Project

General

Profile

07/23/2014

Attendance: Adam, Chris, Dominick, Craig, Matthew, Gavin, Nate, Satish

Need meetings soon:
— How to run the different configurations. (can we discuss it here with Jon and Ryan?)
— File size discussion.
— Understand the NuMI processing needs – does it makes sense to add the functionality to only run on the runs with 4 continuous diblocks.

Nate has some file size slides:
— Reco on cosmics currently takes 200 minutes.
— PID could take 5xReco
— A factor of 4 in file size probably works for production.
— Craig’s comments on file-size issues:
— Concatenating file from the NuMI stream (and others) is critical. The shear number of files we have is one of our biggest hurdles, and it doesn’t have to be. It should be fixed online (at the DAQ stage).
— The DDT files so far (and at least in the near future) have a small effect on production. We simply Raw2root them in keep-up mode. They don’t need standard Reco or PID for exotics analysis, and that should remain the case — they will likely need to do most of their own “custom" processing. As long as the total DDT file footprint remains less than the 10 Hz NuMI spill size I think it isn’t a big problem for production. I spoke with Martin and he agrees with these sentiments.
— The required cosmic sample and pre-selection are important choices to nail down. With a factor of 30 (or more) between the cosmic rate and preselection choices, we need to decide what to process going forward and make sure it is reasonable. It could dominate production efforts if we are not careful.

Timeline discussion:
----------------------------

—> What is the status of current efforts:
—> ND CAF with Reco (no PID)
—> Some PID done for ND (20%)
—>

—>  What can we do before September?
— NuMI Reco checks (database and resource needs)
— Update ND PID
— Tuning MC/ND: Several iterations for tuning — someone in the Sim group should do it.
—> We should provide G4 only files
—> They should tell us what they need
—> ND Cosmic:  Coment for Ryan:  We can run full Reco on this — not only through cosmic tracker.  
—> Low beam intensity?  May want to do this immediately after tuning?  Check slicer?  
—> Need low-intensity fcl. 10x lower?
—>  Alignment study:  We can do the alignment study now - tell us what to do with the fcl.  
—> We already have “un-staggered geometry now”.
—>  GENIE + CRY overlay — simultaneous generation — 100,000 spills.    500 files, 200 spills each.  
—> Nate will do it. Brian said it works.
—> We can do it now.
—>   MRE/MRCC/cosmic beam — probably not in the first analysis.  
—> If it isn’t part of the first analysis we would prefer to
—> When Kanika runs it next time she should try to do it in the SAM framework in a consistent way with production.

—> We would like to re-do the FD simulation if possible. As soon as Sim is validated we tag and can re-do the simulation.

Production questions:
—> Is the current MC cosmic statistics good enough?
—> Need to understand the requested statistics.

—> Need to understand how to do the running with different configurations. POT weighting and bad channel masks
—> For cosmic we need three different configurations
—> For beam we need all of them —>

—> Is the far detector MC simulation already considered validated?

—> FD NUMI Reco needs to start as soon as there is a calibration tag. Late september is probably too late and two weeks may not be enough.
— calibration requires bad channels — need to be updated.
— Dominick will do some NuMI data Reco tests.

—> Reco needs to be validated by the end of August?
—> Production group needs to start NuMI processing sooner…