[G12] Minutes from today's meeting (10/29)

[Craig Bookwalter]

Hi folks,
  Minutes from today's meeting can be found at the following link and
pasted below:

http://clasweb.jlab.org/rungroups/g12/wiki/index.php/27Oct09#Agenda


Agenda

    * Pass1 status (JLab, FSU)
    * Craig's troubles from last week: solved
          o he used just skim 2, which did not include high-momentum pions
(these are all in skim 1)
          o t distribution from skims 1+2, beam energy cut included
          o e beam distribution from skims 1+2, all timing cuts and vertex
cuts applied
    * Craig's new problem: MC t distribution does not have same slope as data
          o plots from 1M MC events (check out 3piFitTSlope.png)
          o the fit produces a t-slope of -3 something
          o the data has a t-slope of -1.027
          o yes i specified this t-slope in my genr8 file (it's in the
above directory)
          o I also don't see the rho in the pi-pi mass (it's in
`std/masses/m_pip1pim.png' and `std/mass/m_pip2pim.png')
    * Diane's pass1 plots ~22% of pass1 files
          o P Pi+ Pi-
    * Johann's Ground State Cascade total yield estimate (using 7.0% of data)

Minutes

    * Attending: Dennis, Johann, Mike, Diane, Craig, Paul, Sasha
    * A vote to discontinue the alternating meeting days passed--we will
now meet solely on Wednesdays at 12 noon.
    * JLab pass1 is on track to finish in early January, and that is not
taking into account cooking at FSU. The HPC continues to confound us
in both native 64 bit and 32 bit mode. Johann will give building a1c
in 32 bit mode on the HPC one last try.
    * FSU had a snafu with the HVAC in their computer room, causing some
data loss -- a temperature monitoring system has been put on order.
Restarting the cluster after a hard shutdown also is troublesome
because Rocks wants to reinstall the OS after a hard shutdown. Paul
suggests that we open up npgrid for Monte Carlo instead since the JLab
farm doesn't look fondly upon MC users (except for jixie).
    * Mike is going to benchmark 32bit a1c vs 64 bit a1c on the same
architecture to see what performance gains can be acquired by
compiling in native 64 bit mode.
    * Craig will re-examine the beam energy as a function of:
          o trigger bits
          o run number
          o re-bin according to t_id
    * Craig will try to guess the number that produces the correct t-slope
after acceptance -- starting with the 3pi mass region around the a2
    * Craig will also produce the same plots for the raw mc to make sure
the dalitz plot is uniform
    * Diane is going to adjust her vertex position for the origin of her
two photons and plot theta in the GJ frame to ensure that the EC
photon requirement is not inherently selecting a region in GJ theta.
    * Johann will find someone who cares about how many ground-state
cascades he's getting.
    * Next meeting: next Wednesday 12 noon.