missing intervalsms workaround

[neuromusic]

No description provided.

[neuromusic]

ok this PR is ready to review. it should do the following...

• fixes cannot assign timestamps to all encoder values #226
• adds a log of logging messages to core_data['log']
• adds a "validate_no_read_errors" function that returns False is there are any errors logged while reading the file (such as the one behind cannot assign timestamps to all encoder values #226)

I would appreciate reviews & try-it-out testing from others (on the pkl files in #226) before we merge it.

cc @dougollerenshaw @Ryval @nicain @derricw

[dougollerenshaw]

Great! Taking a look right now.

[dougollerenshaw]

This looks good to me. I was able to open one of the previously failing files and make a summary plot. Behavior looks exceptionally good for this mouse on this day:

The only red flag I can see is when I plot running data. There are some big negative transients in the running speed plot:

I haven't been looking at the running speed regularly enough to know whether these transients are happening regularly. But they're almost certainly not representative of the animals actual running behavior. Could this have something to do with the missing data in the file?