New binning cirteria to separate trials #32
Description
We came up with a new trial separation strategy to deal with the trial change being on the middle of a bin.
Before, the bin where the trial changed would be duplicated so it could be found both in the last bin of the previous trial and the beginning of the new trial, which makes sense because that bin contains info for both trials. The issue is that, then, some bins are duplicated in pyaldata and it becomes complicated combining the trials and working with the full data. Plus, having duplicities in the data feels like an important source of errors.
Now, we want the binning criteria such that the changing bin will be the first bin of new trial only, to avoid duplicities. This way, if you time-concatenate all the trials in the session, you will recover the original data (without any duplicities).
Furthermore, small clarification that the fields "idx_trial_start" and "idx_trial_end", are not python indexes. They are the first and last bins of that trial, so if you want to use them as index for python, you should do [idx_trial_start:idx_trial_end+1]