Draft:Add muonDLL

Add muonDLL into Hlt reconstruction. Needs LHCb!3498 (closed) and ParamFiles!45

added RTA label

marked this merge request as draft

changed title from Fixed formatting to Draft:Add muonDLL

I need some help in making this work as it does not compile yet. I have added a new set of constants where I think they are needed, but the tests run in Moore during the compilation fail in finding the new set of values.

Does this use the CondDB for the MuonDLL? I thought we agreed here Moore#402 (comment 5367016) to put it into ParamFiles, which is easier.

An example of an algorithm doing this would be here: https://gitlab.cern.ch/lhcb/Rec/-/blob/master/CaloFuture/CaloFuturePIDs/src/CaloChargedBremAlg.cpp

I guess you have your new conditions in your own repository? My guess is you have to (manually) set your local conddb environment variable GITCONDDBPATH (that is, in case in your log file ToolSvc.GitSIMCOND won't show your local repo)

I second that ParamFiles would be a better place. In addition I don't understand why you need LHCBCOND, that is currently not used by Run3 software

@marco, I put things in LHCBCOND because I thought that was the place were constants were taken, but that was a mistake. I will inspire from @mveghel to put them in the right place, but I will move all the parameters as they will not change either, not only the ones related to the DLL.

we did, but before putting them into somewhere new I wanted to check that everything worked as expected by simply extending the current place where the constants are taken. But apparently I am not even able to extend the current framework.

I have never worked with Conditions like this, so I cannot really help.

Maybe @mveghel knows more, as he tried to use the CondDB as well initially.

anything I try to add as conditions, even changing the xml file to PARAMFILES, does not work.

I keep on getting the error when compiling Moore MuonIDHlt1AlgPr ERROR ConditionDerivation failed to retrieve PARAM/ParamFiles/data/MuonIDGlobal.xml StatusCode=INVALID_ROOT

This is the same thing I was getting when using SIMCOND, so there must be something profoundly wrong that I am doing, that I am not able to understand here.

Others can correct me, but I don't think you should put the xml file in PARAMFILES, as this is essentially just a folder one can put in whatever one wants. It is not a DB in the sense of the CondDB.

https://gitlab.cern.ch/lhcb-datapkg/ParamFiles/-/tree/master/data

Looking at what other algorithms do, you could:

• Put the context of your xml file in a json file, and then read this in the algorithm to fill the std::vectors (like here: https://gitlab.cern.ch/lhcb/Allen/-/blob/master/main/src/InputReader.cpp#L121)
• Or if you have the original histograms, you can also directly use this (this is e.g. what is done for the Calo, https://gitlab.cern.ch/lhcb/Rec/-/blob/master/CaloFuture/CaloFuturePIDs/src/CaloChargedBremAlg.cpp#L105)

I think it's not a matter of the format of the file, in PARAMFILES there are plenty of different formats. The problem is that whatever I add seems to be not picked up properly (even without adding any new functionality).

I will adopt the first option as there are no histograms for the muonID, just some vectors of numbers. I will keep digging.

Indeed, I don't think it depends on the format. What I wanted to say is that you should access things in PARAMFILES differently from how you access things in the CondDB, i.e. you should explicitly open the file in the code itself.

Also note that I spent some time to fix the other issues with the flags here: Moore!1455 (merged) !2785 (merged) LHCb!3468 (merged)

It might be easier to start from this set of MRs than from yours, as it will avoid that you have to merge them in. I included everything from your latest MR, except the access to the DB / PARAMFILES.

ok, I will start from this

added 184 commits

• 7f516094...35e118ca - 180 commits from branch master
• a0487045 - add MuonDLL
• ac740b05 - change name of conditions
• 3557b3bb - Fixed formatting
• 978bc0b9 - read parameters from paramFiles

Compare with previous version

added 1 commit

• 7f857b41 - Fixed formatting

Compare with previous version

@rvazquez I See that you started committing things to this MR. As said before, I made quite some updates to the class to fix other issues, so you should start from !2785 (merged) to put things in, otherwise merging will be a bit of a pain I am afraid.

yes, don't worry I will take into account your changes

you can also just push to my branch, that might make things easier

or even if you are done, you can merge it and then I will rebase

We can do that, but it depends on quite many things, and I need to modify it a bit to not break the tests. If that is easier for you, I can do it...

No, don't worry once I'm done with my tests I will commit to your branch

added 1 commit

• 1f17f25d - propagate muonDLL info

Compare with previous version

added 1 commit

• 07779f48 - Fixed formatting

Compare with previous version

mentioned in merge request LHCb!3498 (closed)

changed the description

mentioned in issue Moore#421 (closed)

closed