BF: Ensure BOLD and label orientations are equal

[mgxd]

No description provided.

[pull-assistant]

Best reviewed: commit by commit

---

Optimal code review plan (1 warning)

ENH: Reorient volumetric data to LAS

...iworkflows/interfaces/cifti.py 48% changes removed in doc+sty: suggestions...

     sty: placate py35

     tst: add doctest

     doc+sty: suggestions from review

     ENH: return image if desired orientation already set

Powered by Pull Assistant. Last update 399cc41 ... e1a8a12. Read the comment docs.

[mgxd]

We were assuming the BOLD image and label image had a shared orientation

niworkflows/niworkflows/interfaces/cifti.py

Lines 364 to 370 in bfffaf8

	ijk = np.nonzero(label_data == label)
	if ijk[0].size == 0: # skip label if nothing matches
	continue
	ts = (bold_data[ijk] if ts is None
	else np.concatenate((ts, bold_data[ijk])))
	vox += [[ijk[0][ix], ijk[1][ix], ijk[2][ix]]
	for ix, row in enumerate(ts)]

However, this wasn't always the case - currently, the BOLD image is in RAS while the label image is in LAS. This adds a check to ensure the orientations match up, and if they don't, convert to LAS (HCP Pipelines default, although https://www.nitrc.org/forum/attachment.php?attachid=342&group_id=454&forum_id=1955 recommends RAS space).

And we can see the difference

[codecov]

Codecov Report

Merging #477 into maint/1.1.x will increase coverage by 0.06%.
The diff coverage is 91.66%.

@@               Coverage Diff               @@
##           maint/1.1.x     #477      +/-   ##
===============================================
+ Coverage        62.86%   62.93%   +0.06%     
===============================================
  Files               41       41              
  Lines             4993     5005      +12     
  Branches           724      727       +3     
===============================================
+ Hits              3139     3150      +11     
- Misses            1703     1704       +1     
  Partials           151      151              

Flag	Coverage Δ
#documentation	33.98% <8.33%> (-0.07%)	⬇️
#reportlettests	100.00% <ø> (ø)
#travis	57.78% <91.66%> (+0.08%)	⬆️
#unittests	46.03% <91.66%> (+0.10%)	⬆️

Impacted Files	Coverage Δ
niworkflows/interfaces/cifti.py	54.54% <91.66%> (+3.13%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bfffaf8...e1a8a12. Read the comment docs.

[mgxd]

any lingering concerns?

I'd like to get this out as a 1.1.11 despite it changing the outputs, since it's more of a bug fix imo (cc @effigies)

[effigies]

Seems reasonable. In the long-term a fully-featured reorientation interface would be good to publish in nipype, but no need to go that far in this context.