DOC: Document the PET notebook

[jhlegarreta]

Document the PET notebook: add markdown cells explaining the steps.

[jhlegarreta]

Still not true as the data does not use datalad yet, but it will eventually (related to issue #112).

[jhlegarreta]

@mnoergaard I would be grateful if you could review this PR. There is a number of things that I do not fully understand, and may require separate work (i.e. issues and PRs)

• What is the purpose of the below statement ? The split is done internally by PETMotionEstimator and the data_train, and data_test are not used. Maybe it was a simple example ?

data_train, data_test = pet_dataset.lofo_split(15)

• In the statement

model = PETModel(dataset=pet_dataset, timepoints=pet_dataset.midframe, xlim=7000)

Why is xlim given a seemingly arbitrary value? Documenting the class may help clarifying this. Further down the notebook pet_dataset.total_duration is used to set the value for xlim.

• Is the below statement necessary?

model.fit_predict(None)

• There is something missing in the before/after correction plot because the PET_before is not being shown.

• Slightly related to the above, wouldn't it be better if we could propose (or leverage nireports if possible) a GIF with a flickering between uncorrected vs corrected? Otherwise the difference between the two can be hard to tell.

• Are we sure that data_test[0] is the data after motion correction in the following statement?

nifti_img_after = nb.Nifti1Image(data_test[0], pet_dataset.affine)

data_test it is assigned by the lofo_split call but it is never used? Maybe the before_mc and after_mc NIfTI filenames are swapped?

• It seems weird to first make the correction with a given model instance, and then instantiate the estimator with a new model, isn't it?

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 74.00%. Comparing base (19af931) to head (b66a580).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #197   +/-   ##
=======================================
  Coverage   74.00%   74.00%           
=======================================
  Files          24       24           
  Lines        1404     1404           
  Branches      166      166           
=======================================
  Hits         1039     1039           
  Misses        299      299           
  Partials       66       66           

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.