FIX: Refactor of LTA implementation

[oesteban]

This PR makes the LTA i/o implementation more consistent with the others, cleaning up some bugs, completing some missing aspects of the implementation (which help reduce the XFailed tests by 4).

Related: #40.

[codecov-commenter]

Codecov Report

Merging #145 (a5b9752) into master (b3b188e) will decrease coverage by 0.37%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #145      +/-   ##
==========================================
- Coverage   98.66%   98.28%   -0.38%     
==========================================
  Files          12       12              
  Lines        1122     1109      -13     
  Branches      172      173       +1     
==========================================
- Hits         1107     1090      -17     
- Misses          5       11       +6     
+ Partials       10        8       -2     

Flag	Coverage Δ
travis	98.28% <100.00%> (-0.38%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
nitransforms/io/__init__.py	100.00% <ø> (ø)
nitransforms/io/fsl.py	97.97% <ø> (+2.97%)	⬆️
nitransforms/io/base.py	100.00% <100.00%> (ø)
nitransforms/io/lta.py	99.45% <100.00%> (+0.12%)	⬆️
nitransforms/linear.py	97.10% <100.00%> (-0.56%)	⬇️
nitransforms/io/afni.py	89.24% <0.00%> (-7.53%)	⬇️

---

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 b3b188e...a5b9752. Read the comment docs.

[effigies]

Are we conflating FreeSurfer LTAs with something like a per-volume list of affines that one might use to motion-correct a BOLD series?

[oesteban]

I think so, if I understand correctly.

You said conflating because FreeSurfer LTA does not enforce that dst and src are the same for all the affines found in the file, right? while for motion-correct, those should be constant across affines - is that the question?

This has not been introduced by this PR, so I would prefer to address it on a separate issue.

[oesteban]

Opened #147. @effigies - would you be okay addressing this on a separate PR?

[oesteban]

In the interest of making progress at other points and getting rid of the xfail marks asap, I'll go ahead and merge.