Agenda:

  1. Test Pass2 -  Output, issues, first looks                                              – Rafo + All analyzers.
  2. Test Pass2 -  Do we want/need FEE skims? If so, what selection criteria would be most useful? What purpose would they serve? Do we need them for all runs? – All
  3. Pushing towards a full MC for 2019 and getting reach estimates.      – Matt Solt
    1. Run 7988 Plots
  4. Towards merging the new readout.                                                    – Kyle
  5. Other Business?

Action items:

  • Analyze the existing 10% unblinded data!
  • Rafo will confirm with Stepan the beam energy:  4.55 GeV. 
  • Once that is known, Takashi can start generating the Madgraph input events.
  • We will need to check that a field map exists for that beam energy. If not, generate a new one. Make sure detector is using the correct field map.
  • Matt S. will place the sensors in the correct positions (L1-L3) and verify.
    • There is still some question about the optimal lateral placement of the L0 sensors. Will start with nominal.
  • Once that is done, Takashi can put the generated MC through SLIC.
  • Kyle and Matt G. will verify the new readout code.
  • Omar and Jeremy will update the conditions database.
  • Once these things are done, we can run the slic output through the readout simulation. This step is the most complicated and will probably be the bottleneck.
  • Matt S. will make a driver that masks out some strips in L0 (and L1?) of the SVT. Then Takashi can run the MC through the recon (both nominal configuration, and the one with masked strips).
  • MC production is complete, Matt S. makes a reach plot with intermediate plots

 


 



  • No labels

5 Comments

  1. Norman Graf

    JLab computing experienced a few glitches over the weekend, but the 10% unblinded sample is now essentially done. Analysis of the output is essential before proceeding with the remaining 90%.

    Dst files were produced, but without the new collections. Omar has hpstr working, but would like to run some comparisons before replacing the current dst maker. Will work with Rafo to get this implemented at JLab.

    Matt Solt has analyzed the DST for one run (7988?). Distributions look reasonable (Matt, please post your slides). However, it appears there are more electrons in the top than bottom, which is anomalous. Also, the vertex X distribution is different from what is currently used in the steering file for the beamspot and target constrained fits. Matt will plot Vx, Vy and Vz for a selected number of runs to check the behavior as a function of run. Including a run-by-run beamspot position would involve changes to the code as well as the conditions database. This will not happen before the break. If necessary, we will patch the constrained vertices in a tweak pass over the v0 skims.

    We are currently not skimming FEEs. No clear consensus on whether these are needed. Norman will review.

  2. Norman Graf

    Matt Solt presented his list of things to do before being able to produce an updated 2019 reach plot. (Matt, please cross-post here).

    Kyle summarized the status of the new readout simulation. Matt G. agreed to review it. Kyle will provide documentation and assist in the review.

    Action items are:

    • Rafo will confirm with Stepan the beam energy (4.5 GeV?).
    • Once that is known, Takashi can start generating the Madgraph input events.
    • We will need to check that a field map exists for that beam energy. If not, generate a new one. Make sure detector is using the correct field map.
    • Matt S. will place the sensors in the correct positions (L1-L3) and verify.
      • There is still some question about the optimal lateral placement of the L0 sensors. Will start with nominal.
    • Once that is done, Takashi can put the generated MC through SLIC.
    • Kyle and Matt G. will verify the new readout code.
    • Omar and Jeremy will update the conditions database.
    • Once these things are done, we can run the slic output through the readout simulation. This step is the most complicated and will probably be the bottleneck.
    • Matt S. will make a driver that masks out some strips in L0 (and L1?) of the SVT. Then Takashi can run the MC through the recon (both nominal configuration, and the one with masked strips).
    • MC production is complete, Matt S. makes a reach plot with intermediate plots.

     

  3. Norman Graf

    I looked at the Moller skims.

    A Gaussian Fit to the core of the target-constrained Møller invariant mass gives a mean of 0.04874 (compared to 0.04848 expected) and a sigma of 1.93E-3.

    Will compare to expectations from MC.

     

  4. Maurik Holtrop

    I have generated a field map for 4.55 GeV to go along with the new detector: HPS-PhysicsRun2019-v0-4pt5, which is now available in "master".
    The fieldmap was uploaded to the GitHub fieldmaps area. Please let me know if there are any issues with it.  

    1. Norman Graf

      Thanks Maurik.