Passing simulation to FEDRA
Last updated
Was this helpful?
Last updated
Was this helpful?
Naturally, the first step is to actually launch a MC simulation. I refer to FairShip simulation section for all details concerning FairShip. Currently, the simulations we are performing are:
Test configuration for charmed hadron's propagations (--TrackingCharm), charmed hadrons and other primary particles passed from Pythia to Geant4: charmed hadrons have MCTrackIDs 1 and 2 ( and MCMotherID 0), other primaries have MCMotherID -1. Charm decay daughters should not contain intermediate states;
Proton background simulation (G4 only): a single 400-GeV proton is passed to Geant4 for propagation and interaction. The produced particles have mother ID 0. Need to select ProcID 23 to study daughters of hadronic inelastic interactions.
Two files are produced, a simfile and a geofile, both .root. (actual names of files depend on the simulation and on FairShip version, so I do not report them, for example ship_conical* and geofile_full*).
This step is need to spread the hit on the whole surface, as it was during the 2018 test beam. 5 and 10 spills configurations:
python $FAIRSHIP/macro/runMufluxDigi.py -cs 1 -f simfile.root -g geofile.root -n ntotevents -ns nspillevents
For now, we simulate the same number of events for all spills (in the reality, the number was different also between spills of the same run).
A file simfile_rec.root is produced.
The code which currently does it is
We launch it with $MACROSSHIP/FEDRA/fromFairShip2Fedra("simfile.root"). Parameters should be read from a .rootrc file. A directory b000001 containing empty subdirectories p001...p057 should be prepared. At the end of the scripts, these directories will be filled with converted couples trees.
FairShip2Fedra.nevents 10000 #how many events to convert
FairShip2Fedra.neventsxspill 1000# for SHiP-charm target mover
FairShip2Fedra.nbrick 1 #it will set the brick numbers in the IDs
FairShip2Fedra.nplates 29
FairShip2Fedra.ngrains 70 #placeholder value, it cannot be 0 because it will not read the couples
FairShip2Fedra.minkinenergy 0.03#cut on kinetic energy
FairShip2Fedra.dosmearing 1 #applying smearing in angle
FairShip2Fedra.useefficiencymap 0 #use angular efficiency map instead of fixed value
FairShip2Fedra.maxtheta 1 #maxangle
FairShip2Fedra.emuefficiency 0.85 #fixed value of efficiency in plates
FairShip2Fedra.angres 0.003 #angular resolution
Additional .rootrc parameters for ship-charm beam movement:
FairShip2Fedra.neventsxspill 7000 #how many events to simulate for spill
For simplicity, all spills in the simulation have the same number of protons, even if in the real data there are slight differences
The next steps is to perfom tracking and vertexing according to FEDRA usual application (see FEDRA basic workflow section for details). (Mis)Alignment for now is not simulated, so we just have to launch:
makescanset -set=2.0.0.0 -dzbase=175 -suff=cp.root -dz=-1315
emtra -set=1.0.0.0 -new -v=2
ln -s b000001.0.0.0.trk.root linked_tracks.root
root -l charm_vertexing.C (copied from $MACROS-SHIP/FEDRA)
The vertex file contain only a vtx tree. I am not saving anymore EdbVertexRec (or EdbPVRec) objects due to crash when the files are too big. The tree should contain all the information to analyze and draw the reconstructed vertices.
Currently tracking and vertices are performed in four different quarters. Tracking is performed locally on lxplus, and is done consecutively. Tracks files are stored in subfolders reconstruction_output/firstquarter (secondquarter, thirdquarter,fourthquarter), then vertexing is performed on parallel with HTCondor.
Added a script to perform also tracking on HTCondor, helps avoiding crashes for memory
After vertex reconstruction, it is useful to produce a list with the main features of the vertices we are looking for: that is, charm decay vertices. Each MC event produces two charmed hadrons, which then decay in two separated points. Our task is to reconstruct these decays. We can use MC information to study how well the events are reconstructed by FEDRA.
Storing TrackIDs of charm hadrons and daughters
Identification of primary vertices
Identification of secondary vertices
Extra-tracks (not associated to any vertex)
This step does not involve the reconstructed files, but the original MC true simulation. It will save the lists of charm hadrons and charm daughters ID into a Pickle python file.
It requires only the simulation file and an output file *.p to save the IDs. It is launched as the following:
Given the format of FairShip simulation for SHiP-Charm, particles produced from the primary vertex have MotherId equal to -1 (the two charms, however, have MotherId 0). Then, I can recognize the primary vertices in the MC simulation, by requiring that most of the associated tracks have MotherID equal to -1.
If more than one vertex in the same MCEvent share this feature (primaries associated to different vertices), I save only the vertex with more tracks. Therefore the associated tracks are added to the list (nota bene: the list is defined for tracks, not for vertices)
It requires the simulation file and the file with vertices reconstructed by FEDRA. It will produce the first version of the CSV list with the vertices. Is is launched as the following:
Here, the lists saved at step 1. will be used. I do a loop on the tracks associated to the vertices: if the MCTrackID matches the one of a charm's daughter, I save the information about track and vertex in the CSV file.
A recent modification: tracks connected to the mother charm hadron are now saved at this step, and not at the following one. This means that a track is considered connected (feature 3), only if it is associated to a vertex. Otherwise it will belong to feature 4 (extra tracks) Alcuni sono collegati alla traccia madre (cioè l'adrone charmato), e quindi l'MCTrackID sarà riconosciuto come adrone charmato, essendo il primo segmento.
It is launched as the following:
We tag as extra tracks charm daughters which have been reconstructed as tracks, but not associated to any vertex.
It is launched as the following:
The aforementioned procedure gives CSV files with the following columns:
ntracks;
ivtx;
itrk;
MCEventID;
MCTrackID;
CharmMotherPdg;
CharmMotherPx;
CharmMotherPy;
CharmMotherPz;
MCMotherID;
predmolt;
preddecaylength;
quantity;
vx;
vy;
vz;
topology
The list can be read with a script to count the different MC topologies, and build a more easiliy readable vertex log:
The number of CSV files to be passed as input can be as much as it is needed.
Primary
Secondary
Connected
Extra
Missing
Primary
10.9%
18.5%
1.9%
9.7%
8.1%
Secondary
-
13.0%
2.2%
8.8%
5.6%
Connected
-
-
0.13%
0.85%
0.76%
Extra
-
-
-
2.4%
4.0%
Missing
-
-
-
-
13.2%
Primary
Secondary
Connected
Extra
Missing
Primary
11.3%
18.5%
1.8%
9.9%
8.1%
Secondary
13.3%
2.1%
8.7%
5.6%
Connected
0.07%
0.8%
0.7%
Extra
2.3%
4%
Missing
12.5%
