FIX: Fix bug with arm64 #11865
FIX: Fix bug with arm64 #11865
Conversation
| @@ -62,7 +62,7 @@ Bugs | |||
| - Allow int-like for the argument ``id`` of `~mne.make_fixed_length_events` (:gh:`11748` by `Mathieu Scheltienne`_) | |||
| - Fix bugs with :mod:`mne.decoding` classes (:gh:`11846` by `Eric Larson`_): | |||
| - :class:`mne.decoding.ReceptiveField` now sets ``rf.fit_intercept_`` during ``fit()`` rather than overwriting ``rf.fit_intercept`` | |||
| - :class:`mne.decoding.SlidingEstimator` now returns the same ``dtype`` as ``X`` (usually ``float64``) in ``sl.predict(X)`` rather than always returning ``int`` | |||
| - :class:`mne.decoding.SlidingEstimator` now returns the same ``dtype`` as ``X`` (usually ``float64``) in ``sl.transform(X)`` rather than always returning ``int`` | |||
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
|
... couldn't resist the temptation to add |
| @@ -1,4 +1,4 @@ | |||
| on: [status] | |||
| on: [status] # yamllint disable-line rule:truthy | |||
There was a problem hiding this comment.
Fun fact -- "on" is truthy and actually you can have truthy keys in YAML. So a slightly more correct alternative here would be to change this to "on": .... But this isn't how I've ever seen a .github/workflows example and it's not how they do it in their docs. So it seemed like the safest option was an inline ignore. See also adrienverge/yamllint#430
Going to see if the "slow rendering" was actually a Numba problem as well as fixing recent CircleCI failures on
main