fft notebook

[samjrholt]

Dependent on:

• no fftshift on last dim for real ffts #410

Summary by CodeRabbit

• New Feature: Improved performance of the Fast Fourier Transform (FFT) functionality by utilizing the faster scipy library and optimizing computation.
• Documentation: Added a comment to explain the purpose of a code block in discretisedfield/mesh.py.
• Test: Enhanced test coverage for FFT calculations in discretisedfield/tests/test_field.py, ensuring correctness in various scenarios.
• Refactor: Adjusted the axes parameter in certain functions of discretisedfield/field.py to exclude the last axis, improving computation efficiency.
• Chore: Imported scipy.fft module as spfft in discretisedfield/tests/test_field.py.

These changes collectively improve the performance and reliability of the code while maintaining compatibility with the external interface.

[samjrholt]

@lang-m When you get a chance could you please look through this. There may be a bug in rfft as the peak in the spectrum is not in the same place.

So the current example is probably more useful for people doing micromagnetics but I also considered changing the example to 2 spatial dimensions and a scalar field as this simplifies the code/plots a little.

[samjrholt]

I have fixed the bug in #410. Please review that first and then merge into this one then please review this.

[lang-m]

The notebook looks really good. Here is some preliminary feedback from a first review. I'll go over it again more thoroughly after reviewing and merging the bug fix.

[codecov-commenter]

Codecov Report

Patch coverage: 94.62% and project coverage change: -3.53% ⚠️

Comparison is base (dfbfdb5) 96.87% compared to head (9f1c19a) 93.35%.
Report is 1163 commits behind head on master.

❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #409      +/-   ##
==========================================
- Coverage   96.87%   93.35%   -3.53%     
==========================================
  Files          28       28              
  Lines        2498     2933     +435     
==========================================
+ Hits         2420     2738     +318     
- Misses         78      195     +117     

Files Changed	Coverage Δ
discretisedfield/field.py	87.01% <ø> (-9.94%)	⬇️
discretisedfield/interact.py	100.00% <ø> (ø)
discretisedfield/plotting/util.py	92.94% <66.66%> (-4.32%)	⬇️
discretisedfield/plotting/k3d_mesh.py	95.00% <75.00%> (-5.00%)	⬇️
discretisedfield/plotting/mpl_field.py	85.85% <83.13%> (-3.27%)	⬇️
discretisedfield/plotting/k3d_region.py	96.00% <83.33%> (-4.00%)	⬇️
discretisedfield/io/hdf5.py	85.33% <88.23%> (-14.67%)	⬇️
discretisedfield/plotting/mpl_region.py	96.77% <88.88%> (-3.23%)	⬇️
discretisedfield/io/__init__.py	89.28% <89.28%> (-10.72%)	⬇️
discretisedfield/plotting/k3d_field.py	95.03% <94.44%> (-0.07%)	⬇️
... and 14 more

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

[lang-m]

Final review for the first part of the notebook (anything before "Visualising and analysing FFTs"). I have added some more suggestions and minor corretions.

I'll try to also review the second half of the notebooks soon.

[github-actions]

CodeRabbit

Walkthrough

The changes made to the codebase focus on improving the Fast Fourier Transform (FFT) functionality in the discretisedfield module. The modifications include utilizing the faster scipy library for computation, optimizing the computation by excluding unnecessary axes, adding comments for code clarity, and enhancing the test coverage for FFT calculations. These changes collectively improve the performance, maintainability, and correctness of the code.

Changes

File	Summary
discretisedfield/field.py	- Use scipy library for faster computation. - Adjust axes parameter to exclude the last axis in certain functions.
discretisedfield/mesh.py	Add a comment to explain the purpose of a code block.
discretisedfield/tests/test_field.py	- Import scipy.fft module as spfft. - Modify or remove assertions related to FFT calculations. - Add additional tests for FFT calculations, including testing 1D rfft and rfftn with no shift in the last dimension. - Create a new field a with specific values and compare its FFT with the expected result using scipy.fft.rfftn.

"With scipy we dance,

Faster FFT, a chance.

Axes trimmed, code refined,

Comments clear, peace of mind.

Tests expanded, correctness assured,

Rabbit's joy, code's allure." 🐇✨

---

Uplevel your code reviews with CodeRabbit Pro

CodeRabbit Pro

If you like this project, please support us by purchasing the Pro version. The Pro version has advanced context, superior noise reduction and several proprietary improvements compared to the open source version.

[lang-m]

Some additional minor comments for the second half of the notebook.

Feel free to merge once all open comments are addressed.

Edit: several of the comments seem to be off by one line (not sure how that has happened). I hope they all make sense. Directly applying the suggestions of course doesn't work for these.