>
> Hi Aaron,
>
> Speaking of the vertex generation, I have the following question. When
> programming DGen, Simon decided to use our own triangular distribution
> random number generator codes which we believe out performs the one used
> in blastmc. He modified blastmc such that if event generator could
> overide blastmc's vertex by providing its own ones. Since I have not
> tried to generate any event files lately, I wonder if that is still true
> or not.
>
> I believe that vertex is more of a generator business so should and can
> be done in generators. However, there have been some arguement and
> shuffling around about where vertex should be generated. If the feature
> of overidable vertex is removed, would you please inform me so I can
> remove the codes generating vertex from DGen.
>
>
> Chi
Hi, Chi:
Yes, I agree that whichever subroutine generates the Monte Carlo routine
kinematics should also generate the vertex point. This is, actually, how
the Monte Carlo is now setup. I don't think it's too relevant to go into
all of the details. However, as long as you make a call to the
appropriate vertex routine within your code BEFORE calling the GSKINE
command (which stores kinematic AND vertex information in GEANT's memory
banks), then whatever vertex style you choose should be the one used.
The most realistic vertex method that the Monte Carlo currently has is
vertex style 0. This style has an extended target of total length 40cm
and has a triangular distribution in (cylindrical) rho about this length,
out to a maximum of 0.5mm, centered in the middle of the target.
Aaron
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST