Difference between revisions of "Reconstruction, July 7, 2009"
From GlueXWiki
(→Minutes) |
|||
(7 intermediate revisions by the same user not shown) | |||
Line 9: | Line 9: | ||
* [[Reconstruction References]] | * [[Reconstruction References]] | ||
+ | * <font color=red>Reconstruction document in progress</font> | ||
= Tentative Agenda = | = Tentative Agenda = | ||
− | # Reconstruction Physics Reactions | + | # Reconstruction: Physics Reactions |
# Generating Background Events (Pythia: 124 μb @ 9GeV beam energy) | # Generating Background Events (Pythia: 124 μb @ 9GeV beam energy) | ||
− | # | + | #* Shared data sets, filtering, et c. |
+ | # Angular distributions in GlueXAmpTools? (seems to be flat regardless of spin, no structure in cosGJ - Blake) | ||
+ | |||
+ | = Discussion = | ||
+ | |||
+ | == Reconstruction == | ||
+ | |||
+ | * It is required that there are 4 photons showers and none are due to charged particles, this gives 6 possible combinations. | ||
+ | * Pi0 and Eta candidates are chosen as the eta-pi0 pair with the largest probability (from the fit) | ||
+ | * Before cuts ~51% of the generated events can be reconstructed. A cut on χ^2 < 3 leaves approximately 39% of generated events, | ||
+ | |||
+ | '''Pi0 and Eta masses''' the line in red is cutting on chi^2 < 3. | ||
+ | |||
+ | [[Image:PI0MASS CHI2cut.png|300px]][[Image:ProbMax PI0 CHI2cut.png|300px]] | ||
+ | [[Image:ETAMASS CHI2cut.png|300px]][[Image:ProbMax ETA CHI2cut.png|300px]] | ||
+ | |||
+ | |||
+ | |||
+ | '''resonance mass''' | ||
+ | * a combination of a0(980), a2(1320), a0(1450), a4(2000) resonances. | ||
+ | |||
+ | [[Image:RESMASS CHI2cut.png|300px]] | ||
+ | |||
+ | |||
+ | |||
+ | '''missing mass''' | ||
+ | * the missing 4-momentum can be calculated from the eta, pi0, beam and target 4-momenta. | ||
+ | |||
+ | [[Image:RECMASS CHI2cut.png|300px]] | ||
+ | |||
+ | |||
+ | |||
+ | '''momentum transfer''' | ||
+ | |||
+ | [[Image:T CHI2cut.png|300px]] | ||
+ | |||
+ | == Background == | ||
+ | |||
+ | |||
+ | = Minutes = | ||
+ | # Matt suggested looking at the accept/reject code to determine whether this is causing the issues for flatness in cosGJ. | ||
+ | #Mihajlo askked how many pairs of eta pi0 pass a chi^2 < 3 cut. | ||
+ | #A discussion of how large data sets were in order to produce a common background data file. Blake looked at his bggen folders to find that a BGGEN hddm file is ~1kB/event (10^7 events is ~10GB) and an HDGEANT hddm file (from filtered BGGEN events) is ~27kB/event. |
Latest revision as of 13:16, 7 July 2009
Time: 11:00 EST/10:00 CST
- See GlueX Communications for detailed instructions
- ESNET Number 8542553
Contents
Documents to Review
- Reconstruction References
- Reconstruction document in progress
Tentative Agenda
- Reconstruction: Physics Reactions
- Generating Background Events (Pythia: 124 μb @ 9GeV beam energy)
- Shared data sets, filtering, et c.
- Angular distributions in GlueXAmpTools? (seems to be flat regardless of spin, no structure in cosGJ - Blake)
Discussion
Reconstruction
- It is required that there are 4 photons showers and none are due to charged particles, this gives 6 possible combinations.
- Pi0 and Eta candidates are chosen as the eta-pi0 pair with the largest probability (from the fit)
- Before cuts ~51% of the generated events can be reconstructed. A cut on χ^2 < 3 leaves approximately 39% of generated events,
Pi0 and Eta masses the line in red is cutting on chi^2 < 3.
resonance mass
- a combination of a0(980), a2(1320), a0(1450), a4(2000) resonances.
missing mass
- the missing 4-momentum can be calculated from the eta, pi0, beam and target 4-momenta.
momentum transfer
Background
Minutes
- Matt suggested looking at the accept/reject code to determine whether this is causing the issues for flatness in cosGJ.
- Mihajlo askked how many pairs of eta pi0 pass a chi^2 < 3 cut.
- A discussion of how large data sets were in order to produce a common background data file. Blake looked at his bggen folders to find that a BGGEN hddm file is ~1kB/event (10^7 events is ~10GB) and an HDGEANT hddm file (from filtered BGGEN events) is ~27kB/event.