9/13/2012 Notes (DDT Workshop)

Paths and such

Tree Diagram

  • Zukai
  • Track Like
    • Rough Muons
      • Calibration Muons
        • Golden Muons
      • Stopped Muons
    • EM Clusters

Trigger result object knows what the paths were

Can put producers in different paths and it won't replicate work done, as long as their are identically defined modules
goldenmuons: [A,B,C,TrackLike, RMF, CMF, GMF]
zukai: [A,B,C,X-otics]
stoppedmuons: [A,B,C,TrackLike,RMF,SMF]
roughmuons: [A,B,C,TrackLike,RMF]

If rough muons path is run first, A, B, and C ran. If Zukai path is run next modules A, B, and C won't run again.

Filters in order make sense, especially if we have a sample like Rough Muon Filter (RMF), Calibration Muon Filter (CMF), Golden Muon Filter (GMF), Stopped Muon Filter (SMF) not so much.

Every decision you want to name, make a path and add everything to the path. The framework will make sure not to duplicate work.
Independent triggers need independent paths.

Can veto things like:
Zukai: [A,B,C,"!RMF",X-otic]
"!RMF" won't run RMF again, but invert the decision previously reached, assuming it was run before in another defined path.

Other Path Specifications:
"-CMF" run filter but don't care what the result is

Let's talk about ...


Can't instrument detector until second block is up, somewhere between October and November(?).
After that point we can start looking for monopoles, January.

We can run tests on computing hardware today, as this is already available.

Beam related triggers should be ready earlier than beam will come back. Beam back March - May.

Can also start testing on NDOS today.

Collaboration Meeting

Each group should have some pseudo module ready to go and have run it on the detector. Groups: [NuE, NuMu, Calibration, X-otics]
We need to have infrastructure in place, BaseDataProducts and the like, and documentation in place.
Martin looked into speed profiling (NDOS -> FD)


Format will be close to DAQ meetings, report on progress of weekly tasks.
Weekly meetings: Monday afternoons 1-2pm Fermi time.
ReadyTalk: 1-866-740-1260
ReadyTalk ID: 6321175
Mailing list:

DAQ Meetings on Thursdays at 12:45pm. Andrew is generally in attendance, Martin will play backup.

Action Items for first meeting

Gavin, Andrew - Make novaddt work on bluearc so we can develop on the vms, need profile build so we can optimize timing, fix DAQHit, get CVS notifications setup so we have an idea when code is updated.
Jan, Evan - Work on putting together some analysis modules, break apart test module and get it working for Monday
Martin - GroupedHitLists, help clean up test module and implement new BaseProducts
Zukai - Make module use DDTBaseDataProducts, Monopole filter running, analyzer work ongoing
Denis - Get him to write out a slightly different raw data object as simulated far detector data