Fix SimpleRegressionResults #174
Conversation
Gilles86
changed the title
[WIP] Memory efficient F test
ENH: Replace DataFrame.as_matrix() with .values
corrected wrong voxel selection in spm_plot_auditory
…sion problem - Same error as nilearn/nilearn#1672
…das version - Updated miniconda installer URL to miniconda3 to fix Pandas incorrect version problem (Same error as nilearn/nilearn#1672). - Updated minimum versions of dependencies in nistats/version.py . - Updated versions of dependencies in for TravisCI builds.
… num must be integer) * 'master' of https://github.com/kchawla-pi/nistats: FIX: Use rint instead of ceil FIX: numpy.linspace parameter num must be integer
…_contrast) Fix memory caching issue when doing .compute_contrast in SecondLevelModel
…l.compute_contrast * 'master' of https://github.com/kchawla-pi/nistats: Spurious commit to trigger the disappeared TravisCI Replaced `ols` in comments with `Ordinary Least Squares` in 3 files for clarity rename test case for clarification first_level_model test case - memory caching computation simplify memory caching if statement simplify test case for memory caching fix memory caching not working properly when doing compute_contrast remove unused import
…lt arg) Fixed: param img_filters no longer has a mutable arg ([]) as default
… mutable default arg) * 'master' of https://github.com/kchawla-pi/nistats: Fixed: param img_filters no longer has a mutable arg ([]) as default
…mination) - Renamed manual-cache-timestamp.txt to manual-cache-timestamp
… from scratch) Enable external contributors to rebuild CircleCI from scratch
… X or Y is provided
… X or Y is provided
|
OK gave it another try. The bigger point is that SimpleRegressionResults does not have access to X or Y anymore, so you would have to provide that explicitly to get to residuals or predictions. Maybe SimpleRegressionResults should at least store X? |
|
This one we will have to fix. No need to submit a new PR necessarily, tell me the command you used to merge master into this. |
|
|
What do you think @Gilles86 , this seems like a good candidate for a new PR. |
|
Hi @Gilles86 ! We have incorporated all of Nilstats functionality into Nilearn's master and if you are still interested, i would urge you to reopen this PR there. We are archiving this repo, so it is read-only. I know this PR has been stalled for a while but I'm hoping to see you on Nilearn! |
This fixes https://github.com/nistats/nistats/issues/150