Skip to content

Draft: HLT1 forward tracking with no UT information

Alessandro Scarabotto requested to merge ascarabo_VeloSciFi_forward into master

This is a draft version of the HLT1 forward tracking with no information from the UT, so in practice we extrapolate Velo tracks directly into the SciFi.

Link here to the first presentation in December for a WP2 meeting:

The algorithm focuses on higher momentum tracks P > 5 GeV and PT > 1 GeV, which is the minimum momentum assumption to decide the size of the search windows inside the SciFi. Also, as no charge information of track is provided by the Velo, left-right search windows are opened (positive-negative charge assumption).

From the software point of view, the code is very similar to the "full" forward. Inside the Looking Forward algorithms Views::UT are replaced by Views::Velo (@dcampora and @thboettc had some ideas to make this more flexible). To run the sequence hlt1_pp_default, UT information is also removed from some muon and Calo algorithms.

This is still a draft MR so many comments on the code should still be removed/cleaned.

Next steps to improve the algorithm:

  • Understand the 10% loss in efficiency at high momentum tracks
  • Many candidates can still be rejected at first stages of the reconstruction to both reduce ghosts and improve throughput

cc @abaillyr @gligorov @dcampora @decianm @dovombru

Merge request reports