FIX: Update ApplyVDM to handle 4D inputs #3500
Conversation
|
@effigies I tried to extend the apply VDM functionality to imaging data containing multiple frames (e.g. fMRI). this is a tiny modification. however, despite the fact that these modifications work for my private data, my commit does not pass the tests (see commit e8fe958). I am a bit puzzled because none of the tests concerns the modified file. furthermore, when I revert the changes (see commit 90385cf), the original commit also does NOT pass the tests, producing the same errors. any hints on how I should handle this? |
|
Hi, thanks for this. Could you try rebasing on the latest master? I suspect the issues you were seeing are related to an incompatible traits release. |
Codecov Report
@@ Coverage Diff @@
## master #3500 +/- ##
==========================================
- Coverage 65.25% 65.25% -0.01%
==========================================
Files 309 309
Lines 40859 40861 +2
Branches 5379 5380 +1
==========================================
Hits 26663 26663
- Misses 13122 13124 +2
Partials 1074 1074
Flags with carried forward coverage won't be shown. Click here to find out more.
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. |
nipype/interfaces/spm/preprocess.py
Outdated
| in_files = InputMultiObject( | ||
| ImageFileSPM(exists=True), | ||
| field="data.scans", | ||
| traits.Either( | ||
| ImageFileSPM(exists=True), traits.List(ImageFileSPM(exists=True)) | ||
| ), |
There was a problem hiding this comment.
Note that InputMultiObject is already a single file or a list of files. Is the goal to make it either a single file, a list of files, or a list of lists of files?
There was a problem hiding this comment.
hey, thanks for the hint about rebasing and your quick review. let me have another look at my logfiles before I answer this. for the moment, I am happy the tests were passed.
There was a problem hiding this comment.
ok, so regarding your question:
InputMultiPath = InputMultiObject
OutputMultiPath = OutputMultiObject
therefore the syntax I use is the same as other interfaces:
SliceTimingInputSpec
RealignInputSpec
RealignUnwarp
...
also, without this modification my nipype - workflow was failing.
but you seem to suggest my modification was not needed, and the documentation of the MultiObjects suggests the same.
but then why my wf was failing? could it be related to the incompatible traits release?
I mean:
- with the newest release, no need for my modification (it would actually be a wrong edit - but also for realign and slicetime interfaces?)
- with the old ones (for example version nipype = "1.8.0" or nipype = "1.8.2") it somehow solves a problem
There was a problem hiding this comment.
Is the change around L340 enough?
There was a problem hiding this comment.
@effigies yes! I have done some tests now. so I push anew commit with only changes around line 340?
I guess so.
There was a problem hiding this comment.
can this pull request be accepted then? if not, what remains to be done?
effigies
changed the title
…ames (e.g. fMRI)
Summary
Fixes # .
List of changes proposed in this PR (pull-request)