Draft: Replace the QEE Inclusive HNL Line with a New Version Triggering on T-track Vertex Only

The merge aims to improve the selection efficiency of the QEE inclusive HNL T-track line to the decay B -> X mu (N -> mu pi).

The new trigger only reconstructs the N -> pi mu vertex, as opposed to the previous selection which reconstructs a mu mu pi final state, with the first mu coming from the B meson.

Attached are the projected number of selected events using the old and new selection assuming 10 inverse femtobarn. Assuming no background, the red contour of 3 events reflects the 95% rejection contour that could be placed with a full analysis.NewSelection.pdf OldSelection.pdf

Bandwidths (BWs) were estimated over a sample of 100,000 minbias events following the QEE BW testing guidelines.

If we persist only the rawbank objects and replace the existing (dysfunctional) inclusive turbo lines, we have an expected BW change of:

QEE Turbo BW; -1.9 MB/s (+5.2 MB/s *)

* BW of new line; it should be in full stream, but this is not possible at the moment as full lines require persist_reco=True

If we instead run the line as a typical full line, we expect:

QEE's Turbo BW: -1.9 MB/s

QEE's Fullstream BW: +20 MB/s

Further details on the physics impact can be found at the QEE WG meeting (here), or in the comment chains below