Standardize Data Filtering

[tomeichlersmith]

Following discussions with @jmuse13 , there are several cuts that could be helpful for everyone. How to actually implement these cuts is still up for debate, but this can be a helpful issue for keeping track of our notes.

Channel-by-Channel

Cuts that are removing digi hits from specific channels.

• [ ] Noise hits - cut out channels who are in ADC mode and max ADC < some configurable threshold (e.g. 25)
• [ ] Multi-hits - cut out channels with multiple TOAs or multiple TOTs <- may or may not be used (e.g. very helpful in muon runs, but very unhelpful in runs where there could be a shower)
• [ ] Bad links - many runs had a few links that were not aligned properly and were reading out bad data. drop channels from these bad links to not confuse downstream analysis https://github.com/LDMX-Software/Hcal/pull/49

Event-wide

Cuts dropping entire events.

• [ ] Events with isolated hits - probably event was noise, a cosmic, or a bug, want to drop it any case
  • "isolated" means the hit doesn't have another hit in adjacent layers with same orientation
• [ ] Shower started before first layer - higher-level reconstruction is virtually impossible if shower begun upstream of all sensitive layers.