You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
FTO calculations are sophisticated and we don't want to replicate or backport what we do in scikit-talk in talkr. Instead, we should be interoperable.
Ideally, we can load the scikit-talk method in talkr and give people proper FTO calculations. Those can be used for basic quality control and sophisticated visualisations of overlap and turn-taking dynamics.
Second best option is to supply a simpler relative timing measure that is NOT the full and proper FTO and so should not be called that. Proposal: call it offset_simple and flag it clearly as not to be used as FTO. This should suffice for basic quality control but cannot be trusted for scientific visualization.
Least favourite from the POV of code redundancy but perhaps most user-friendly would be to replicate the full functionality of FTO calculation in R. I do think there is an argument to be made that you should be able to redo and adjust FTO calculations in R.
The text was updated successfully, but these errors were encountered:
FTO calculations are sophisticated and we don't want to replicate or backport what we do in scikit-talk in talkr. Instead, we should be interoperable.
scikit-talk
method intalkr
and give people properFTO
calculations. Those can be used for basic quality control and sophisticated visualisations of overlap and turn-taking dynamics.offset_simple
and flag it clearly as not to be used as FTO. This should suffice for basic quality control but cannot be trusted for scientific visualization.The text was updated successfully, but these errors were encountered: