search

cms-L1TK / cmssw

Fork of CMSSW where improvements to L1 tracking code are developed.
http://cms-sw.github.io/
Apache License 2.0
4 stars 5 forks source link

L1 tk dev 12 0 0 pre4 hph before rebase (not to be merged) #96

Closed Jingyan95 closed 2 years ago

Jingyan95 commented 2 years ago

The crash described in PR#94 is fixed with this new PR. I will create a separate PR to include changes made to ntuple maker that is requested by Chris.

tomalin commented 2 years ago

I've verified L1TrackNtupleMaker_cfg.py runs and gives unchanged results.

tomalin commented 2 years ago

@tschuh points out that for the new KF, the class TrackerTFP/interface/LayerEncoding.h determines the KF layer encoding, so you should check if calling this would significantly simplify the HitPatternHelper.

tomalin commented 2 years ago

I advise making clear in comments which part of HitPatternHelper are only needed to handle the case of the Old KF. (And also say somewhere in the code what is meant by Old KF).

Jingyan95 commented 2 years ago

@tschuh points out that for the new KF, the class TrackerTFP/interface/LayerEncoding.h determines the KF layer encoding, so you should check if calling this would significantly simplify the HitPatternHelper.

I am a bit hesitant at this one because: 1) TrackerTFP package already depends on TrackTrigger to set up the configuration. Is it really a good idea to ask TrackTrigger to use a product from TrackerTFP ? 2) Also that laye-encoding function written by Thomas is binned by eta/cot/z0. I find it unnecessary for HitPatternHelper, which has access to accurate track parameters.

skinnari commented 2 years ago

merged via #146 so closing this.