Closed oesteban closed 5 years ago
I guess my question is whether it makes sense outside of fmriprep. As it is, we pipe a lot of options in through workflow arguments, which, on a first glance would seem to make it difficult to abstract to usability in another program.
Let's start with 2 for now. I personally would have interest in seeing the anatomical workflows in niworkflows, but it is true that I know how to use them too well. Other users might be clueless.
As we are considering to branch a dMRIPrep out, I guess this makes more sense now.
The anatomical workflow is mature enough to be detached from fmriprep's base code.
This would make the anatomical workflow more visible and will allow: 1) Move the anatomical workflow tests to niworkflows. 2) Cache all the anatomical workflow between circle builds (speed up build time of fmriprep).
WDYT?
(option 2 could happen before 1, that would also be okay)