new EpochsTFR class + make _compute_tfr private #8
Conversation
|
@agramfort you're doing this on my branch? shall I be doing something? |
|
@kingjr it's a PR but it's rebased on master. At least you see the diff. there is a bug so far with multitaper "I think" avg_power != power / nave if you have time to look. bed time here |
|
if you want you can fetch my branch and force push my branch to your PR so it's up to date and we'll have travis. that would be cool |
|
a snippet to replicate the mtm power error? |
|
See failing test on my branch On Sun, Jul 31, 2016 at 12:28 AM +0200, "Jean-Rémi KING" [email protected] wrote: a snippet to replicate the mtm power error? — |
mne/time_frequency/tests/test_tfr.py
Outdated
| power = power.average() | ||
| power2, itc = tfr_multitaper(epochs, freqs=freqs, n_cycles=freqs / 2., | ||
| time_bandwidth=4.0, average=True) | ||
| # --- end debug test |
There was a problem hiding this comment.
@kingjr here is the broken test
… mtm to multitaper
No description provided.