fix(Arbor): try to reproduce baseline performance (CEPCv4, nnH_gg BMR ~3.7%).

read from LCIO (SimCalorimeterHit + Track): BMR ~3.74%

• script: Examples/options/LCIO_read_Arbor.py
• setup.sh: add updated k4LCIOReader environment

complete simulation and reconstruction (with Arbor) in CEPCSW: BMR ~4%

• script: Examples/options/CEPCV4_simu_reco_Arbor.py
• cellID is still not compatible
• disagreement in low energy range of energy distributions of both charged and neutral PFO

requested review from @lintao

Hi @wangyuexin,

Do you need a new version of k4LCIOReader? If yes, I need to deploy it into /cvmfs. Then, please remove this part.

Hi @lintao,

Yes, a new version of k4LCIOReader is needed, otherwise some collections (TrackerHitPlane) can not be read in. Please help to deploy it into /cvmfs. Thank you!

Dear @zoujh,

In the https://github.com/key4hep/k4LCIOReader repo, the last commit is 9 months ago. Do you have another repo for the new version?

Tao

Hi @wangyuexin,

Jiaheng's repo is moved to https://code.ihep.ac.cn/cepc/externals/k4LCIOReader now. I have deployed it in /cvmfs.

Could you check whether it works or not? If works, please remove the code in the setup script.

BTW: in order to let CI/CD work, please setup your own gitlab runner following this instruction.

Thank you @lintao, it works! Also thank you for your guidance on gitlab runner setup, I will try.

changed this line in version 2 of the diff

Hi @wangyuexin,

If not necessary, I suggest to avoid to use SimHitMergeAlg to merge hits in same cell. At github, I ever submit a PR to fix hug time spend while merging, and CalorimeterSensDetTool has provided the merge.

Another question, is the 4% result obtained after updating TPC tracking?

Best, Chengdong

Hi @fucd,

Thank you for your reviewing, and sorry for my delayed reply due to my confusion between "Start a review" and "Add comment now" in this gitlab merge request review.

For calorimeter hit merge, do you mean it has been implemented in CalorimeterSensDetTool and I can simply remove the part of SimHitMergeAlg in the script? Or should I import CalorimeterSensDetTool?

For BMR of 4%, yes, it is the result obtained with updated TPC tracking. Compared with the old version (BMR ~4.44%), the charged PFO is more consistent in high energy range, while there is still slight disagreement in the low energy range. I will further check some differential distributions at track level (before arbor).

added 1 commit

• 954a17c6 - fix(Arbor):cellID convert

Compare with previous version

added 18 commits

• 954a17c6...0e7da1c8 - 14 commits from branch cepc:master
• 58e1ad10 - Merge remote-tracking branch 'upstream/master'
• dc55cfb7 - fix(Arbor): cellID(layerNum) in G2CD
• fa2d217c - fix bug
• ebe0296e - update the script Examples/options/CEPCV4_simu_reco_Arbor.py

Compare with previous version

closed