[BUG] Fixed timestamps when there are more than one event

[stuwilkins]

Attn @cmazzoli

[coveralls]

Coverage remained the same at 27.303% when pulling 312171f on stuwilkins:fix_timestamps into 4a67124 on NSLS-II-CSX:master.

[tacaswell]

This will silently drop images that do not have a time stamp which may make it difficult to realign the timestamps with the images later. Maybe something like

timestamps = [i['data'][tag] if tag in i['data'].keys() else np.nan for i in img]

which will fill in np.nan for the missing values. Or, as these are dictionaries coming back

timestamps = [d['data'].get(tag, np.nan) for d in img]

[stuwilkins]

I see what you say, but this does not place np.nan for fccd timestamps that don't exist but rather for events that don't have a tag. Your suggestion would make the output list be of length of all events, not just events that have a tag. Perhaps both behaviors would be best to add with a fill kwarg

[codecov-io]

Current coverage is 27.12%

Merging #48 into master will increase coverage by -0.17%

@@           master        #48   diff @@
========================================
  Files          11         11          
  Lines         304        306     +2   
  Methods         0          0          
  Branches        0          0          
========================================
  Hits           83         83          
- Misses        221        223     +2   
  Partials        0          0          

Powered by Codecov. Last updated by 4a67124...4029428

[coveralls]

Coverage decreased (-0.2%) to 27.124% when pulling 61ed7f2 on stuwilkins:fix_timestamps into 4a67124 on NSLS-II-CSX:master.

[coveralls]

Coverage remained the same at 27.303% when pulling dbef079 on stuwilkins:fix_timestamps into 4a67124 on NSLS-II-CSX:master.