Closed danjgale closed 5 years ago
Merging #69 into master will not change coverage. The diff coverage is
0%
.
@@ Coverage Diff @@
## master #69 +/- ##
=======================================
Coverage 91.66% 91.66%
=======================================
Files 6 6
Lines 348 348
=======================================
Hits 319 319
Misses 29 29
Impacted Files | Coverage Δ | |
---|---|---|
atlasreader/atlasreader.py | 97.83% <0%> (ø) |
:arrow_up: |
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 4b6a2e4...9c2dffb. Read the comment docs.
Quick question, after your PR, does it accept filenames, as well as nifti images, like in nilearn
? That would be ideal.
I think I would leave filename
if it also accepts just strings, otherwise this might be confusing to certain users. And I think it's also coherent to nilearn
's approach, no?
It accepts filenames or nifti images, as in nilearn
– I should've been a little clearer with that in the PR! Just edited it above.
nilearn
typically names these parameters with _img
(e.g., NiftiMasker
).
I'm just worried that if we don't call the main input parameter not filename
, users will feel obliged that they load the image first with nibabel
.
But @rmarkello has more experience with nibabel
and nilearn
, let's see what he thinks.
Sorry for the delay in getting to this; a local conference in Montreal had me tied up last week! Just getting back to these things.
This all seems fine to me! Going to go ahead and merge.
Simple fixes for #68. Just changed the documentation for the
filename
parameter increate_output
, and changed default output prefix to"atlasreader"
.Now that I think about it, the use of
filename
is maybe a misnomer. Perhaps we could change it toimg
to show that it is not just a file name, but can also be anyNIfTI-like
object. Thoughts?