Move to ctapipe-stage1 for the r0 to dl1 step

[maxnoe]

Opening this issue for discussion what needs to be done to be able to switch to the ctapipe stage1 tool for the r0 to dl1 processing of LST data.

Please comment with any other issues that need to be resolved before we can make the switch.

Things missing / unresolved issues in ctapipe

• Muon analysis as part of stage 1 without having to do the R0 to R1 IO / calibration twice

  • Related PR in ctapipe: Outsourced muon processor ctapipe#1618
  • Related Issue in ctapipe: merge stage1 and reconstruct-muons ctapipe#1353

• Cleaning using thresholds based on pedestals

  • For this we need to fill the monitoring information (event.mon.tel[tel_id].pedestal.charge_std) and implement a new cleaning component that uses this to set tailcuts / other cleaning thresholds.

• Pointing information in the DL1 files

  • Currently, the pointing information is not stored in a good way by the stage1 source. It uses a table structure that is suited for observed data (time / pointing combinations that then should be interpolated to the actual pointing of an event). But because of how the simulations are done, ctapipe tries to be clever and store the unique pointings of the events. For LST, it should just store the monitoring table, not the unique pointings, which are all the pointings, since each event has a slightly different time.
  • ctapipe issues: DataWriter should not write pointing info from each event ctapipe#1562 Store event-wise pointing at least for MC ctapipe#1484 Discussion on separation of Event, Monitoring, and Service/header data ctapipe#1041

Necessary steps in lstchain

• Adapt steps after dl1 to the new format
• Contribute to the general dl2 solution in ctapipe