At the January 2026 collaboration meeting, it was agreed that the basic workflow for most oscillation analysis toolkits (MaCh3, GUNDAM) will likely be to create standalone "friend tree" style files with NuSystematics-generated weights in them. These would be made by Jaesung Kim's cafnusyst. However, this would be almost unusable for CAFAna, which expects the relevant information for a given event to be in its file.
It was agreed that cafnusyst could also be configured to output CAFs that were updated to include the syst weights it's outputting, but that would need to be done by the users that need it (i.e., for the CAFAna analysis).
At the January 2026 collaboration meeting, it was agreed that the basic workflow for most oscillation analysis toolkits (MaCh3, GUNDAM) will likely be to create standalone "friend tree" style files with NuSystematics-generated weights in them. These would be made by Jaesung Kim's cafnusyst. However, this would be almost unusable for CAFAna, which expects the relevant information for a given event to be in its file.
It was agreed that
cafnusystcould also be configured to output CAFs that were updated to include the syst weights it's outputting, but that would need to be done by the users that need it (i.e., for the CAFAna analysis).