Strange value for date in SANS2D event file

[Gesner Passos]

ws = LoadEventNexus(Filename=r'/archive/ndxsans2d/Instrument/data/cycle_13_2/SANS2D00019689.nxs')
proton_charges =  ws.getRun().getLogData("proton_charge")
print proton_charges.firstTime(), proton_charges.nthTime(1), proton_charges.lastTime()

It says that the first event was 1721-04-19 - quite hard to believe ;)

[Gesner Passos]

These values could come from framelog/proton_charge.

[Gesner Passos]

produces the same result

filename=r'/archive/ndxsans2d/Instrument/data/cycle_13_2/SANS2D00019689.nxs'
ws = CreateWorkspace([1], [1])
LoadNexusLogs(ws,filename)
proton_charges =  ws.getRun().getLogData("proton_charge")
print proton_charges.firstTime(), proton_charges.nthTime(1), proton_charges.lastTime()

[Gesner Passos]

The same happens if you change 'proton_charge' to 'period_log'

[Gesner Passos]

It was found that the data file had a strange value.

Open the file (hdfview): ndxsans2d/Instrument/data/cycle_13_2/SANS2D00019689.nxs

Open the dataset: framelog/proton_charge/time

Got to the last value of this dataset and you will find the following value: -1.3019562E10. 

[Gesner Passos]

I got the following answer from the responsible for the data acquisition:

in the instrument log file I see: 
 
    2013-07-29 10:08:13  (5954) NeXusEventCallback: [Warning] (NeXusEventCallback::allFrameCallback) Invalid DAE time value 0 (offset -1.30196e+010 seconds) for frame 6350

this only happened for the final frame in the experiment – times for all other frames are OK.

I tested 100 of files for the same instrument and this failure occurred in just one.

I consider we should not try to fix this right now. Or we should come up with a more general decision on how we will deal with this kind of issues.

[Karl Palmen]

When I loaded the file with the load logs checkbox ticked and then sampled log, I then imported the proton charge log and got a message "Found entries in the proton_charge sample log with invalid pulse time. In viewing the table. I see the first time is a large number. The graph plotted along with it starts with a large negative number suggesting the very early date.

Other files I tried are OK.

I pass this. A new ticket can be opened if this or similar problem occurs with other files/runs or we have a good idea how to deal with such rogue data.

[Stuart Campbell]

This ticket has been transferred to github issue ​8457