Eshin Jolly
Eshin Jolly
@ljchang Separate from the issue we were having about hosted datasets on neurovault going missing, currently we can't build our docs because `nilearn` has changed their API and removed `_get_dataset_dir`...
On some macOS and linux builds we're encountering this issue: https://github.com/joblib/joblib/issues/1544 Can't do much right now except recommend using a different python version with nltools
These are bug-fixes but many are enhancements that would be likely candidates for moving into `burnt-ends`. We had also discussed creating a stats output class #172 which should be straightforward...
We should clean up plotting and updated methods to make it easier to tweak the underlying `nilearn` calls. We should also fix-up plotting with `Adjacency`: Some relevant issues: - #378...
We had also discussed how subclassing is handled #277 but this might be informed by the decisions we make in #437 Relevant issues: - [ ] #405 - [ ]...
We had talked about sublcassing at some point #254 which might be worth re-discussing Relevant issues: - [ ] #371 - [ ] #381 - [ ] #383 - [...
Relevant issues: - [ ] #126 - [ ] #150 - [ ] #391 - [ ] #420 - [ ] #434
To leverage newer faster tools (e.g. GPU, parallelization) with potentially nicer APIs, we're interested in exploring different backends or providing a flexible system to swap out backends on-the-fly: Numpy alternatives:...
During this move it would be great to create a "10 minutes to nltools" tutorial, maybe grabbing some notes from your classes @ljchang ? Some of these I have miscellaneous...