Improve throughput for FastJetBuilder algorithm
NEEDS: Moore!4977
To tackle the first point of Moore#969.
The MR aims at changing the way PVs are associated to particles when making jets. The idea is to increase the HLT2 rate by improving this algorithm which takes a considerable amount of time (about 1% of the entire HLT2). The test is based on of the relying to the pointer to PVs already associated to the particles instead of using the IRelatedPVFinder toolHandle with a dedicated algorithm to find the best PV.
The throughput improvement is significant, as with the change the algorithm is 3x faster (processing now on avg at ~1s).
In the majority of the cases, the PV found with the two methods corresponds, but in some cases it differs, as shown in the plot below
See below differences observed in the output log when executing the algorithm
BEFORE
Hlt2QEE__JetBuilderGetSelJetsSimpleJets INFO Number of counters : 4
| Counter | # | sum | mean/eff^* | rms/err^* | min | max |
| "Nb of Inputs" | 1230 | 88252 | 71.750 |
| "Nb of Reconstructed Jets" | 1230 | 6925 | 5.6301 |
| "Nb of Reconstructed Jets with pT>10GeV" | 1117 | 1117 | 1.0000 |
| "Nb of Reconstructed Jets with pT>20GeV" | 102 | 102 | 1.0000 AFTER
Hlt2QEE__JetBuilderGetSelJetsSimpleJets INFO Number of counters : 4
| Counter | # | sum | mean/eff^* | rms/err^* | min | max |
| "Nb of Inputs" | 1230 | 88252 | 71.750 |
| "Nb of Reconstructed Jets" | 1230 | 6934 | 5.6374 |
| "Nb of Reconstructed Jets with pT>10GeV" | 1111 | 1111 | 1.0000 |
| "Nb of Reconstructed Jets with pT>20GeV"