problem with tracker hit associations (VXDTrackerHitRelations)?

[doloresgarcia]

using the nightlies, the last CLDConfig and CLD_o2_v06, VXDTrackerHitRelations (as well as other relation collections for the tracker hits) rec and sim index are filled with -2s

an example file can be found here:
/eos/experiment/fcc/ee/datasets/CLD_tracking/condor/Pythia_eval/Zcard_CLD_v3/1/out_reco_edm4hep_edm4hep.root

[Zehvogel]

I would suspect this check

https://github.com/key4hep/k4EDM4hep2LcioConv/blob/73793abf6794c48e27c0a584e645d2f0a0377f48/k4EDM4hep2LcioConv/include/k4EDM4hep2LcioConv/k4Lcio2EDM4hepConv.ipp#L856

Still need to figure out why it fails though, maybe something in the translation mappings has changed unexpectedly recently @tmadlener?

[tmadlener]

When did it start failing? My suspicion would be that it's an issue related to TrackerHitPlane vs TrackerHit in the filling of the association. We have removed the association from TrackerHitPlane to SimTrackerHit in EDM4hep (key4hep/EDM4hep#331). The corresponding changes in the converter are: key4hep/k4EDM4hep2LcioConv#73 and key4hep/k4EDM4hep2LcioConv#76

[Zehvogel]

I think I found one suspicious bit but I don't think that it was the source of this error key4hep/k4EDM4hep2LcioConv#83

As there are still associations created but just default initialised (which I would think from the -2) this check should not have been the problem.

Does it maybe just not work because the typeMapping.trackerHits that is also passed for TrackerHitPlanes contains a mapping to TrackerHit3D and not to the TrackerHit interface? I think that would fit with the above error :)

[andresailer]

Does it maybe just not work because the typeMapping.trackerHits that is also passed for TrackerHitPlanes contains a mapping to TrackerHit3D and not to the TrackerHit interface? I think that would fit with the above error :)

indeed