Description of the new MVD code

development

author Jeonghwan Park(Yonsei Univ.)
revision Sangsu Ryu modified to fit into new framework PHOOL.

naming convention

The new MVD code written in C++ consists of many MVD own classes. These classes have MVD own naming convention-all classes are preceded by subdetector name "Mvd"(or "Mvc" or "Mvb") to differentiate themselves from classes of other subdetectors. Some base classes have "b" in front of "Mvd"(or "Mvc" or "Mvb").

documentation

New Mvd code is documented automatically using Doc++ which is PHENIX offline document generator. Doc++ in PHENIX

mvd class

user level module

However in the user perspective the code consists of two STAF styled modules-mMvdResp and mMvdReco. These are meant as interface between  MVD own classes and current STAF style PHOOL environment. This is the only way figured out so far to call MVD classes in PHOOL.
  These modules are named to observe PHENIX naming convention-all modules start with "m" and then subdetector specifier "Mvd"(or "Mvb" ro "Mvc") follows. mMvdResp gets ROOT formatted PISA data as input and gives DCM formatted data as its output which is in turn input to mMvdReco.

mMvdResp --  convert GEANT his to DCM format

mMvdReco gets this DCM formatted data which can be either from simulation or real PHENIX data as input and gives STAF styled tables which can be written to DST. These tables are dMvdVertexOut, dMvdMultOut and MvddNdEtaOut. It also makes some useful diagnostic histograms. This histogramming feature is implemented in this module temporarily although it should be replaced by ROOT macro.

mMvdReco -- gets DCM format as input and calculate vertex, dN/dEta and multiplicity
                               incoporate some diagnostic histograms as well


how to setup new code



comments to ryu@p2hp2.lanl.gov
last modified Dec\23\99
Sangsu Ryu