521 to_vtk

[samjrholt]

User description

Make to_vtk work with different dim names.

---

PR Type

enhancement, tests

---

Description

• Enhanced the to_vtk method in discretisedfield.Field to only allow conversion of 3D fields and dynamically handle dimension names based on mesh.region.dims.
• Added comprehensive tests for the to_vtk method, ensuring it handles fields with both standard and custom dimension names correctly.

---

Changes walkthrough 📝

	Relevant files
Enhancement	field.py Enhance `to_vtk` to Support Dynamic Dimension Names            discretisedfield/field.py Restrict to_vtk conversion to only 3D fields. Adapt coordinate retrieval to use dynamic dimension names from mesh.region.dims. +19/-4
Tests	test_field.py Add Tests for `to_vtk` Method with Dynamic Dimensions        discretisedfield/tests/test_field.py Added a new test test_to_vtk to verify the conversion of fields to VTK format. Extended testing to handle fields with custom dimension names. +44/-0

---

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

[github-actions]

PR Description updated to latest commit (ad79700)

[github-actions]

PR Review 🔍

⏱️ Estimated effort to review [1-5]	2, because the changes are localized to specific methods in a single file and are well-documented with a clear description of the enhancements. The addition of tests also helps in understanding the impact of these changes.
🧪 Relevant tests	Yes
⚡ Possible issues	Possible Bug: The to_vtk method assumes that mesh.region.dims will always have exactly three dimensions. If not, this could lead to errors or unexpected behavior.
🔒 Security concerns	No

Code feedback:

---

relevant file	discretisedfield/field.py
suggestion	Consider adding a check to ensure that mesh.region.dims contains exactly three dimensions before proceeding with the conversion. This would prevent potential runtime errors if the dimensions are not as expected. [important]
relevant line	can be converted. The ``x``, ``y``, and ``z``, spatial coordinates of

---

relevant file	discretisedfield/field.py
suggestion	To improve code readability and maintainability, consider refactoring the repeated code for setting coordinates into a loop or a helper function. This change would make the code cleaner and easier to manage if modifications are needed in the future. [medium]
relevant line	vns.numpy_to_vtk(

---

[github-actions]

PR Code Suggestions ✨

Category	Suggestions
Enhancement	Add a check for the 'nvdim' attribute to ensure it is set before converting to VTK. Consider adding a check to ensure that the nvdim attribute is properly set before proceeding with the conversion to VTK format. This will prevent potential errors or undefined behavior if nvdim is not set or incorrectly set. discretisedfield/field.py [3384-3386] +if not hasattr(self, 'nvdim'): + raise ValueError("The 'nvdim' attribute must be set to convert to VTK format.") Scalar fields (``nvdim=1``) contain one VTK array called ``field``. Vector fields (``nvdim>1``) contain one VTK array called ``field`` containing vector data and
	Enhance the test_to_vtk function by adding assertions to verify the data in the VTK arrays. To ensure the robustness of the test_to_vtk function, add assertions to check the correctness of the data in the VTK arrays, not just their names. discretisedfield/tests/test_field.py [2811-2815] assert cell_data.GetArrayName(0) == "norm" +assert np.allclose(cell_data.GetArray(0).GetTuple(0), expected_norm_values) assert cell_data.GetArrayName(1) == "x" +assert np.allclose(cell_data.GetArray(1).GetTuple(0), expected_x_values) assert cell_data.GetArrayName(2) == "y" +assert np.allclose(cell_data.GetArray(2).GetTuple(0), expected_y_values) assert cell_data.GetArrayName(3) == "z" +assert np.allclose(cell_data.GetArray(3).GetTuple(0), expected_z_values) assert cell_data.GetArrayName(4) == "field" +assert np.allclose(cell_data.GetArray(4).GetTuple(0), expected_field_values)
Maintainability	Use a loop to set X, Y, and Z coordinates to enhance code readability and maintainability. To improve code readability and maintainability, consider using a loop to set the coordinates instead of repeating similar code blocks for X, Y, and Z coordinates. discretisedfield/field.py [3426-3445] -rgrid.SetXCoordinates( - vns.numpy_to_vtk( - np.fromiter( - getattr(self.mesh.vertices, self.mesh.region.dims[0]), float - ) - ) -) -rgrid.SetYCoordinates( - vns.numpy_to_vtk( - np.fromiter( - getattr(self.mesh.vertices, self.mesh.region.dims[1]), float - ) - ) -) -rgrid.SetZCoordinates( - vns.numpy_to_vtk( - np.fromiter( - getattr(self.mesh.vertices, self.mesh.region.dims[2]), float - ) - ) -) +for dim, setter in zip(self.mesh.region.dims, [rgrid.SetXCoordinates, rgrid.SetYCoordinates, rgrid.SetZCoordinates]): + setter(vns.numpy_to_vtk(np.fromiter(getattr(self.mesh.vertices, dim), float)))
Bug	Add error handling for fields with dimensions other than 3 in the to_vtk method. Ensure that the to_vtk method handles cases where the field dimension (ndim) is not equal to 3, either by raising an exception or providing a fallback mechanism. discretisedfield/field.py [3381] +if self.ndim != 3: + raise ValueError("Only fields with ndim=3 can be converted to VTK format.") Only fields with ``ndim=3`` can be converted.
Possible issue	Add checks to ensure the existence of required attributes in mesh.vertices before accessing them. To avoid potential issues with attribute access, use a more robust method to check for the existence of mesh.vertices attributes based on mesh.region.dims. discretisedfield/field.py [3429-3443] +for dim in self.mesh.region.dims: + if not hasattr(self.mesh.vertices, dim): + raise AttributeError(f"Mesh vertices do not have the dimension '{dim}'") getattr(self.mesh.vertices, self.mesh.region.dims[0]), float getattr(self.mesh.vertices, self.mesh.region.dims[1]), float getattr(self.mesh.vertices, self.mesh.region.dims[2]), float

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.39%. Comparing base (f1b4cda) to head (d0cf601).
Report is 86 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #527      +/-   ##
==========================================
- Coverage   93.48%   93.39%   -0.10%     
==========================================
  Files          28       31       +3     
  Lines        3025     3286     +261     
==========================================
+ Hits         2828     3069     +241     
- Misses        197      217      +20     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.