Tracks to MCP particle link broken when using data in DD4HEP from recent Boole. Potential VP issue.
It seems that, in DD4HEP, the link from reconstructed track to MCParticle is broken when using data from recent Boole master versions of 2025. This worked OK in the Boole of Sep2024.
The problem is seen when running Moore (Panoptes) from the digi files. This calls the method match_track(...) in the file ..Rec/Pr/PrMCTools/src/PrTrackAssociator.cpp In this for the hits (smartIDs) of VP, it fails at the line idlinks.applyToLinks( ids.lhcbID(), &mcParts, &truthCounters { ...}
(It is OK for the hits from the other tracking detectors).
For reference, I attach a copy of this file with some debug prints added and the resulting log files. For this the full software chain Gauss-Boole-Moore is used. I used two Boole versions, one from Sept2024 and another from recent Boole from 2025. Hence there are two log files that are attached. They are both using the same sim file made from Gauss. It can be seen that when the Boole-Sept2024 is used and then Moore is run, it works OK. But when the Boole-2025 is used, followed by running the Moore the failure mentioned above, happens.
This then leads also the printing of 0 in the Boole log file for
PrLHCbID2MCParticle_d0bb6bbd INFO Number of counters : 1 | Counter | # | sum | mean/eff^* | rms/err^* | min | max | | "#removed null MCParticles" | 37738 | 0 | 0.0000 |
This may be a simpler way to monitor this problem, without adding any extra debug prints, if someone is running Moore.
I used the recent master versions of all the packages for the full software chain. I also used the Detector- MR-694 suggested by Tom. This problem happens with/without lhcb-conditions-database- MR-111. For reference I also attach the options file I used to run Panoptes.
I have been discussing this issue with @tlatham and @hcroft recently. We are not aware of a solution yet. Today I saw a message in the 'Simulation mattermost' which says that, @gcorti and others are aware of this problem and they are testing a potential fix. This is good news.
If there is any MR that would fix this problem in any of the packages, please let us know.
For RICH this is important, since the monitoring of the PID performance in simulated data needs this link to work. (I can provide the small sim file/digi file I used for this, if anyone needs it). @jonrob and @zexu for information.
Boole_2025_Test4_PrintVP_MCP_WithLocalDB.log.txt
Boole_Sept2024_Test5_PrintVP_MCP_WithLocalDB.log.txt
PrTrackAssociator.cpp-WithDebugPrint.txt
Thanks for any related information.