Weights appear not to propagate correctly in some cases with muon splitting on.

Laurie Nevay reporter

changed status to resolved

This was found to be due to a flag of per-secondary weights being turned on in the Geant4 wrapper process. This in turn, weirdly, caused weights to be reset (by default) back to 1.0. This is because switching this in the wrapped decay process affected all particles that use the decay process where there is only one instance of that process.

So the incoming weights were reset to 1.0 only for an action by decay for a particle that wasn't biased and happened to produce a muon. An example we see this for is lamba decays that produce a muon but that were considered rare enough not to include in splitting. We found this example in around 1 in 10^5 events. It explains the high weight muons we saw in FASER monte carlo samples.

This was fixed by switching the flag on and off each time we call the process so that the single decay process remains in its original state.

2023-12-14T16:37:22+00:00

Comments (1)

Laurie Nevay reporter
- changed status to resolved
This was found to be due to a flag of per-secondary weights being turned on in the Geant4 wrapper process. This in turn, weirdly, caused weights to be reset (by default) back to 1.0. This is because switching this in the wrapped decay process affected all particles that use the decay process where there is only one instance of that process.

So the incoming weights were reset to 1.0 only for an action by decay for a particle that wasn't biased and happened to produce a muon. An example we see this for is lamba decays that produce a muon but that were considered rare enough not to include in splitting. We found this example in around 1 in 10^5 events. It explains the high weight muons we saw in FASER monte carlo samples.

This was fixed by switching the flag on and off each time we call the process so that the single decay process remains in its original state.
- 2023-12-14T16:37:22+00:00
Log in to comment