Make kikuchipy compatible with PyEBSDIndex 0.2

hakonanes commented 11 months ago

Description of the change

This PR aims at making kikuchipy compatible with just released PyEBSDIndex v0.2.0 and to exploit its many new features. New features of main interest are:

Improved PC optimization via particle swarm optimization (PSO), allowing broader search limits in the PC values (default is 0.2 or 20% of the detector dimensions; quite wide!)
Hough indexing of all Laue classes (not only FCC or BCC). So far, only FCC and BCC (m-3m), HCP (6/mmm), and the triclinic case (-1) have been tested.

kikuchipy v0.8.* has PyEBSDIndex as an optional dependency with a compatible release specification (~= 0.1.2), so people installing this release from PyPI will only get PyEBSDIndex v0.1.2. However, our conda-forge feedstock only asks for >= 0.1.2... People installing from conda-forge will most likely get PyEBSDIndex v0.2.0, which does not work with the current kikuchipy release. A big problem. So we should get this merged and released as 0.9.0 quickly. I plan to get this done within Tuesday next week.

Progress of the PR

[x] Docstrings for all functions
[x] Unit tests with pytest for all lines
[x] Clean code style by running black via pre-commit
[x] Update all affected tutorials (and the PM tutorial):
- [x] Pattern matching
- [x] PC plane fit
- [x] PC plane extrapolation
- [x] ESTEEM3 workshop
- [x] HI
- [x] Hybrid indexing

~What we should do but perhaps don't have time~:

~[ ] Add patterns from crystals having symmetries from all Laue classes to the data module~
~[ ] Test HI of these new patterns~
~[ ] Add examples showing how to index each of these patterns to the docs~

For reviewers

[ ] The PR title is short, concise, and will make sense 1 year later.
[ ] New functions are imported in corresponding __init__.py.
[ ] New features, API changes, and deprecations are mentioned in the unreleased section in CHANGELOG.rst.
[ ] New contributors are added to release.py, .zenodo.json and .all-contributorsrc with the table regenerated.

CSSFrancis commented 11 months ago

@hakonanes Have you considered adding https://github.com/apps/review-notebook-app to kikchipy. It is fairly good at showing diffs for jupyter notebooks?

I can review this if you want someone to (superficially) look at the changes.

hakonanes commented 11 months ago

Thanks for the suggestions, @CSSFrancis. I'll consider ReviewNB. I'll ask you for a review once the PR's ready!

review-notebook-app[bot] commented 8 months ago

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

hakonanes commented 8 months ago

@CSSFrancis, this work is ready for merging. If you want to have a quick look, that would be very much appreciated! I plan to merge tomorrow Friday.

I'll get to reviewing and answering your stuff in orix, diffsims, and pyxem shortly!

The only outward API changes are the possibility to (listed in the changelog):

pass a reflector list per phase for use in Hough indexing
pass parameters for particle swarm optimization (PSO) of projection/pattern center (detector-sample geometry calibration)
optimizing one projection center at a time with PSO

Other changes:

Tutorial notebooks: use improvements to optimization and indexing accuracy in PyEBSDIndex v0.2; improve explanations of many steps and analyses.
Replace interactive EBSD master pattern 3D plot with a static plot in tutorials for visualization and kinematic simulations. This was needed because PyVista have discontinued support for the Jupyter backend we used, Panel. The currently supported Jupyter backends cannot provide interactive 3D plots. But they're working on it!

hakonanes commented 8 months ago

And thank you for suggesting ReviewNB. It sure beats me comparing the changes to rendered tutorial pages in current docs and the newly built docs from a PR by eye!

hakonanes commented 8 months ago

Now that you've both reviewed and contributed to the docs, is it OK if I add you to the package credits? This includes the kikuchipy/release.py file, the all-contributors table in the README, and in the Zenodo file. You'll be listed as "Carter Francis" with the following entry in the Zenodo file

   {
      "name": "Carter Francis",
      "orcid": "0000-0003-2564-1851",
      "affiliation": "University of Wisconsin Madison"
    },

hakonanes commented 8 months ago

I plan to release 0.9.0 once this PR is merged and some other smaller things are done 🚀

CSSFrancis commented 8 months ago

Now that you've both reviewed and contributed to the docs, is it OK if I add you to the package credits? This includes the kikuchipy/release.py file, the all-contributors table in the README, and in the Zenodo file. You'll be listed as "Carter Francis" with the following entry in the Zenodo file
   {

      "name": "Carter Francis",

      "orcid": "0000-0003-2564-1851",

      "affiliation": "University of Wisconsin Madison"

    },

Yea that works for me :)

pyxem / kikuchipy