Now that upward muon lists are being exchanged, I can look at the
results of my analysis and see what kind of problems are present. In
looking at the upmus present in the MONTARULI/RONGA sample, but not in
my sample, I found they primarily came in three classes. Two of the
problems should not have existed, and the third is due to the
differences in the way tracking is performed.
problem 1:
nfs mount on our Unix cluster can be a little fussy. On occassion, the
links between machines will drop out. When this happens, the data disks
are not not necessarily visible to the machine processing the data.
This is stupid. This should not be a problem. I have to be more
careful and make sure that all the runs I expected to be analyzed are in
fact analyzed.
problem 2:
VAX and SGI treat logicals differently. I am running both at Gran
Sasso and locally. Except for a couple of file handling issues, the
software is the same. I was trying to do some arithmetic with logicals,
but the actual value of a TRUE is different on the VAX and SGI. I had
seen and corrected this once before, but it managed to sneak back in.
I have eliminated all older software versions with this problems, so
this should not come up again.
problem 3:
In my analysis, I apply a cut on the number of expected central streamer
tube planes in the track, or the number of lateral planes actually
present in the track. For some of the tracks near cracks between
modules, a track can fail this requirement. Also tracks between a
horizontal and vertical scintillator box at very large angles might get
lost due to this cut.
I am reprocessing the data taking problems 1 and 2 into account, and
will post an updated upward muon list.
Colin