From: Tracy Usher [usher@slac.stanford.edu]
Sent: Thursday, April 13, 2006 1:07 PM
To: Strickman, Mark
Cc: Benoit Lott; Chekhtman, Alexandre (Forn Natl); Dubois, Richard; Eric
Nuss; Frederic Piron; Grove, Eric; Johnson, Neil; Lavalley, Claudia;
Makeev, Andrey (Forn Natl); Mark Strickman; Patrick Smith; Richard
Hughes; Thierry Reposeur
Subject: McIntegratingHits


Hi Mark,
  I reminded myself of what was done a few months back with the 
aim to help Andre in his studies. Basically, McIntegratingHits contains two
maps: one relating a pointer to an McParticle to the energy deposited 
by this particle in the crystal, the other relating the PDG particle 
id of the particle to the energy it deposits in the crystal. The former 
is what you might normally like to us, unfortunately, this is not filled 
unless you are running G4Generator in "full" mode (i.e. keeping all the 
McParticles in the TDS) as opposed to the standard "prune" mode. 
Unfortunately, trying to run "full" mode when simulating heavy ions may 
not be possible, I can imagine that a typical event might result in 
an McParticle tree so large as to eat nearly all the available memory...
  So, you are probably stuck using the PDG particle id to deposited 
energy map. An issue here is the numbering scheme for heavy ions. 
I don't remember the details but as I recall, PDG doesn't define 
heavy ions so G4 makes its own ID based on the ions Z and A. 
The solution to this problem, I believe, is for us to define our 
own heavy ion table instead of letting G4 create it on the fly, 
but that idea didn't seem to take wings when midly proposed a 
couple of months ago.
  Ok, now for the bad news. The above information exists only in 
the TDS version of McIntegratingHit. The PDG id to deposited energy 
map is NOT written out to the root version of McIntegratingHit. 
This could be easily implemented if that is important.
  It is probably worth trying to play with the two methods of looking 
at the energy maps, the first to see if there are really events which 
would cause the McParticle tree to blow its stack and the latter to 
see if it is useable.

  Tracy