpy values to Python built-in types. +4/-3      mesh.jinja2 Convert mesh.n to list in mesh template                                    discretisedfield/html/templates/mesh.jinja2 - Converted `mesh.n` to list using `tolist()` method. +1/-1      region.jinja2 Convert region.pmin and region.pmax to list in region template discretisedfield/html/templates/region.jinja2 Converted region.pmin and region.pmax to list using tolist() method. +2/-2

---

💡 PR-Agent usage: Comment /help on the PR to get a list of all available PR-Agent tools and their descriptions

[github-actions[bot]]

PR Reviewer Guide 🔍

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 No relevant tests

🔒 No security concerns identified

⚡ Key issues to review **Possible Bug:** Ensure that the use of `datetime.datetime.now(datetime.UTC)` is correctly implemented. The correct usage should be `datetime.datetime.now(datetime.timezone.utc)`. **Data Conversion:** Verify that the use of `.item()` for converting numpy and xarray data types to Python built-in types does not lead to loss of precision or other unintended side effects, especially in large datasets. **Data Structure Conversion:** Ensure that the conversion of numpy arrays to lists using `.tolist()` in the Jinja2 templates does not impact performance or expected behavior, particularly with large arrays.

[github-actions[bot]]

PR Code Suggestions ✨

Category	Suggestion	Score
Possible issue	Use the recommended datetime.timezone.utc for timezone-aware datetime objects ___ **Replace datetime.UTC with datetime.timezone.utc for compatibility with Python 3.9 and later versions, which deprecate datetime.UTC.** [discretisedfield/io/hdf5.py [76]](https://github.com/ubermag/discretisedfield/pull/532/files#diff-39def737ba78f85af8f515ef3c1c5602391c99bcb75eaeb189ff4a366cd6d576R76-R76) ```diff -utc_now = datetime.datetime.now(datetime.UTC).isoformat(timespec="seconds") +utc_now = datetime.datetime.now(datetime.timezone.utc).isoformat(timespec="seconds") ``` Suggestion importance[1-10]: 10 Why: The suggestion correctly identifies the deprecation of `datetime.UTC` and provides a compatible alternative, improving code compatibility with Python 3.9 and later versions.	10
Possible bug	Add error handling for potential None values in min/max calculations ___ **Add error handling for cases where array[dim].min() or array[dim].max() might return NaN or None, which would cause item() to fail.** [discretisedfield/plotting/hv.py [740]](https://github.com/ubermag/discretisedfield/pull/532/files#diff-3e50151dd1046211ca111a090ea920ee499420b5fb3bfa96e04319fc95a90a0cR740-R740) ```diff -dim: np.linspace(array[dim].min().item(), array[dim].max().item(), ni) +dim: np.linspace(array[dim].min().item() if array[dim].min() is not None else 0, array[dim].max().item() if array[dim].max() is not None else 0, ni) ``` Suggestion importance[1-10]: 8 Why: The suggestion addresses a potential bug by adding error handling for None values, which could cause the `item()` method to fail, thus improving the robustness of the code.	8
	Add checks to ensure region.pmin and region.pmax are iterable before converting to list ___ **Ensure that region.pmin and region.pmax are always lists or arrays before calling tolist() to avoid potential errors if the properties are not iterable.** [discretisedfield/html/templates/region.jinja2 [3-4]](https://github.com/ubermag/discretisedfield/pull/532/files#diff-db22f8be37d96fc9cf5d95e42c0fd376b7c7c5fd6d626ea3967ca821a2f7ae70R3-R4) ```diff - pmin = {{ region.pmin.tolist() }} - pmax = {{ region.pmax.tolist() }} + pmin = {{ region.pmin.tolist() if isinstance(region.pmin, (list, np.ndarray)) else [region.pmin] }} + pmax = {{ region.pmax.tolist() if isinstance(region.pmax, (list, np.ndarray)) else [region.pmax] }} ``` Suggestion importance[1-10]: 7 Why: The suggestion improves code robustness by ensuring `region.pmin` and `region.pmax` are iterable before calling `tolist()`, preventing potential runtime errors.	7