Does the counterpart to time_as_index exist? The event array returned by mne.find_events returns the position in samples and I want to add events with make_fixed_length_events while providing the argument start and stop based on my events. Is there a convenience function to convert from sample to time somewhere in MNE?
Not to my knowledge; I think in all examples, we’d do something like
t = sample_index / info['sfreq']
That’s what I thought. I do get that there is some error checking/tricks in the conversion from time to index that has to be covered in time_as_index; but I’m not sure if it doesn’t apply also to the conversion in the other direction, especially with the origin/first_samp/meas_date system.
Actually I don’t think so, therefore the example I showed should do the Right Thing ™
But for sake of consistency, sure, why not add index_as_time? 
Even though maybe sample_index_as_time() would be more explicit?
Ok, thanks. I’m not familiar enough with the system of origin/first_samp/meas_date for now to make sure there isn’t some hidden logic somewhere and to add this. I’ll keep it in mind for a future me.
Same 
And I think 99% of our users would agree
1 Like
reminds me of a PR when I needed “index_as_time” as well and it wasn’t there 
I had a brief discussion with @larsoner on this: [MRG] ENH, FIX: Add tmin/tmax parameters to get_data methods, fix bug in find_bads_ecg by sappelhoff · Pull Request #9556 · mne-tools/mne-python · GitHub
In that PR it turned out that “index_as_time” wasn’t needed, so we didn’t add it.
1 Like