Tightened B vertex chi2 cuts in three- and four-body IDDL spruce lines and...
Tightened B vertex chi2 cuts in three- and four-body IDDL spruce lines and loosened hadron IP cuts in four-body IDDL spruce lines to match the cuts between HLT2 and sprucing
To check rates, I ran over 100000 events using this script: https://gitlab.cern.ch/lhcb/Moore/-/snippets/3292
Output before change:
LAZY_AND: SpruceRD_BuToHpEEDecisionWithOutput #=59741 Sum=346 Eff=|(0.5791667 +- 0.0310459)%|
LAZY_AND: SpruceRD_BuToHpEE_SameSignDecisionWithOutput #=59741 Sum=308 Eff=|(0.5155588 +- 0.0293009)%|
LAZY_AND: SpruceRD_BuToHpMuMuDecisionWithOutput #=59741 Sum=206 Eff=|(0.3448218 +- 0.0239834)%|
LAZY_AND: SpruceRD_BuToHpMuMu_SameSignDecisionWithOutput #=59741 Sum=117 Eff=|(0.1958454 +- 0.0180882)%|
LAZY_AND: SpruceRD_B0ToHpHmEEDecisionWithOutput #=59741 Sum=100 Eff=|(0.1673892 +- 0.0167249)%|
LAZY_AND: SpruceRD_B0ToHpHmEE_SameSignDecisionWithOutput #=59741 Sum=121 Eff=|(0.2025410 +- 0.0183942)%|
LAZY_AND: SpruceRD_B0ToHpHmMuMuDecisionWithOutput #=59741 Sum=56 Eff=|(0.09373797 +- 0.0125204)%|
LAZY_AND: SpruceRD_B0ToHpHmMuMu_SameSignDecisionWithOutput #=59741 Sum=31 Eff=|(0.05189066 +- 0.00931742)%|
LAZY_AND: moore #=59741 Sum=2312 Eff=|( 3.870039 +- 0.0789134)%|
Output after change:
LAZY_AND: SpruceRD_BuToHpEEDecisionWithOutput #=59741 Sum=290 Eff=|(0.4854288 +- 0.0284361)%|
LAZY_AND: SpruceRD_BuToHpEE_SameSignDecisionWithOutput #=59741 Sum=279 Eff=|(0.4670160 +- 0.0278941)%|
LAZY_AND: SpruceRD_BuToHpMuMuDecisionWithOutput #=59741 Sum=183 Eff=|(0.3063223 +- 0.0226093)%|
LAZY_AND: SpruceRD_BuToHpMuMu_SameSignDecisionWithOutput #=59741 Sum=105 Eff=|(0.1757587 +- 0.0171372)%|
LAZY_AND: SpruceRD_B0ToHpHmEEDecisionWithOutput #=59741 Sum=146 Eff=|(0.2443883 +- 0.0202010)%|
LAZY_AND: SpruceRD_B0ToHpHmEE_SameSignDecisionWithOutput #=59741 Sum=160 Eff=|(0.2678228 +- 0.0211449)%|
LAZY_AND: SpruceRD_B0ToHpHmMuMuDecisionWithOutput #=59741 Sum=81 Eff=|(0.1355853 +- 0.0150548)%|
LAZY_AND: SpruceRD_B0ToHpHmMuMu_SameSignDecisionWithOutput #=59741 Sum=52 Eff=|(0.08704240 +- 0.0120654)%|
LAZY_AND: moore #=59741 Sum=2279 Eff=|( 3.814801 +- 0.0783707)%|
Relative difference:
| Decision | Relative Efficiency Change (%) |
|---|---|
| SpruceRD_BuToHpEEDecisionWithOutput | -16.18% |
| SpruceRD_BuToHpEE_SameSignDecisionWithOutput | -9.42% |
| SpruceRD_BuToHpMuMuDecisionWithOutput | -11.17% |
| SpruceRD_BuToHpMuMu_SameSignDecisionWithOutput | -10.26% |
| SpruceRD_B0ToHpHmEEDecisionWithOutput | 46.00% |
| SpruceRD_B0ToHpHmEE_SameSignDecisionWithOutput | 32.23% |
| SpruceRD_B0ToHpHmMuMuDecisionWithOutput | 44.64% |
| SpruceRD_B0ToHpHmMuMu_SameSignDecisionWithOutput | 67.74% |
| moore | -1.43% |
To calculate bandwidth, I assume an input rate of 75kHz and using that the output DST file is 45/48 MB before/after the change. I've been told this is not quite accurate as there is quite a lot of overhead in the DST. Anyway, here are the numbers:
| Decision | Before change (GB/s) | After change (GB/s) | Relative change (%) |
|---|---|---|---|
| SpruceRD_BuToHpEEDecisionWithOutput | 0.0085 | 0.0077 | -9.4 |
| SpruceRD_BuToHpEE_SameSignDecisionWithOutput | 0.0075 | 0.0074 | -1.3 |
| SpruceRD_BuToHpMuMuDecisionWithOutput | 0.0050 | 0.0048 | -4.0 |
| SpruceRD_BuToHpMuMu_SameSignDecisionWithOutput | 0.0029 | 0.0028 | -3.4 |
| SpruceRD_B0ToHpHmEEDecisionWithOutput | 0.0024 | 0.0039 | +62.5 |
| SpruceRD_B0ToHpHmEE_SameSignDecisionWithOutput | 0.0030 | 0.0042 | +40.0 |
| SpruceRD_B0ToHpHmMuMuDecisionWithOutput | 0.0014 | 0.0021 | +50.0 |
| SpruceRD_B0ToHpHmMuMu_SameSignDecisionWithOutput | 0.0008 | 0.0014 | +75.0 |
| moore | 0.0565 | 0.0603 | +6.7 |