[FIX] Explicitly strip extensions #115
Conversation
|
@rmarkello it'd be great to get your review, here ! |
emdupre
force-pushed
the
num-vol
branch
3 times, most recently
from
094e4c1
to
6794bf1
Compare
tedana/utils/utils.py
Outdated
| return np.atleast_3d(fdata), data[0] | ||
| fdata = np.atleast_3d(fdata) | ||
| ref_img = check_niimg(data[0]) | ||
| else: |
There was a problem hiding this comment.
I don't think this will work as-is for z-concatenated images. The if isinstance(data, list) will always be triggered because of argparse, so the else clause here never will. I think we can get rid of this whole if-else block and move the code below this else section up under if len(data) == 1.
tedana/utils/utils.py
Outdated
| # create reference image | ||
| ref_img = img.__class__(np.zeros((nx, ny, nz)), affine=img.affine, | ||
| header=img.header, extra=img.extra) | ||
| ref_img.header.set_sform(ref_img.header.get_sform(), code=1) |
There was a problem hiding this comment.
This should probably be moved back next to ref_img.header.extensions = [] so that it triggers regardless of the input data!
emdupre
force-pushed
the
num-vol
branch
3 times, most recently
from
7380280
to
3beb3d9
Compare
Codecov Report
@@ Coverage Diff @@
## master #115 +/- ##
==========================================
+ Coverage 49.58% 49.72% +0.14%
==========================================
Files 32 32
Lines 2033 2035 +2
==========================================
+ Hits 1008 1012 +4
+ Misses 1025 1023 -2
Continue to review full report at Codecov.
|
Closes #82.
Changes proposed in this pull request:
load_datato explicitly strip extensions for both zcat images and listed echo files