Filtered Dijet Simulation Request
I posted some details on the filter performance here. I would like to make a simulation request using this filter, which focuses on dijets in the barrel with the dijet pT thresholds at 7 and 10 GeV.
I have attached the proposed kumac.
The simulation will be for the 2009 geometry.
There is still software needed for the filtered simulation to go forward:
- StPythiaEvent - needed to write events rejected by BFC filter
- StTriggerFilterMaker - filter in BFC for trigger
- Upgrade of PYTHIA to fix PYTUNE
I'm proposing running the full filter for 2 < partonic pT < 9 GeV, only trigger filter for 9 < partonic pT < 15 and no filter for 15 < partonic pT.
For each of the partonic pT bins, the following table shows the how many events will be produced in each file type, file size, and how long they'll take to achieve L = 1 pb^(-1):
| Bin | Pythia Filter Accept | Total Filter Accept | Events in pythia.root files | Events in MuDst.root files | pythia + MuDst size (GB) | time (s) |
time/event passing PYTHIA filter (s) |
| 2-3 | 4*10^-4 | 4.9*10^-6 | 3.24*10^6 | 3.9*10^4 | 18.8 | 1.6*10^8 | 47.43 |
| 3-4 | 2.5*10^-3 | 4.5*10^-5 | 3.25*10^6 | 5.9*10^4 | 19.5 | 5.5*10^7 | 15.87 |
| 4-5 | 6.5*10^-3 | 2.2*10^-4 | 2.0*10^6 | 6.8*10^4 | 13.3 | 2.4*10^7 | 10.7 |
| 5-7 | 2.2*10^-2 | 1.6*10^-3 | 3.1*10^6 | 2.2*10^5 | 23.2 | 2.4*10^7 | 5.94 |
| 7-9 | 0.10 | 1.2*10^-2 | 2.3*10^6 | 2.7*10^5 | 18.0 | 1.7*10^7 | 3.4 |
| 9-11 | 1 | 0.14 | 5.5*10^6 | 7.9*10^5 | 41.5 | 5.7*10^7 | 3.1 |
| 11-15 | 1 | 0.30 | 2.2*10^6 | 6.6*10^5 | 23.4 | 3.3*10^7 | 4.8 |
| 15-25 | 1 | 1 | 3.9*10^5 | 3.9*10^5 | 9.1 | 1.4*10^7 | 4.6 |
| 25-1000 | 1 | 1 | 1.1*10^4 | 1.1*10^4 | 0.3 | 3.8*10^5 | 4.5 |
The total footprint for ends up being ~170 GB and the total time ends up being ~3.9*10^8 seconds (~12.4 CPU years)
Simulation request settings:
Type of request: simulation with filtering
Number of events: read from table above, the request form doesn't really make sense for filtered request
Field: reversed full field
Collision type: pp200
Centrality: n/a
BFC tags: We need some updates to Library, so this part is not entirely known
Production Tag: needs new library
Geometry for Simu: 2009a
Geometry for Reco: 2009a
Library:
Vertex: Reco vertex with VFPPVnoCTB
Pileup: no
MC Event Generator: use attached kumac
- mattheww's blog
- Login or register to post comments