Problem:

All vertices, regardless of rank, have 0 returned for mu->PrimaryVertices()->nCrossCentralMembrane().

Looking directly in the MuDst, mNCrossCentralMembrane is zero.

From 2008:

Figure 1: CM Match Distribution from 2008 (http://drupal.star.bnl.gov/STAR/blog-entry/rjreed/2008/sep/15/ppv-revision-1-29-low-pt-high-luminosity-fmsslow-trigger-evaluation-part-2) bfc options were: '"DbV20080712 B2008a ITTF IAna hitfilt ppOpt VFPPVnoCTB l3onl emcDY2 fpd ftpc trgd ZDCvtx NosvtIT NossdIT Corr4 analysis OSpaceZ2 OGridLeak3D beamLine debug logger" - run in stardev in Oct' 08.  This only looks at index 0 vertices with positive rank.

Figure 2: # of CM TPC tracks from index 0 vertices with positive rank.  Definitely enough statistics that some should have been seen.

Current file under analysis:

 st_W_10095048_raw_1180001.daq

Library SL09b

'bfc options "VtxSeedCalG pp2009a ITTF BEmcChkStat QAalltrigs btofDat Corr3 OSpaceZ2 OGridLeak3D beamLine -VFMinuit  VFPPVnoCTB -dstout evout

Look at Single Event (from log):

StiMaker:INFO  - PPV:: # of input track          = 2069
StiMaker:INFO  - PPV:: dropped due to flag       = 0
StiMaker:INFO  - PPV:: dropped due to pt         = 842
StiMaker:INFO  - PPV:: dropped due to PCT check  = 322
StiMaker:INFO  - PPV:: dropped due to DCA check  = 553
StiMaker:INFO  - PPV:: dropped due to NHit check = 119
StiMaker:INFO  - PPV:: # of track after all cuts = 233

StiMaker:INFO  - PPVend  eveID=19,  list of found 6 vertices from pool of 233 tracks
Vertex ID=1 nUsedTrack=12 gPtSum=21.701 Lmax=1.0027e+06 match: any=5-3 CTB=0-0 BEMC=5-3 EEMC=0-0 TPC=0-0  Vz=33.35 +/-0.1

StiMaker:INFO  - PPV::fit() nEve=1 , 6 traks with good DCA, matching: CTB=0 BEMC=5 EEMC=1
StiMaker:INFO  - PPV::buildLikelihood() pool of nTracks=233
StiMaker:INFO  - PPV::buildLikelihood() 6 tracks w/ matched @ Lmax=2693.700560
StiMaker:INFO  - PPV:: iLow/iMax/iHigh=2824/2834/2844
 accM=10.000000  accW=24.500000  avrW=2.450000
  Z low/max/high=32.349998 33.349998 34.349998, kSig=14.828822, sig=0.067436
 

In isPostCrossingTrack every single track that has a z change sign is labelled as being a PCT with 2 or more hits in the wrong z.  Based on previous statistics, we would expect to see ~1 out of every 3 index 0 vertices with at least 1 TPC CM track.  Looking at the r,z,sector of each hit in these "PCTs", in every single case, the sector does not switch from East to West (or vice versa) when crossing the center.  This is even the case if the number of wrong hits required for PCT is increased to 20.

StiMaker:INFO  - PPV:: # of input track          = 2069
StiMaker:INFO  - PPV:: dropped due to flag       = 0
StiMaker:INFO  - PPV:: dropped due to pt         = 842
StiMaker:INFO  - PPV:: dropped due to PCT check  = 35
StiMaker:INFO  - PPV:: dropped due to DCA check  = 746
StiMaker:INFO  - PPV:: dropped due to NHit check = 161
StiMaker:INFO  - PPV:: # of track after all cuts = 285

StiMaker:INFO  - PPV::fit() nEve=1 , 6 traks with good DCA, matching: CTB=0 BEMC=5 EEMC=1
StiMaker:INFO  - PPV::buildLikelihood() pool of nTracks=285
StiMaker:INFO  - PPV::buildLikelihood() 6 tracks w/ matched @ Lmax=2693.700560
StiMaker:INFO  - PPV:: iLow/iMax/iHigh=2824/2834/2844
 accM=10.000000  accW=24.500000  avrW=2.450000
  Z low/max/high=32.349998 33.349998 34.349998, kSig=14.828822, sig=0.067436
 found  PPVertex(ID=1,neve=1) z0 =33.35 +/- 0.07
 

Interestingly enough, the same vertex is found in both cases.....

Problem:  Given luminosity, the number of real PCTs is much greater than # TPC CM tracks.  Assume that a vertex corresponding to a real collision near the central membrane can be found.  Look at all tracks that appear to extrapolate to that vertex which do not fail for othewr reasons (nHit, etc.)  Some of these tracks should cross the membrane. 

Current task:  Find ~5 such vertices in different events.   View event.  Examine all associated tracks.