Mat, Jim, Kyle, Fernanda, Keith, Jaroslav, Siva, Michael, Louise
Kyle gave an update on DDT DQ plots. (The note-taker missed the
NearDet Run Selection¶
At DQ meeting, decided to ramp up effort due to timescale to make ND good runs selection.
Michael, Fernanda, Hayes and Jim working with Barnali.
A lot of what Barnali showed had good selections where things were known to be bad.
Run 10391 timing was off.
Distributions look sensible if you restrict to a good runs period.
Items came up during the effort at IU:
In Nearline they mix numi/cosmic triggers so that there is data when the beam is down.
Good runs selection should ONLY look at "numi" trigger data.
Need to pull in Xuebing's empty spill fraction metric. In order to that you need to be applying
a good beam requirement.
There is an issue with some processing - for example, files that are 256b small.
There are lots of rate-based selections.
Cuts are quite aggressive for ND right now, although distributions look better.
Fernanda ran through her slides (doc-db 12055)
Shifters (Jan and Susan) report PE distribution in FD that doesn't match standard: ECL
Do we understand this?
Jaroslav gave an indication that zoom feature would be useful.
It would be useful to have reference plots.
When we made these checklists we were doing commissioning and the number of plots has increased to something that is quite overwhelming.
Nearline provides quick analysis checks for good physics analysis. OnMon should be used for checking basic functionality.
GoodRuns - Date
When we edit checklist, what's the process? Feature changes in ECL propagate into the past too but changes go through Jaroslav!
Would it be possible to come up with a few plots to always have in front of shifters?
Fernanda will put together a front page showing three key plot overviews of the snapshot of FD/ND status.