Dorr mri

[PolarBean]

Description

This adds the dorr mouse mri atlas to ccf translator with a deformation to perens stereotaxic mri created by @saarah815

What is this PR

• Bug fix
• Addition of a new feature
• Other

References

brainglobe/brainglobe-atlasapi#657

How has this PR been tested?

ive manually checked that it produces the correct results transforming to and from perens stereotaxic mri and allen

Checklist:

• The code has been tested locally
• Tests have been added to cover all new functionality (unit & integration)
• The documentation has been updated to reflect any changes
• The code has been formatted with pre-commit

[alessandrofelder]

@IgorTatarnikov has kindly volunteered to have a look at this given his experience of working on the Dorr atlas with @saarah815.

Before this, could you ensure the tests pass, please? (Not sure how busy you are atm)
From a quick look at the CI error message, it looks like brainglobe_ccf_translator/config.py imports click but that is not specified as a dependency in pyproject.toml

[PolarBean]

Hey @alessandrofelder so theres a bit of a backlog of PRs. This is merging into #26 which merges into #23.

I think the tests will always be too heavy to pass in an action, and the repo tests pass offline. But yeah some problems like this click import thing should definitely be fixed. Thank you

[PolarBean]

eventually we will do something like this so that the tests are more light weight and runnable via gh action #28

[PolarBean]

I will run it again offline just to be sure and post the results here

[codecov]

Welcome to Codecov 🎉

Once you merge this PR into your default branch, you're all set! Codecov will compare coverage reports and display results in all future pull requests.

Thanks for integrating Codecov - We've got you covered ☂️

[PolarBean]

genuinely shocked that some of the VMs did not crash and are passing 🎉

[PolarBean]

also passes on my machine offline

❯ python -m pytest
=============================================================================== test session starts ================================================================================
platform linux -- Python 3.11.7, pytest-7.4.0, pluggy-1.6.0
rootdir: /home/harryc/github/brainglobe-ccf-translator
configfile: pyproject.toml
plugins: anyio-4.2.0, cov-7.0.0
collected 12 items                                                                                                                                                                 

tests/test_PointSet.py .....                                                                                                                                                 [ 41%]
tests/test_Volume.py ...                                                                                                                           [ 66%]
tests/test_VolumeSeries.py .                                                                                                                       [ 75%]
tests/test_space_utils.py ...                                                                                                                      [100%]

===================================================================== tests coverage =====================================================================
____________________________________________________ coverage: platform linux, python 3.11.7-final-0 _____________________________________________________

Name                                                                           Stmts   Miss  Cover
--------------------------------------------------------------------------------------------------
brainglobe_ccf_translator/PointSet.py                                             57      6    89%
brainglobe_ccf_translator/Volume.py                                               58     14    76%
brainglobe_ccf_translator/VolumeSeries.py                                         72     10    86%
brainglobe_ccf_translator/__init__.py                                              9      2    78%
brainglobe_ccf_translator/config.py                                               65     36    45%
brainglobe_ccf_translator/deformation/__init__.py                                  1      0   100%
brainglobe_ccf_translator/deformation/apply_deformation.py                       190     62    67%
brainglobe_ccf_translator/deformation/forward_transform.py                        50     50     0%
brainglobe_ccf_translator/deformation/interpolation/NearestNDInterpolator.py      59     59     0%
brainglobe_ccf_translator/deformation/interpolation/__init__.py                    1      1     0%
brainglobe_ccf_translator/deformation/route_calculation.py                        40     16    60%
brainglobe_ccf_translator/read_write.py                                           23     23     0%
brainglobe_ccf_translator/space_utils.py                                          35      4    89%
--------------------------------------------------------------------------------------------------
TOTAL                                                                            660    283    57%
============================================================ 12 passed in 1961.65s (0:32:41) =============================================================

[alessandrofelder]

I think the tests will always be too heavy to pass in an action, and the repo tests pass offline.
so that the tests are more light weight and runnable via gh action #28

OK! Sound like a plan :) I'd suggest noting on future PR descriptions with this situation that you don't expect tests to pass (because reviewers might wait for you to fix them before taking time to review otherwise). (But ideally CI tests pass before anyone is asked to review.)

I'd also (possibly naively) say you can drop the CI runs for Python 3.9 and 3.10 (that were cancelled - maybe because they hit the runner VM limits somehow) - other BrainGlobe tools support Python 3.11+ only anyway.

[PolarBean]

Good idea :) Done

[PolarBean]

Are you ok with me merging this (and the other PR's as theyre sort of merging in a chain).

[IgorTatarnikov]

Sorry it took so long to review!

It looks great, I just left a few comments that aren't blocking. It would be good to avoid storing the shapes of the atlases as hard coded values. It's always best to have one source of truth!

[IgorTatarnikov]

Will perens be a common intermediary? It might be worth saving the metadata for the perens atlas as a constant somewhere to import into other files.

[PolarBean]

i can access it through brainglobe it just then requires installing the whole atlas to get the shape. I think this will be fixed in v2 and its also nice to use our tools for things like this so ill change it.

[IgorTatarnikov]

I wonder if the shapes could be loaded from somewhere.

The shapes could be accessed via brainglobe-atlasapi, however, this would entail downloading the entire atlas which may slow things down. This will be solved with BrainGlobe Atlas V2 as you'd only fetch the metadata file and no data to return the shape.

[PolarBean]

ah aha i see i replied to your other comment before reading this! I am ok with using the api for this :)

[PolarBean]

one really weird issue is the dorr shape here is not actually the shape of dorr! but instead the shape of dorr if it were 25 micron voxels (to match perens). Ill jsut add scaling for this.

[PolarBean]

No stress! and I agree about storing atlas shapes, I'll look into a better alternative

[PolarBean]

the jobs are again running out of resources but the parts of the code theyre testing havent changed since the last time it passed. I just tried to rerun some of them. This is however ready for review. Thank you

[IgorTatarnikov]

LGTM 🎉

Thanks for addressing the comments!

[PolarBean]

Thanks for your review!