HF PWG Meeting Minutes by Zhenyu - 2015/05/07

1) MTD Data production - Lijuan
Lijuan presented the proposal to prioritize std_mtd data stream in the next month or so to allow sufficient amount of data produced before the Stony Brook Collaboration week, so that we can decide whether or not one can submit MTD abstracts to QM15. In a study done by Shuai Yang, Omega, rho and Jpsi peaks are observed with 365M MB events, and it is expected that a factor of 50 increase in statistics with the full Run14 dimuon data. Lijuan projected that one month of data production with 90% of CPUs devoted to st_mtd data production, will allow 20% of full std_mtd data produced, i.e, 0.2*3.2B events processed or 0.2*2.0B events produced. The proposal was unanimously accepted at the HF PWG meeting and an email was sent to Zhangbu, and PACs to support the proposal.

2) PicoDST production - Xin
Xin presented the status and plan of the picoDST production. A small bug in st_physics picoDST production was discovered, which led to about 5% event loss for MB and 10% overall. Some changes needed for MTD deltaT calibration or requested by other PWGs are also incorporated in the new picoDST production. It is expected that by the end of the week, the picoDST production should be able to catch up with the MuDST production. 

3) J/psi polarization analysis in pp 500 GeV - Barbara
Barbara presented updates on Jpsi polarization in Run11 pp500 GeV. She investigated the DsmAdc vs Adc0 correlations, using Run11 online DB, where some of the towers have different relationship between these two variables. It was suggested to plot DsmAdc vs pT and Adc0 vs pT to check which variables are problematic, and check if the simulation can reproduce the observed feature. It was suggeted to ask the trigger and EMC experts to see if they have any insight. 
Barbara also presented signal extraction using both bin counting and fit methods. A question was raised why the former has better uncertainties than the latter and Barbara will look into it.

4)  D0 reconstruction using KFParticle - Amilkar
|VpdVz - VzOffline| cut - add distribution to see what is the width and how many sigma represents this cut. Rosi commented  -it is ~1 sigma, we could use |VpdVz - VzOffline|<6 cm
Cuts values - taken from Jonathan, based on his simulations - link will be added in the slide for a reference. Not much signal at low pT, needs some work to tune cuts.
slide 3,4 - different resolution for D0, D0-bar, this could due to fluctuations, will be checked when more data available
KF vs helix swimming resolution - it should be better for KF particles (based on simulations done by Michale). Will be confirmed when high stat. results are available.
slide 6 - there are correlations between variables and a rectangular cuts may be inefficient. TMVA could be used (but it has to be done with care to avoid issues which we have seen in the talk from Jonathan ~3 weeks ago, namely creation of fake signal)

The rest of presentations were postponed to the next week.