Reconversion (Oct 2025)

[tch285]

Implementations in derived data needed before reconverting:

• DistanceToBadChannel
• NLM
• Confirm how matched track indices are stored in the derived data format (resolve MatchedTrackIndex not being filled #5 first)
• Clarify what noTrigSel means as opposed to 0
• Time-dependent RCT flag for EMCal bit stored

[tch285]

RCT flags now saved as of AliceO2Group/O2Physics@9bfc210, will have to reimplement the actual selection in analysis. The bits are saved in raw format for JBCs and JCollisions, then checked with the RCTChecker (https://github.com/AliceO2Group/O2Physics/blob/5cb870101460ad343ea52fdc1e487a2990ac5ea5/Common/CCDB/RCTSelectionFlags.h#L54)

[tch285]

NLM implemented in O2 by Florian in PR#14943

[tch285]

After discussion with Nima, a 1 in noTrigSel means it didn't pass any of the JE triggers (except it looks like it's unfilled). Found workaround on analysis level for MB data.

[tch285]

Matched tracks now filled in trees, but more info may need to be stored, like track's selection bit. Also needs some work on analysis level to parse this information properly with uproot.

[tch285]

DBC will not be implemented in the JE derived data until the frequent CCDB calls can be optimized. To be ignored for now.