1. test done locally in DEV with files from Yifei.

The files are AuAu MB events with |Zvertex| < 40 cm and sigma_vertex = 40 cm ; run over 1k events

QA PLOTs : here 

There is no pileup hits since the purpose of this production was to create the .event.root to produce the pileup hits

     2. correction in the analysis :

We have seen that some vertex selected have no primary tracks.

It's because in the analysis code, the 'good' vertex was selected using the rank :

 Float_t currentRank = event->primaryVertex(v)->ranking();

 if(event->primaryVertex(ipr)->ranking()!=highestRank) continue;

However, this is not always true.

I tried the proposal from Yifei to remove these vertices :

 StThreeVectorD PrimPos  = event->primaryVertex(v)->position();

      if(fabs(PrimPos.x())<0.05 && fabs(PrimPos.y())<0.05 && fabs(PrimPos.z())<40)

{

 cout <<" current vertex x= "<<PrimPos.x() <<"  y = "<<PrimPos.y() <<"  z= "<<PrimPos.z() << endl;

 goodVertex = v;

}

 QAplots: here

     3. test production in DEV from scratch : here

     4. test in SL08f with updated StiKalmanTrackFinder.cxx : here 

     5. test in SL08f witn new beam pipe : here

     6. test in DEV after fullTOF config (modification of gemetry.g) : here

     7. test in SL08f : final config; here

     8. single tracks plots : here

     (an update with beta taken into account for the sigma_dca is here )

     9. tpc hit errors investigation : here

    10. mini production w/o SSD (w/o IST) in reconstruction chain : here