Tal Yarkoni

> 1. then stimulus properties could somehow represent not only details like cropping/full frame, colorspace, etc. I think this is analogous to the movie example, but I still think it's...

> for movie, i'm thinking of things like commercial clips that are shown, and i'm sure that certain clips cannot be shared. I don't know that we can do anything...

As @robertoostenveld notes, there is no current problem representing medication information in a BIDS dataset, because one is free to include any and all relevant variables in `scans.tsv`, `sessions.tsv`, or...

That would be great! Neurosynth has a REST API that should be more or less operational, so you should already be able to grab images pretty easily. E.g., if you...

Thanks, this list is helpful. For most if not all of the above, working with masked operations shouldn't be too hard. E.g., while einsum won't natively do any masking, I...

I was wrong, it's not straightforward. Will leave open, but doubt I'll be able to work on it.

I think the current observed behavior is actually what we want. In principle the `y` values *could* all be exactly `0`; it's just extremely unlikely. Whereas the same is not...

> That sounds good to me! In addition to the initialization flag, do you think it would useful to include dof in the model summary? I'm sure we could also...

That sounds reasonable to me. The current implementation already allows varying n per voxels, so that's not a problem. It's just a question of how to present that to the...

That's cool to illustrate the conceptual point that the two are different, though I'm not convinced it makes sense to add as a PyMARE example. Generally speaking, if you have...