RD: implement cone and vertex isolation, FT information and PV track persistence in B24l; rate reductions in B2ll
MR to add parent and child cone and vertex isolation, FT information and PV track persistence to B2MuMu, B2EE (also Same Sign), B24Mu, B24E, B22Mu2E lines
RD isolation implementation status tracking issue: #644 (closed)
Performance
HLT2 calculated with 100k HLT1-filtered minbias events (from expected_2024_min_bias_hlt1_filtered)
Sprucing to be calculated using HLT1-HLT2-filtered minbias events (from upgrade-minbias-hlt2-full-output-Aug2023) though MANY more events needed than used so far (~ 40k)
Bandwidths
| Case | HLT2 bandwidth / GBs-1 | HLT2 change w.r.t master / Gbs-1 (%) |
Spruce bandwidth / GBs-1 | Spruce change w.r.t master Gbs-1 (%) |
|---|---|---|---|---|
| Master | 0.823 | N/A | N/A | |
| !2948 (merged) | 0.821 | -0.002 (-0.24 %) | --- |
Rates
| Case | Channel | HLT2 rate / Hz | HLT2 change w.r.t master / Hz (%) |
Spruce rate / Hz | Spruce change w.r.t master / Hz (%) |
|---|---|---|---|---|---|
| Master | BToEE | 11.40 +/- 11.40 | N/A | N/A | |
| Master | BToMuE | 22.80 +/- 16.12 | N/A | N/A | |
| Master | BToMuMu | 11.40 +/- 11.40 | N/A | N/A | |
| Master | BToHH | 410.40 +/- 68.42 | N/A | N/A | |
| Master | BTo4E | 57.00 +/- 25.49 | N/A | N/A | |
| Master | BTo2Mu2E | 0.00 | N/A | N/A | |
| Master | BTo4Mu | 0.00 | N/A | N/A | |
| !2948 (merged) | BToEE | 0.0 | - 11.40 +/- 11.40 (-100%) | 0.00 | |
| !2948 (merged) | BToMuE | 11.40 +/- 11.40 | - 22.80 +/- 16.12 Hz (-50%) | --- | |
| !2948 (merged) | BToMuMu | 0.00 | - 11.40 +/- 11.40 (-100%) | --- | |
| !2948 (merged) | BToHH | 410.40 +/- 68.42 | 0.00 (0%) | -- | |
| !2948 (merged) | BTo4E | 57.00 +/- 25.49 | 0.00 (0%) | --- | |
| !2948 (merged) | BTo2Mu2E | 0.00 | --- | ||
| !2948 (merged) | BTo4Mu | 0.00 | --- |
Deprecated bandwidth numbers (to be removed before merging)
HLT2 bandwidths calculated on turbo_lines from Hlt2Conf.lines.rd, using https://gitlab.cern.ch/-/snippets/2694
Spruce bandwidths calculated on sprucing_lines from Hlt2Conf.lines.rd, using https://gitlab.cern.ch/-/snippets/2918
In each case, 10000 events were used. Changes with respect to master are expressed as absolute differences and as a fraction (the ratio of the difference to the bandwidth in master).
| Case | HLT2 bandwidth / GBs-1 | HLT2 change w.r.t master / Gbs-1 (%) |
Spruce bandwidth / GBs-1 | Spruce change w.r.t master Gbs-1 (%) |
|---|---|---|---|---|
| Master | 0.2499 | N/A | 0.2451 | N/A |
| !2948 (merged) | 0.4992 | +0.2493 (+99.76%) | 0.3257 | +0.0806 (+32.88%) |
| !2948 (merged) w/o vertex isolation | 0.4801 | +0.2302 (+92.12%) | 0.3257 | +0.0806 (+32.88%) |