1M event campaign: Goal, scale up to the level of days of FD data. Use PDHD simulation and PDHD data, include OSP and SPNG and metric evaluation as in full-chain. May be run by the prompt-processing system itself. This campaign will produce a lot of data useful to a lot of collaborators and should be executed with close coordination with DUNE production team.
This output is expected to be used for some form of AI/ML training and thus depos must be saved (at least). If we can find a reasonable way to save out G4 trajectories, all the better. We should save initial kinematics.
1M event campaign: Goal, scale up to the level of days of FD data. Use PDHD simulation and PDHD data, include OSP and SPNG and metric evaluation as in full-chain. May be run by the prompt-processing system itself. This campaign will produce a lot of data useful to a lot of collaborators and should be executed with close coordination with DUNE production team.
This output is expected to be used for some form of AI/ML training and thus depos must be saved (at least). If we can find a reasonable way to save out G4 trajectories, all the better. We should save initial kinematics.