I have put another simulation file into the esc1/calcom/mc repository.
This one has 10,000 events, with a Q^2 cutoff of 0.35, and half-strength
B-field. I ran it with the AUTO 1 option, and the time per event dropped
from 9 seconds per event to 3 seconds per event. (That means you can
calculate this many events in about 8 cpu hours.) There is a documentation
file for this run.
I will try another longer run with a higher Q^2 cutoff. The above run
had the lower cutoff because Celeg reported errors of not conserving
energy. TY gave me a workaround to be able to continue working until
the problem is fixed.
I noticed the egn disk is about 3/4 full. We probably need to think
about cleaning up, i.e. getting rid of some of the old simulations
which are less useful now.
- Will