MicroBooNE Monte Carlo Challenge 5 status and actions
→
US/Central
WH5SW (Boardroom) (Fermilab)
WH5SW (Boardroom)
Fermilab
Description
ReadyTalk connection: 866-740-1260
Code: 8867778
Modification key: uBooneMCC5
Attending Tracy Usher, Leon Rochester, Tingjun Yang, Lynn Garren, Gianluca Petrillo, Yun-Tse Tsai, Herb Greenlee, Wes Ketchum, Eric Church, Kazu Terao, Zarko Pavlovic, Erica Snider
Purpose of the meeting
larsoft_data
Purpose of the meeting
- What is left to do and who is available to do it
- To make sure the project and experiment are on the same page.
- LG: need to know the status of the code so that we can start preparing for the release.
- Also need to know about the status of larsoft_data
larsoft_data
- Expectation is that files there do not change on the time scale of about a month — so more slowly than integration release schedule
- WK and KT note that files may change rapidly during detector commissioning, but otherwise agree that existing content will not change often by the current definition
- The data is similar to calibration data, but not suitable for a DB.
- Discussion of why it’s not in a code repository. uB doesn’t like current solution because it is too hard to change and test
- Current version of files passed to LG should be correct. She will make a new version of larsoft_data immediately after this meeting
-
PFParticle has been merged
- Change to LarPandora that use PFParticle have not yet been merged
- In a branch. Andy Blake is responsible for this. Believe it is ready
-
Outstanding feature branches
- LarPandora code that uses PFParticle
- LarAna code that calculates momentum from multiple scattering
- Eric also has something (changes to dirt filter??) that he is trying to get in. It builds. Will get it in by 4:30
-
We should make an integration tag now??
- Yes
-
KT: simulation is good to go. One minor fcl file change outstanding. Should not change the physics.
- Resource usage within simulation has been checked, and appears to be ok
-
GP has looked at resource usage through simulation. All ok, except G4 occasionally uses > 2 GB (but less than 3 GB)
- KT noted that condor and memcheck provide consistent answers in 100 event samples, but that GP’s measurements are lower, sometimes by as much as 500 MB (though not sure of the exact number). They will work together to understand the difference.
-
Will create new larsoft_data immediately after this meeting using the files previously provided. (Completed during the meeting)
- Will push it to cvmfs
-
Will start cutting integration release today based on content of head at 4:30
- Eric will try to get his changes in by then (completed during the meeting)
- Will expect to cut a production release on Thursday after the Analysis Tools meeting. This should be ready by Friday morning
-
uB will test and inform us of any problems or changes that are needed
- Time is not critical, so can deal with these changes and any needed patch releases over reasonable length of time
- Follow-up: discuss ways to make modifying / testing larsoft_data easier
There are minutes attached to this event.
Show them.