[Mne_analysis] using the autoreject package

[Neeraj Sharma]

        External Email - Use Caution        

Hi List,

Following up on this - it got fixed.
- on seeing the source code found that the location field in my EOG
channels had NaNs and this was the issue.
- may be good to through a warning.

Best regards,
Neeraj

On Tue, Feb 4, 2020 at 1:07 PM Neeraj Sharma <neerajww at gmail.com> wrote:

> Hi List:
>
> Excuse me for this naive question.
> Does the "autoreject" take a lot of execution time?
> I downloaded the pip install -U autoreject package.
> On using:
> ar = AutoReject()
> epochs_clean = ar.fit_transform(epochs)
> Running autoreject on ch_type=eeg
> 0.00% Creating augmented epochs | ...... Since 20+ mins.
> The epochs object is 32 channel x 1400 epochs x 512 samples.
>
> Should I use the latest version or is something weird in my epoch object.
> The example shared in the website works fine though.
>
> Best regards,
> Neeraj
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20200204/863d0dd7/attachment-0001.html