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

d0 bias of displaced tracking #138

Open tomalin opened 2 years ago

tomalin commented 2 years ago

The fitted d0 of tracks from the displaced tracking is biased with respect to the truth. The PR https://github.com/cms-L1TK/cmssw/pull/124 eliminated some of the bias, but as the plots ( from @cgsavard )that can be found inside the review of that PR by show, some bias remains. This is not understood.

skinnari commented 1 year ago

@cgsavard you mentioned in #124 that you thought another correction is needed, correlated with the z shift. can you please elaborate? is this based on looking at the displaced math from @gershtein ?

cgsavard commented 1 year ago

Yes, we made a correction to d0 which fixed one of the biases seen but not the other. Therefore, I thought another higher order may be needed from the displaced math Yuri calculated. I spent some trying trying out a few orders to see if I could fix it but did not manage to do so. It's possible that some other fix is needed, not corresponding to the displaced math, but I'm not certain.