feature: calculate subvolume defocus values

[McHaillet]

This PR adds support for calculating defocus value for each subvolume specified with --volume-split in template matching. As particles experience a different defocus based on their position in the ice and depending on the tilt angle, this potentially provides detailed defocus gradient correction. However, I also added a page on the site with a benchmark for measuring the effects of defocus gradient correction and found that the effect is very small. I thought this would be cool to add.

• implementation
• add benchmark

• I left the version number as we did not release 0.7.2 yet.
• I made a change in angles.py to read the log_level as an integer as this is what is available in the TMJob. I also removed a section there that overwrote the log_level as a string. (should have put this in a separate PR, sorry for that)

[github-actions]

File	Coverage		Missing
All files	87%	✅
src/pytom_tm/correlation.py	85%	❌	98-99 131
src/pytom_tm/entry_points.py	79%	❌	28 113-130 235-280 385-404 540-558 1030-1041
src/pytom_tm/extract.py	76%	❌	127-138 205-212 256-261 354-372
src/pytom_tm/io.py	80%	❌	20 40 58 76 89 110-116 133-137 144 150 162 169 216-218 288 325-328 433-441 466
src/pytom_tm/mask.py	80%	❌	76 94 104-105
src/pytom_tm/parallel.py	89%	✅	14-15 93 104-105 111 176
src/pytom_tm/plotting.py	16%	❌	19-21 24-31 35-43 47-50 53-69 72-73 76-77 97-102 118-126 132-149 164-172 176-193 197-205 209-216 222 228 233 247-397
src/pytom_tm/tmjob.py	97%	✅	364-365 381 391-396 475 598 647
src/pytom_tm/utils.py	78%	❌	18-20
src/pytom_tm/weights.py	96%	✅	59 314-318 325 535 540 545

Minimum allowed coverage is 86%

Generated by 🐒 cobertura-action against ed489e6

[sroet]

Some issues and feedback.

Other points:

• Would this need an option to override the tilt-axis? Or are those always "close enough" to the y-axis?

[sroet]

this seems broken in github MD rendering

[McHaillet]

I guess it is not supported in native markdown: https://squidfunk.github.io/mkdocs-material/reference/images/#image-captions

The md_in_html extension (in the mkdocs.yml) enables it.

[McHaillet]

I meant the captions are not supported

[McHaillet]

* Would this need an option to override the tilt-axis? Or are those always "close enough" to the y-axis?

The tilt axis of the tilt series is always aligned with the y-axis of the tomogram. It's the common convention as far as I know. So should be safe.

[sroet]

Thanks for the changes! LGTM!