Q? Epochs default SSP

What is the motivation of having proj=True by default when creating mne.Epochs()?

It’s been a difficult to track annoyance in multiple cases for me 😦. IMO this kind of operation should be explicitly asked by the user.

[larsoner]

@kingjr also doing raw.del_proj() is a one-liner that should get you where you want to be.

We’ll keep this proj=False idea in mind when dealing with #2310. In the meantime I’ll close this.

[dengemann]

It was not so obvious that ssp projs would become obsolete for many. My approach would also be: just delete them from the info if you have them in your data. Working with neuromag data means handling SSP explicitly very often. Maybe we should document this. Of course we have a bunch of such parameters, e.g. baseline, which is rarely useful for time frequency/ resting state etc. and can give surprising results. I would baseline prefer to be None by default. But inertia … On Tue, 18 Oct 2016 at 23:22, Jean-Rémi KING [e-mail] wrote:

For now I’m fine, just explicitly set the Epochs params. It’s more like an API problem IMO, but I understand how the inertia limits this kind of correction.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/mne-tools/mne-python/issues/3664#issuecomment-254643214, or mute the thread https://github.com/notifications/unsubscribe-auth/AB0fisP2DdwlCTh6J99F2xbWDV27jwgxks5q1TiBgaJpZM4KXhCf .