Attempting to fix symmetric R3 fullchain ART test
I tried to patch few things just to make it work but maybe few things could be better aligned.
1] I find there is a bit of inconsistency in nomenclature of certain branches.
For instance:
MM_SDO but MMRDO (MM_RDO?)
N_Digits_MM but nSDO_MM
(for other see changes)
2] Also there are no Sim hits and digits collection passed to RDO file, so the validation check alg (checkNSWValTree.py) that runs after reco step failed because it expected those. Apparently this is expected for them to not be passed down (I can imagine because of the size, not sure if we want to have them at least for validation or downstream or not). I created flags to only check these collection when necessary and available.
3] Looks like RDO container is needed/requested for the MMPRDVariables but the key has not been set to anything and there is no way how to configure it outside of the class. Not sure what the most optimal thing is here but for it to be configurable I extended this class for an extra argument. Let me know if you prefer a different way
4] I guess when the createDCubeRecoHistograms_withSel.py was made it only focused on TGCs and CSCs because I don't see MM and sTGC, nor RPCs and MDTs. In fact, it is a bit strange as currently there is a weird combination of container collections in the expected input file (No PRDs or RDOs for MDT or RPC yet they have the SDOs which MM and sTGC does not have? I am a bit confused what is expected here). Below is an example of what containers are present by grepping for localPosX:
Br 21 :PRD_MM_localPosX : vector
Br 44 :PRD_sTGC_localPosX : vector
Br 57 :RDO_MM_localPosX : vector
Br 75 :RDO_sTGC_localPosX : vector
Br 89 :SDO_MDT_localPosX : vector
Br 109 :SDO_RPC_localPosX : vector
Br 125 :SDO_TGC_localPosX : vector
5] I managed to run both thread 1 and 5 reco tests but the following comparison with serial and 1vs5 tread show quite a few inconsistent branches (for more see the logs). Is this expected?
Tagging potential interesting parties @jojungge @minlin @mbarel @pscholer