Clemens Brunner

Re `MNE_` prefix, what about `SUBJECTS_DIR`?

Hm. I think the problem is that the t/f map is computed by averaging over epochs. Therefore, filling with NaNs to make all epochs the same length would probably work,...

> what do you mean stretch / scale? To compute a t/f map, I need to average over epochs of the same length. If I have variable-length epochs, I need...

Hm, not sure I understand how that would work. I'll try playing around a bit with dynamic time warping (e.g. using [fastdtw](https://github.com/slaypni/fastdtw)). I think DTW could be applied to each...

Actually, it was much simpler than I thought. Basically, I resampled all individual TFRs (for each variable-length epoch) to a common length. Then I average these new epochs to create...

Plus, the x axis labels are incorrect, because in the warped results (bottom row), the data from 0s to 5s are actually the warped time points from cue until the...

😛 yes but this was not the point. I think I'll prepare a short example, but I need to find some open dataset that features variable length epochs. Any ideas?

I think we decided not to integrate this into MNE because it's probably not interesting for most people. However, I have been working on writing up a blog post on...

The EEGLAB example data contains markers that would lead to variable length epochs. What they do in their tutorials is to make them fixed length. So yes, EEGLAB (and MNE)...

Yes, EEG data locked to gait cycles is a great example where this would be useful. I am currently restructuring my blog at https://cbrnr.github.io/ (I'm switching to Quarto), and after...