CosPhi Status

Aus HERMESwiki
Version vom 13. Januar 2011, 16:11 Uhr von Francesca (Diskussion _ Beiträge)
(Unterschied) ← Nächstältere Version _ Aktuelle Version (Unterschied) _ Nächstjüngere Version → (Unterschied)
Zur Navigation springen Zur Suche springen

Page maintainer: Rebecca and Francesca

Workingpage.jpg This is a working page which will continuously change. The content is not reviewed and is intended for people working on this specific topic.

This page catalogues the on-going work on the cos(nphi) analysis. It is only meant collect studies, information and ideas for the purposes of aiding in discussion. Please consult Francesca or Rebecca before using any of the content on this page.
For the first "released" information (hadrons) about this analysis see the Release page, for realesed plots and numbers look at the 08-048 to 08-051 entries here (obsolete)
For the last "released" information (pions) about this analysis see the Release page, for realesed plots and numbers look at the 10-021 to 10-011c entries here
For the last "released" information (kaons vs. pions vs. hadrons) about this analysis see the Release page, for realesed plots and numbers look at the 10-065 to 10-076 entries here
For the last "released" information (pions vs. hadrons) about this analysis see the Release page, for realesed plots and numbers look at the 10-077 and 10-078 entries here
For information on the parameterization see here /user/giordano/public/Param and read the README file, and also check with Francesa
For the latest discussions see the Draft 78 email archive


January 2011


December 2010

Analysis:

  • Pmatrices: check 2007 H using weighted Pmatrices
    • new run list running > output in /user/giordano/Pmatrices/Outdir/evt_cosphiCuts_ratio_Pepsi5
  • Q2: it seems we are not sensitive to Q2 (too small range), still to do: calculate proper error for the parametrization
  • Projection with multiplicities (eliminate sys. error), when ready
  • Pepsi with mig matrix reweighted with gmcTrans > running gmctrans for pions

November 2010

Analysis to do (postponed after 1st circulation):

  • Pmatrices: check 2007 H using weighted Pmatrices
  • Q2: it seems we are not sensitive to Q2 (too small range), still to do: calculate proper error for the parametrization


Text (highest priority):

  • Figures: PAW or ROOT? code needed
  • Projection with multiplicities (eliminate sys. error), when ready
  • charge difference for all hadron type?
  • Suggestions:
    • change order in Results section: Hadrons (move from appendix), Pions, Kaons, Comparison section
    • prefactors will not be removed
    • add evt appendix?
  • Discussions needed:
    • what definition for BM effect??

October 2010

Analysis, To do:

  • Pions: redo Pepsi & add systematic for zero vector : done
  • Kaons:
    • test on RICH contamination: eliminated form sample hadrons with momenta between 8 and 11 GeV, to eliminate large contamination from proton
    • test on Pmatrices: check differences when changing the cross section distribution (cross section from data): the plot show the comparison between standard Pmatrices and new ones that include cuts from cosphi analysis and multiplicities from data.
  • PEPSI with data stats (test, not sys) -- will test how bad bad-bins are in the data?


  • Q2:
    • will need new pepsi and zero sys done
    • 4-D fit: show different contributions separately: done => is there a bug? try in 1D: done


  • Year dependence:
    • time dependence studies finalized:
      • 2006 e rejected
      • 2005: some discrepancies between data and MC is seen, large sinphi observed in MC, but not in data, checked the MC production, the magnet is off, and the beam position is right, anything else to check?
    • smoothing accounts only partially for statistic fluctuations: better way to evaluate this sys?
      • use average of differences instead of max? > in this way we account for different statistical power of data: yes
      • n times randomly changed yields and pepsi


  • Finalized misalignment studies:
    • Alexander & Larry produced a test 2005 (05t0) data production including shifted dipole > no difference with standard 2005 (05d2) production
  • Use multiplicities or flag projected plots with "Integrated with Pythia"
    • first test using a parametrization of multiplicities show no significative difference wrt use Pythia cross section
    • Sylvester will provide the multiplicities in our binning

August 2010

To be discussed:

  • Release hadrons/kaons at next release meeting? Missing:
    • Hadrons : redo year-dep sys, (add zero sys ? or not, to be compatible with released pions?) check diff hadron/pion with RICH sys
    • Kaons : xc for projected, finalize sys
    • write the report: deadline monday 23rd

Paper

  • check Norik email about MC cross section tests
  • draft!!!!

Analysis

  • Redo all year dep sys wo 2006e
  • Pions: redo Pepsi
    • PEPSI with data stats (test, not sys) -- will test how bad bad-bins are in the data?
  • Hadrons: add the zero sys.
    • If the RICH systematic on pions is larger than the difference with hadrons then we are "safe" to publish without mention of kaons (but it is probably too small??)
  • Kaons: xc done for 900, do for projected and add other systematics
    • For the kaon PEPSI is the normalization done correctly for p=0?
  • Q2:
    • will need new pepsi and zero sys
    • 4-D fit: show different contributions separately: done => is there a bug? try in 1D
  • Year dependence:
    • time dependence studies finalized: 2006 e rejected, 2005 under study
    • smoothing accounts only partially for statistic fluctuations: better way to evaluate this sys?
      • n times randomly changed yields and pepsi
      • not linear (4-D) smoothing? done => worst
  • Finalize misalignment studies:
    • Alexander & Larry are producing a test 2005 data production including shifted dipole
  • Use multiplicities (?)
    • Gunar's suggestion, flag projected plots with "Integrated with Pythia"

July 2010

New Hydrogen XC  (?)

Moments: Diff 1% cosphi ecosphi cos2phi ecos2phi %%% Cov_cosphi Cov_cos2phi off diag

  • Pi + 127 13 202 12 %%% 28 27 310991
  • Pi - 143 4 180 3 %%% 15 14 254554
  • h + 116 9 208 8 %%% 29 30 349546
  • h - 139 6 173 7 %%% 14 16 268253


Kaons XC

Moments: Diff 1% cosphi ecosphi cos2phi ecos2phi %%% Cov_cosphi Cov_cos2phi offdiag

  • Hyd K+ 67 8 97 7 %%% 16 14 173578
  • Hyd K- 64 3 74 2 %%% 13 11 68470
  • Deu K+ 54 10 74 12 %%% 18 18 115566
  • Deu K- 34 3 31 3 %%% 5 5 22659

Moments: Diff 1% cosphi ecosphi cos2phi ecos2phi

  • 2000 H K+ 45 2 59 3
  • 2000 H K- 27 4 33 3
  • 2006e+ H K+ 64 6 67 4
  • 2006e+ H K- 51 5 41 6
  • 2007 H K+ 74 10 91 7
  • 2007 H K- 61 1 55 3
  • 2000 D K+ 77 17 87 17
  • 2000 D K- 22 4 36 5
  • 2005 D K+ 20 1 30 1
  • 2005 D K- 8 0 14 0
  • 2006e+ D K+ 18 2 21 2
  • 2006e+ D K- 18 1 15 1
  • 2007 D K+ 27 5 30 2
  • 2007 D K- 15 1 13 2


Yields:

  • MC (OK)
    • 2006 e+ H K+: Fra: 5506440. Reb: 5506425 Diff/Fra 0.00024 Diff 1306. all 1 except 36 of 2, 2 of 3, 1 more than 4
    • 2006 e+ H K-: Fra: 2835345. Reb: 2834542. Diff/Fra 0.00039 Diff 1094. all 1 except 31 of 2, 1 of 3
    • 2000 H K+: Fra 1801264. Reb: 1801269. Diff/Fra 0.00031 Diff 562. all 1 except 8 of 2
    • 2000 H K-: Fra: 939103. Reb: 939097. Diff/Fra 0.00036 Diff 336. all 1 except 4 of 2
    • 2000 D K+: Fra: 1750177. Reb: 1750186. Diff/Fra 0.00031 Diff 549. all 1 except 6 of 2, 2 of 3
    • 2000 D K-: Fra: 999777. Reb: 999769. Diff/Fra 0.00035 Diff 345. all 1 except 5 of 2
    • 2006 e+ D K+: Fra: 1748052. Reb: 1748037. Diff/Fra 0.00028 Diff 489. all 1 except 3 of 2
    • 2006 e+ D K-: Fra: 986117. Reb: 986104. Diff/Fra 0.00031 Diff 301. all 1 except 5 of 2
    • 2005 D K+: Fra: 1649511. Reb: 1649506. Diff/Fra 0.00031 Diff 519. all 1 except 8 of 2, 1 of 3
    • 2005 D K-: Fra: 937123. Reb: 937115. Diff/Fra 0.00026 Diff 247. all 1 except 4 of 2, 1 of 3
  • Born (OK)
    • Hyd K+ Fra 6720974.0 Reb 6721023.0 Diff/Fra 0.39E-05 Diff 26: all of 1 except 3 of 2
    • Hyd K- Fra 3510762.0 Reb 3510796.0 Diff/Fra 0.15E-04 Diff 51: all of 1
    • Deu K+ Fra 6309983.0 Reb 6310035.0 Diff/Fra 0.57E-05 Diff 36: all of 1 except 3 of 2
    • Deu K- Fra 3602115.0 Reb 3602146.0 Diff/Fra 0.12E-04 Diff 44: all of 1 except 1 of 2
  • Data
    • 2007 H K+ Fra 260101.9 Reb 260104.9 Diff/Fra 0.92E-04 Diff 24: all of 1 except 1 of 2
    • 2007 H K- Fra 87984.4 Reb 87981.3 Diff/Fra 0.17E-03 Diff 15: all of 1
    • 2000 H K+ Fra 97725.8 Reb 97725.1 Diff/Fra 0.16E-03 Diff 16 : all of 1
    • 2000 H k- Fra 33038.1 Reb 33039.2 Diff/Fra 0.21E-03 Diff 7 of 1
    • 2000 D K+ Fra 122754.6 Reb 122753.6 Diff/Fra 0.20E-03 Diff 25 of 1
    • 2000 D k- Fra 47834.3 Reb 47834.3 Diff/Fra 0.13E-03 Diff 6 of 1
    • 2005 D K+ Fra 62089.3 Reb 62091.5 Diff/Fra 0.19E-03 Diff 12 of 1
    • 2005 D k- Fra 24125.4 Reb 24125.4 Diff/Fra 0.25E-03 Diff 6 of 1
    • 2006 e+ H K+ Fra 241944.4 Reb 241945.1 Diff/Fra 0.12E-03 Diff 30 of 1
    • 2006 e+ H K- Fra 81284.9 Reb 81285.0 Diff/Fra 0.74E-04 Diff 6 of 1
    • 2007 D K+ Fra 26213.9 Reb 26212.8 Diff/Fra 0.19E-03 Diff 5 of 1
    • 2007 D K- Fra 66378.3 Reb 66375.5 Diff/Fra 0.17E-03 Diff 11 of 1
    • 2006 e+ D K+ Fra 57948.6 Reb 57947.6 Diff/Fra 0.86E-04 Diff 5 of 1
    • 2006 e+ D K- Fra 22516.7 Reb 22516.7 Diff/Fra 0.18E-03Diff 4 of 1




To do

Rebecca's To Do list

  • check first moments of cosphi, cos2phi, cos3phi, sinphi for every 10k events for 06e,06p,07 data
  • Year dependence t-tests don't have any obvious conclusion
  • (check that bins with large chi2 also get dropped in projection)
  • XC kaons


Paper

  • check Norik email about MC cross section tests


Summary of collaboration meeting:

  • Pions: redo Pepsi
    • PEPSI with data stats (test, not sys) -- will test how bad bad-bins are in the data?
  • Hadrons: add the zero sys.
    • If the RICH systematic on pions is larger than the difference with hadrons then we are "safe" to publish without mention of kaons (but it is probably too small??)
  • Kaons: do the xc and add the other sys
    • For the kaon PEPSI is the normalization done correctly for p=0?
  • Q2:
    • will need new pepsi and zero sys
    • what to do? leave proj vs. y or move to proj. vs Q2?
    • 4-D fit: show different contributions separately
    • include the "r" term?
  • Year dependence:
    • finalize time dependence studies
    • smoothing accounts only partially for statistic fluctuations: better way to evaluate this sys?
      • n times randomly changed yields and pepsi, or n times randomly change 900 moments and project?
      • not linear (4-D) smoothing?
  • Finalize misalignment studies:
    • Alexander: check a few runs using a complete description of the magnet
    • Francesca: pepsi with a MC including dipole shift
  • Use multiplicities (?)
    • Gunar's suggestion, flag projected plots with "Integrated with Pythia"

June 2010

  • Time dependence:
    • Rebecca - stability of yields vs (x,y,z,pt) in each detector quadrant for each year or subsection of each year, 06e- with and without recoil magnet;
    • Rebecca - p0 parameter: compare different years to look for something strange in one year (2006 e-?);
    • Rebecca - plots chi^2 distribution of each year
    • DONE - Be careful about BB in difference years -- already done, if one or both years have a BB, sys=0, error_sys=100 (used for smoothing);
    • Postpone/Disregard - check statistic fluctuations using reweighted Pythia: do "10" times 4 different (independent) unfolding (1 for each year statistic);
    • add non-linear terms to the smoothing. Is this really meaningful? How smooth do we expect it to be?
    • do we really want to reduce our systematic when our default is always to rely on it to cover all our other issues?
  • Comparison to old hadrons
    • the difference is due to the z binning, but instability present only in 2006 e-, other years are stable --> this is now a test of time dependence.
    • redo the projection of only 2006 e moments using the same kine region for the 2 different binnings
  • Kaons (there will undoubtedly be a lot more work, do we really want to go there with limited manpwoer? better to publish something than nothing?)
    • XC
    • finish systematics
  • Pions: redo pepsi for pions including the Kaons parametrization
  • Postpone until multiplicities are done Projection: use multiplicities to XC pythia in 3D? Comparison of pt distribution for DISng/Pythia/GMCtrans
  • Q2 dependence (drop from the paper?)
    • use parameterization to check for dependence. Then what? how to include this in the paper? What if there is still a dependence?
    • parametrization (x-z-pt)+(x-z-pt)*1/Q^2 + prefactors
    • somehow fit cos(phi) and cos(2phi) to see if slopes are consistent. Does this even make sense? So what if they aren't consistent, there are other terms...
  • Beam slopes & Fringe field effects
    • See Alexander's plots and Francesca's Slides:
    • Discussion with Alexander: the beam slope+ fringe fields may generate an azimuthal modulation (sinphi??)
    • Alexander will have a look at the 900 moments to look for momentum dependence > No evident momentum dependence: no fringe field effects?
    • check if Pythia+beam slope generate a sinphi, or another modulation > no
    • check if "sinphi" is sensible to the beam slope (change hadron theta_x) > no (neither cos3phi)
    • To do:
      • Alexander: check results of the beam finder for a few runs using a more complete fringe field description
      • Francesca: generate a Pythia production using a 2mm shift of dipole and do a Pepsi

May 2010

To do for the next Cosphi Analysis meeting (27th May):

    • Hadron / Pion comparison year-by year
      • The difference (pi - h) is stable between years H+ H- D+ D-

April 2010

Minute of Release meeting

  • Q^2 dependence need to be understand:
    • check pi+ - pi-
    • try to use the complete prefactors & include 1/Q factors
  • pi/hadron comparison:
    • comparison in 900?
    • Kaons: Rebecca has already kaon moments (to be xchecked?)
    • Francesca: kaon fit and pepsi challenge
    • K + pi = hadron momenta?
    • new hads with old binning
    • check pi/hadron ratio year by year
      • look at year dependence of kaons
  • more tests on the beam slopes > meeting with Alexander friday 30 April
  • discussion with Henk: W2/nu dependence?

Missing for the analysis

  • pi--pi+ : possible to calculate systematic in 900?
  • zero bin test with reweighted pythia
  • Projection: comparison bt Pythia cross section and multiplicity, in multidim.

Missing for the paper

  • ADD mention of additional moments!!
  • draft

March 2010

XC special


XC hadrons


  • Moments 900 (diff>1%)
    • 2000 H h+
      • cosphi 125 ecosphi 18
      • cos2phi 173 ecos2phi 16
    • 2000 H h-
      • cosphi 94 ecosphi 12
      • cos2phi 121 ecos2phi 13
    • 2005 D h+
      • cosphi 118 ecosphi 7
      • cos2phi 136 ecos2phi 6
    • 2005 D h-
      • cosphi 84 ecosphi 6
      • cos2phi 92 ecos2phi 5
    • 2000 D h+
      • cosphi 119 ecosphi 12
      • cos2phi 160 ecos2phi 13
    • 2000 D h-
      • cosphi 130 ecosphi 11
      • cos2phi 145 ecos2phi 11


  • Data (OK)
    • 2000 H h+ : Fra 798863.0 Reb 798865.0, differences: of 1: 22. of 2: 2
    • 2000 H h- : Fra 447516.0 Reb 447516.0, differences: of 1: 23.
    • 2000 D h+ : Fra 1019355.0 Reb 1019344.0, differences: of 1: 17
    • 2000 D h- : Fra 658269.0 Reb 658272.0, differences: of 1: 32.
    • 2005 D h+ : Fra 522380.0 Reb 522377.0, differences: of 1: 33.
    • 2005 D h- : Fra 340543.0 Reb 340546.0, differences: of 1: 21
    • 2006 e+ H h+ : Fra 1968548.0 Reb 1968530.0, differences: of 1: 35. of 2: 1
    • 2006 e+ H h- : Fra 1115756.0 Reb 1115749.0, differences: of 1: 28.
    • 2006 e+ D h+ : Fra 480758.0 Reb 480757.0, differences: of 1: 15.
    • 2006 e+ D h-: Fra 312566.0 Reb 312562.0, differences: of 1: 33.
    • 2006 e- H h+ : Fra 710061.0 Reb 710054.0, differences: of 1: 24.
    • 2006 e- H h- : Fra 396977.0 Reb 396975.0, differences: of 1: 28.
    • 2007 H h+ : Fra 2114817.0 Reb 2114805.0, differences: of 1: 27.
    • 2007 H h- : Fra 1202331.0 Reb 1202320.0, differences: of 1: 36.
    • 2007 D h+ : Fra 549855.0 Reb 549845.0, differences: of 1: 30
    • 2007 D h-: Fra 360528.0 Reb 360526.0, differences: of 1: 17.
  • Born (OK)
    • H h+ differences: of 1: 10. of 2: 2
    • H h- differences: of 1: 6. of 2: 2
    • D h+ differences: of 1: 13.
    • D h- differences: of 1: 8.
  • Matrices (OK)
    • 2000 H h+ Fra: 13782517. Reb: 13782534., Diff 2734. all 1 except 111 of 2, 13 of 3, 4 of 4, 2 more than 4, Diff/Fra=0.2*10^-3
    • 2000 H h- Fra: 8925905. Reb: 8925918., Diff 2003. all 1 except 84 of 2, 12 of 3, 4 of 4, 1 more than 4, Diff/Fra=0.2*10^-3
    • 2000 D h+ Fra: 9654995. Reb: 9655011, Diff 2133. all 1 except 81 of 2, 10 of 3, 2 of 4, Diff/Fra=0.2*10^-3
    • 2000 D h- Fra: 13265744. Reb: 13265750., Diff 2643. all 1 except 105 of 2, 22 of 3, 3 of 4, 1 more than 4, Diff/Fra=0.2*10^-3
    • 2005 D h+ Fra: 12264034. Reb: 12264021., Diff 1768. all 1 except 51 of 2, 13 of 3, 3 of 4, 1 more than 4, Diff/Fra=0.1*10^-3
    • 2005 D h- Fra: 8916121. Reb: 8916109., Diff 1561. all 1 except 66 of 2, 10 of 3, 1 of 4, Diff/Fra=0.2*10^-3
    • 2006 e+ H h+ Fra: 41236352. Reb: 41236448, Diff 4280. all 1 except 233 of 2, 40 of 3, 17 of 4, 21 more than 4, Diff/Fra=0.1*10^-3 (checked, ok)
    • 2006 e+ H h- Fra: 26417776. Reb: 26417848., Diff 3201. all 1 except 159 of 2, 42 of 3, 15 of 4, 9 more than 4, Diff/Fra=0.1*10^-3
    • 2006 e+ D h+ Fra: 12969011. Reb: 12968983., Diff 2236. all 1 except 82 of 2, 10 of 3, 8 of 4, 2 more than 4, Diff/Fra=0.2*10^-3
    • 2006 e+ D h- Fra: 9369632. Reb: 9369626., Diff 1785. all 1 except 67 of 2, 8 of 3, 4 of 4, Diff/Fra=0.2*10^-3
    • 2006 e- H h+ Fra: 14937841. Reb: 14937835, Diff 2336. all 1 except 89 of 2, 16 of 3, 5 of 4, Diff/Fra=0.2*10^-3
    • 2006 e- H h- Fra: 9560166. Reb: 9560182, Diff 1813. all 1 except 43 of 2, 6 of 3, 3 of 4, 1 more than 4, Diff/Fra=0.2*10^-3

February 2010

Missing for the paper

  • Hadrons -- XC + systematic
  • Systematic for special plots
  • pi--pi+ : possible to calculate systematic in 900?
  • talk to theorists?
    • email Gamberg!!!
  • zero bin test with reweighted pythia
  • project with multiplicities???
  • look at correlations in the projected plots (Rebecca has the #s)
  • ... draft ...

Rebecca's list

  • figures
    • change colors to black and white?
    • how to format special plot figures?

Francesca's list

  • hadrons+ sys (XC)
  • sys for special plots
  • zero vector sys

Topics to discuss in drafting

  • Projection: comparison bt Pythia cross section and multiplicity, in multidim.
  • More in HERMES section?
  • A better PRD temple?
  • What systematic checks to include?
  • smoothing?
  • HTC?
  • background model test?
  • remove smearing matrix and gmctrans sys?
  • ADD mention of additional moments!!

Still missing:

Done:

Just in case

Special Plots

  • Cherry, prefactor removed (avg Q2 in y plot) D, H, pi+, pi-

January 2010

Fixed

  • cos3phi:
    • the beam slope does not affect cos3phi moments
    • The acceptance's cos3phi is of same order of acceptance's cosphi >> use residual cos3phi as estimate of residual acceptance effects
  • Acceptance for 2000 and 2007 (target changed) are similar
  • The year depepndence is better with HRC, anyway the final moments show a shift wrt HTC, it was decided to use HTC (larger systematic)

To do

  • check acceptance for pi-
  • check vertex/track probability distributions for 2005 & 2006e-: do we need a different cut here?

XC

==== Summary ====

* BORN: everything is OK

* DATA: everything is OK, except 2006 e- : wrong TRDcorrectiong (R) and target bug (F).... FINE!!

* MC:   
  ** new Matrices:
   *** 2000 ok
   *** 2006 e- ok
   *** 2006 e+ Fra have 100 events more, is it fine?
 ** old:
  *** 2000: found a difference in selecting events when using HRC, changing in progress.... FINE!!
  *** 2005  missing runs, running... FINE!
  *** all matrices show difference of 0.2E^-3 (#diff /all events), which is roughly doubled wtr to the past (jan'09) 
     it was ~0.09E^-3 and now we have roughly 2 times the number of bins (which difference we had in the last release?)

* Weights and DIS:
  ** DISng:   weights ok, DIS to be xc again....
  ** Data:    DIS to be xc again
  ** Born:    OK
  ** MC:      Deuterium OK
              Hydrogen  To be checked !!!!

Details

  • Final Moments: bins not within 1% agreement plotted here


Unfolding: 2006 e- both with Rebecca's input

  • Normalized data: 2 diff > 1%
    • 9364 3.1035e-10 3.1451e-10 -4.1636e-12 0.013416 0.98676
    • 9451 4.7356e-10 4.8172e-10 -8.1577e-12 0.017226 0.98307
    • 55 diff > 0.1%
  • Normalized inverted data covariance (bck included): 0 diff > 1%
    • 5321 diff > 0.1%
      • all less than 0.13%,
  • Normalized smearing matrix: 0 diff < 1%
    • 6418 diff > 0.1%
      • 5878 of these are < 0.2%


  • term0 = SX: 23274 diff < 1% (out of 396939 non-zero numbers)
    • should we worry about this or only the final results?


  • 900 Moments:
    • 22 diff >1% in cosphi
    • 34 diff>1% in cos2phi
    • Counting only moments with diff>1% && ( moment>error __ diff/err > 0.5% ) we have the following diff
      • 528 cos -0.055830 +- 0.378090 -0.053697 +- 0.378428 -0.002133 0.038202 0.005641
      • 827 cos -0.097280 +- 0.332830 -0.100257 +- 0.332633 0.002977 0.030602 0.008945
      • 126 cos2 -0.040860 +- 0.196190 -0.041987 +- 0.196370 0.001127 0.027584 0.005745
      • 311 cos2 -0.032500 +- 0.096780 -0.031932 +- 0.096753 -0.000568 0.017492 0.005874
      • 464 cos2 0.005660 +- 0.067480 0.005286 +- 0.067479 0.000374 0.066058 0.005541
      • 827 cos2 -0.098850 +- 0.301880 -0.096546 +- 0.301657 -0.002304 0.023306 0.007632
    • Counting only moments with diff>1% && ( moment>error __ diff/err > 1% ) we have no diff
  • Final Covariance
    • 33700 diff of >1% out of 879844 non-zero numbers
  • Projection
    • Differences due to Rebecca not including highest z bin, and Francesca cutting on moments>1, fixed -- OK


Weights and DIS:

  • DISng DIS / ievgen / extraweight
    • 00D (updated)
      • Reb 2594603.0 / 69189498 / 1
      • Fra 2595048 / 69189498 / 1 -- OK
    • 00H (updated)
      • Reb 2942456.25 / 69208197 / 1
      • Fra 2942914. / 69208197 / 1 -- OK
    • 05D (updated)
      • Reb 2449247.75 / 64306777 / 1
      • Fra 2449385. / 64306777 / 1 -- OK
    • 06D (updated)
      • Reb 2563993.75 / 66313166 / 1
      • Fra 2564163 / 66313166 / 1 -- OK
    • 06He- (updated)
      • Reb 2939086.25 / 66993839 / 1
      • Fra 2939280.25 / 66993839 / 1 -- OK
    • 06He+ (updated)
      • Reb 2907514.25 / 66324755 / 1
      • Fra 2907708.25 / 66324755. / 1 -- OK


  • Data DIS
    • 00D (updated) -- OK
      • Reb 8111893
      • Fra 8109982--> (BUG FIX 2/16/10) 8111505
    • 05D (updated) -- OK
      • Reb 5235674
      • Fra 5235672
    • 06D (updated) -- OK
      • Reb 4640460
      • Fra 4640459
    • 07D
      • Reb 5394745 -- OK
      • Fra 5394745
    • 00H (updated) -- OK
      • Reb 6393362
      • Fra 6392956
    • 06He+
      • Reb 18892600 -- OK
      • Fra 18892758
    • 06He- (updated) -- OK
      • Reb 6693541
      • Fra 6693572
    • 07H
      • Reb 20515740 -- OK
      • Fra 20516298


  • Born ievgen / extraweight
    • H
      • Reb 345000000 / 0.1654099971
      • Fra 345000000 / 0.165410 -- OK
    • D
      • Reb 345000000 / 0.1553229988
      • Fra 345000000 / 0.155323 -- OK


  • Rec ievgen / extraweight
    • 00D
      • Reb 584604955 / 0.1724569947
      • Fra 584604955 / 0.17245698 -- OK
    • 05D
      • Reb 585603244 / 0.1724600047
      • Fra 585603244 / 0.172460094 -- OK
    • 06D
      • Reb 593842567 / 0.1724559963 --> New: 593844046 / 0.172456
      • Fra 593842567 / 0.172455817 -- OK (New: no 7100327 + 7100602) 593844046 / 0.172456
    • 00H
      • Reb 569077719 / 0.1858499944
      • Fra 569582271 / 0.185850292, 569077719 / 0.185850 without 416 -- OK
    • 06He-
      • Reb 637995676 / 0.185854002 -- OK
      • Fra 637995676 / 0.185854554
    • 06He+
      • Reb 1767500814 / 0.185854, 1766993961 / 0.1858540028, 7100001 and 7103427
      • Fra 1767500814 / 0.18585379, 1766995192 / 0.185854 without 7100001 -- OK
    • New 06He+ (No 7100850 & 7100327)
      • Reb 1766488232 / 0.185854
      • Fra 1766488232 / 0.185854 -- OK


  • Yields:

Diff >1% unless otherwise notes

  • BORN D
    • pi+ 15 diff, all of 1 except 2 of 3 -- OK
    • pi- 11 diff, all of 1 except 1 of 2 -- OK
  • BORN H
    • pi+ 14 diff, all of 1 except 2 of 2 -- OK
    • pi- 9 diff, all of 1 except 2 of 2 -- OK


  • MC 00D
    • pi+ 1869. all 1 except 62 of 2, 10 of 3, 2 of 4 Diff/Fra = 0.2 10^-3
    • pi- 1622. all 1 except 39 of 2, 10 of 3, 1 of 4 Diff/Fra = 0.2 10^-3
    • old
      • pi+ 3142 diff: 2973 diff of 1 , 129 of 2, 27 of 3, 8 of 4 and 5 more than 4 ----->>>> rerunning
      • pi- 2619 diff:


  • MC 05D
    • pi+ 1474. all 1 except 38 of 2, 7 of 3, 2 of 4, 1 more than 4 Diff/Fra = 0.2 10^-3
    • pi- 1411. all 1 except 57 of 2, 7 of 3 Diff/Fra = 0.2 10^-3
    • old:
      • pi+ 1994. all 1 except 60 of 2, 15 of 3, 5 of 4, 2 more than 4
      • pi- 1937. all 1 except 80 of 2, 12 of 3
  • MC 06D & 07D
  • pi+ 1904. all 1 except 60 of 2, 14 of 3, 4 of 4, 1 more than 4 Diff/Fra = 0.2 10^-3
  • pi- 1609. all 1 except 48 of 2, 7 of 3, 2 of 4 Diff/Fra = 0.2 10^-3
    • old:
      • pi+ 2581. diff. all of ~1 except 85 of 2, 17 of 3, 6 of 4, 1 more than 4 --?
      • pi- 2162. diff. all of ~1 except 72 of 2, 11 of 3, 2 of 4 --?
  • MC 00H
    • pi+ 1945. all 1 except 70 of 2, 10 of 3, 1 of 4 Diff/Fra = 0.2 10^-3
    • pi- 1528. all 1 except 61 of 2, 6 of 3, 3 of 4, 1 more than 4 Diff/Fra = 0.2 10^-3
  • MC 06e-H
    • pi+ 1987. all 1 except 65 of 2, 15 of 3, 3 of 4 Diff/Fra = 0.2 10^-3
    • pi- 1606. all 1 except 33 of 2, 4 of 3, 2 of 4, 2 more than 4 Diff/Fra = 0.2 10^-3
    • old:
      • pi+ 2676. diff, all of ~1 except 105 of 2, 19 of 3, 5 of 4 --?
      • pi- 2158. diff, all of ~1 except 58 of 2, 5 of 3, 2 of 4, 2 more than 4 --?
  • MC 06e+H & 07e+H:
    • pi+ 3789. all 1 except 194 of 2, 34 of 3, 12 of 4, 16 more than 4 Diff/Fra = 0.1 10^-3
    • pi- 2933. all 1 except 151 of 2, 31 of 3, 12 of 4, 7 more than 4 Diff/Fra = 0.1 10^-3
    • old:
      • pi+ 6963. diff, all of ~1 except 337 of 2, 67 of 3, 25 of 4, 18 more than 4 --?
      • pi- 5458. diff, all of ~1 except 286 of 2, 50 of 3, 16 of 4, 9 more than 4 --?



DATA

  • 00D
    • pi+ 58 diff, all of 1 -- OK ---> BUG FIX --> 25 diff, all of 1, OK
    • pi- 58 diff, all of 1 -- OK ---> 2/16/10 -> 37 diff, all of 1, OK
  • 05D HTC
    • pi+ 34 diff, all of 1 or less -- OK
    • pi- 18 diff, all of ~1 -- OK
  • 06D HTC
    • pi+ 13 diff, all of ~1 -- OK
    • pi- 31 diff, all of ~1 -- OK
  • 07D HTC
    • pi+ 30 diff, all of ~1 -- OK
    • pi- 19 diff, all of ~1 -- OK


  • 00H
    • pi+ 33 diff, all of 1 except 1 of 2 -- OK
    • pi- 18 diff, all of 1 -- OK
  • 06e-H HTC
    • pi+ 23 diff. all of 1 -- OK
    • pi- 28 diff. all of 1 -- OK
  • 06e+H HTC
    • pi+ 36 diff, all of ~1 --OK
    • pi- 31 diff, all of ~1 --OK
  • 07H HTC
    • pi+ 31 diff, all of ~1 --OK
    • pi- 39 diff, all of ~1 --OK


07H HRC

  • DATA
    • pi+ 77 diff, all of ~1 except 6 of 2, 2 of 3 -- OK
    • pi- 48 diff, all of ~1 except 2 of 2, 1 of 3 --OK

December 2009

To Do

XC

  • All D results to see if the year dependence is real
    • 00D moments have a problem
  • Francesca add smearing matrix statistical errors


PEPSI

  • Need final MC
  • Need final moments to redo para
    • also do param with various # of parameters
  • Need to include with/without final BB list (check of effect of BB in silver region)
  • Include with/without RICH to see effect


Model dependence

  • Need new pythia statistics that are running
  • Need gmc_trans


Misalignment / Top/Bottom / Additional moments / Year Dependence

  • Should discuss beam slopes (effect top/bottom)
  • Include difference between with and without additional moments as a systematic


Final sysmatics

  • How to calculate --> write some code
  • What to include:
    • PEPSI (RICH, BB, bin size)
    • Additional Moments
    • Model dependence (all 3? only the worst?)
    • Something more for year dependence / top/bottom / misalignment / beam slopes ??

November 2009

  • Pepsi Challenge
    • Pepsi challenge with Reweighted Pythia (comparison 1 vs. 2 GeV) using
    • Systematic (half difference) from Pepsi challenge with Reweighted Pythia (comparison 1 vs. 2 GeV) using


Tweek summary

  • Tweek slides
    • long term plans (Monday) [1]
    • Tuesday studies [2]
    • Thursday follow up studies [3]
    • Friday summary [4]
  • Highest priority
  • priority

To do: Analysis

  • HTC:
    • Change vertex finder from 1-to-beam to 2-to-beam, add the correct variables for Fiducial Volume
    • with the new vertex finder check again
      • the differences HRC/HTC
      • the cuts on probabilities
      • the Top-Bottom differences
      • the additional moments
      • 2 -> 1 GeV ???
  • Misalignment:
    • improved after new vertex finder?
    • new fiducial volume accounting for real position of detectors
    • tomography of tracks
    • harder Fiducial Volume cuts
    • is it possible to have new geometry file accounting for mis. in 2006/2007 data?
    • use Top/Bottom studies? Year dependence? Additional moments?
  • 06/07 Year dependence
    • look at phi distributions
    • randomly mix events (not time dependent) and see what a Ttest gives
  • Moving from 2 to 1 GeV
    • Plot phi distributions to see what is happening
    • Try to understand what is happening with a PEPSI challenge
    • check RICH systematics (Francesca: send Rebecca Reweighted Pythia + RICH)
  • Include smearing matrix statistics in fitting procedure
    • test using various different b vectors
    • how much effect does it have on the moments?
  • final PEPSI-challenge
    • Increase pythia statistics
    • Vary # of parameters used in input model
    • include RICH
    • include & not BBlist from data to get a systematic on the missing BB in the cherry
  • Model dependence tests GUNAR!!!
  • Final systematics evaluated for projections and fully differential bins

To do: Draft

  • Report need to be ready at the end of november
    • discuss again on TUESDAY
  • paper
    • Add longer descriptions, unfolding appendix
    • discuss on FRIDAY

T-week summary

  • The last pT bin (1.3-1.7) has too low average Q^2: it was decided to drop it
  • Moving from 2 to 1 GeV: changes in some bins > does the systematic from RICH change?
    • Pmatrices don't show anything pi+ pi-
    • Differences in data kinematics are negligible
  • Additional moments: small effect, but similar in all years: misalignment? > check again after new HTC vertex finder
  • 2006e+ vs. 2007: systematic difference according to t-test, agreement seems better after including cos3phi
  • Top-Bottom: differences for both leptons&hadrons in top and bottom detectors:
    • Misalignment?
    • do a tomography of tracks, to see where they fall
    • which are the real positions of detectors? Mail Alexander
    • Harder fiducial volume
    • HTC correct the misalignment during tracking, but if mis. cause acceptance effects these are not corrected for. Is it possible to have the new geometry file?
  • HTC vs. HRC: HTC and HRC results are close when no cuts are applied to HTC
    • Probability cuts: changing the cuts slightly change the results. With cuts 0.01 the MC seems already to eliminate all bad tracks/verteces, while cuts for data should be hardened to 0.02-0.03, but no differences are observed in the extracted results. Redo these tests with new vertex finder
  • Binning tests: results depend slightly on the choise of binning. Systematic for this is accounted in the final PEPSI-challenge

October 2009

XC status for HTC

September 2009

To do (Francesca)

  • Code: event-list > vector, eventlist > matrix
  • Mixed events
  • 2007 HRC vs. HTC
  • Test HTC vertex probability (Data vs. MC)
  • After XC:
    • Bin test: change phi/kine binning
    • Additional moments
    • Other sys checks (Top-Bottom, Up/Down ...)

August 2009

Rebecca's Work

Average / Simultaneous Fit / Combining data years

  • Previous studies led to three different ways to combine data from different years.
  • The latest write-up of the math shows that in fact all three methods are different.
    • However, a fourth method, where the *parameters* from each data year (rather than the moments) are averaged corresponds exactly to the method of simultaneously fitting the data year, both in math and as demonstrated in these plots, showing 4 pT bins vs pT and vs. z
  • (New) Plots can be found here
  • (Old) Slide are here

Given that the method are in fact different, I believe that (from investigating the math) simultaneously fitting the years is the best option

Normalization

DIS cross section

The data is normalized to "cross section" level by doing Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle n(i)*{\frac {DIS_{MC}}{DIS_{data}}}*{\frac {extraweight}{IEVGEN}}}

For this to work properly we need a MC with an accurate DIS cross section. We first used pythia for this, but as pythia doesn't include elastic Bethe-Heitler we switched to disNG. However, to be sure that we are not too sensitive to this number I ran the unfolding for 2000 Hydrogen with

  • DISxsec from new disNG productions
  • DISxsec*2
  • DISxsec/2

The comparison plot shows that even these drastic changes have very little influence on the final moments.

Normalization to SIDIS

We discussed the possibility of normalizing to the SIDIS count in data and rec MC. I have checked the number of SIDIS after the current normalizations and get for 00H

  • Data
    • raw SIDIS 567311
    • Normalization factor (2944201/6405565)(1/69208200)
    • Normalized SIDIS 3.767e-3
  • MC
    • raw SIDIS 10226419
    • Normalization factor (0.18585/569582271)
    • Normalized SIDIS 3.350e-3

I guess these agree well enough (well within the factor of two that I've tested) to make us feel safe and happy.

2000 Neon

pi+ and pi- plots comparing 2000 Deuterium and Neon moments with the same set of bad bins (all bins that are good for both data sets).

The neon data is unfolded with the smearing matrix and DIS normalization from 00D. To see the possible size of the effect of using the wrong normalization, see the Normalization section above, with accompanying plot

Bad Bin Cut Tests

  • Latest slides are C0.4_cut_v3.1.pdf
  • Including (and with links to additional plots)
    • Comparison plots with different cuts and BB lists
    • Error comparisons
    • Plots of bins added when the cut is relaxed
    • Phi distributions
    • PEPSI Challenge plots with Reweighted Pythia

Smearing Matrix statistical error propagation

  • Math
  • Stuffs

Plot layout options

All the options for plot layout are presented in these slides, and the corresponding plots are in the same directory

Things yet to be done

  • Implement HTC
  • Finish working on propagation of smear matrix statistics
  • Write up and implement math to propagate 500-bin systematics to the projections
  • ( Vary # kinematic bins - Francesca has done a preliminary version )
  • Vary # phi bins
  • Estimate a systematic on the (cherry picked) projections due to bad bins using the parameterization
  • ( Mixed event sample -- Francesca is working on this )

July 2009

Minutes for cosphi meeting of July 16th, 2009

Xcheck

June 2009

Minutes for DC78 meetings of June 11th, 2009

New MC productions

Statistics

New 1D projection method and Q^2 binning

Binning studies


Additional modulations in the fit: cos3phi && (sinphi __ cos4phi)

February 2009

Topics to discuss

  • what type of draft are thinking? #pages? Journal?
  • what is needed for release?
    • any sort of 500 moments release?
      • systematics in 4D?
    • release of fit parameters??

Xcheck Status

XC status (Rebecca input files)

Plots and tests

  • 2007: unfold the 07 production using
    • /mcdata06/DATA/RICH_SYSTEMATIC_STUDIES/PYTHIA/RESULTS_P06_Q2cut_RAD_MSTP18_3_2004C_MSEL2_STD from Elke
    • 2006 MC (wrong beam charge) (Francesca)


Bad Bins tests:

               good bins         bad bins
  for pi+        13401             2321
  for pi-        11167             1756


  • Rebecca's Bad Bin studies
    • plots
      • Standard Francesca 0.4 cut
      • Francesca 0.3 cut
      • Francesca 0.5 cut
      • Rebecca cut, 1 filled bin, 12 bin gap allowed (NoBB)
    • View of BBs here


pi+/pi- difference:

RICH systematics


Rebecca's Born Cross section plots

    • Plot the born (MC) xsec in 500 bins, in the same format as the bad bin plots, to see where xsec is small
    • Xsec plots
    • Compare to Bad Bin plots
    • Mostly ok


Rebecca's studies about potential bad bin savings with more data / pre-combined years

    • Combine "raw" 00 and 06 data to estimate potential saving in BBs
    • Combine all data (00H, 00D, 05D, 06H) to estimate what x4 data give for BBs
    • pi+ Compared to 00H and Compared to 06H
      • After checking data remove 1325 bins; After checking born remove 1328 bin; After checking mig remove 1494 bins; After remove Fran bad bins 1818 bins.
        • Not MC limited
    • k- Compared to 00H and Compared to 06H
      • After checking data remove 2194 bins; After checking born remove 2205 bins; After checking mig remove 2348 bins; After remove Fran bad bins 3575 bins.
        • Not MC limited
      • Green - always a good bin
      • Purple - Not a good bin for this year alone, but good when 00H and 06H are averaged
      • Blue - Not good except when 00H and 06H raw data is combined
      • Red - Not good expect when 00H, 00D, 05D, 06H raw data is combined
    • Can we implement a "giant chi^2" to fit all years simultaneously?
      • Is that worth implementing given that we have the 07 data?
        • 00+05+06 ~20M DIS; 06e+ + 07 ~30M DIS H and ~30M DIS D


Parametrization


Ami's rho plots

    • Since we believe the cross section of pi+ and pi- from rho decay are identical, by looking at kinematic distributions of these particles we can see "only the acceptance"


P_T distribution's plots (Alessandro)

Comparison of P_T distributions for h+/h- in Hydrogen/Deuterium target, in data/PYTHIA productions.

Things that came up in the xcheck

  • Make fiducial and DIS cuts while choosing a lepton or after?
  • What about when g1Track.g1MTrack=NULL... decided to ditch this track only, but??
  • "instability" in the moments due to input data
  • negative yields, what to do??
    • 72 bins with negative yields (06do pi+)
    • -0.28, -0.19, -0.15, -0.14*13, etc., 50 are <0.1
    • Pions are fine, most within (or nearly within) statistical error
      • see /user/rlamb/analysis/xcheck/Feb2/00Hpip_F_FIn/BadNegBins and /user/rlamb/analysis/xcheck/Feb2/06Hpip_F_FIn/BadNegBins
    • check kaons
  • Note that DIS count now includes removing charge symetric events
  • Fixed code to deal with multiple links to the g1MTrack lepton

Other updates:

  • New grid MC productions
  • HTC (?)
  • disNG xsec for normalization

Nov09 HTC 2track-to-beam

How it's done

  • Count DIS with 1track to beam
  • Loop through every pair of tracks in the event
    • Do 2track to beam
    • save events with a hadron and a lepton in the fiducial volume

Questions

  • Should we throw out events if there is no DIS lepton?
  • Vertex / Track probabilities
    • for DIS lepton
    • for events
  • Check with Alexander that we have the proper fiducial volume input
  • change vertD cut? remove?
  • ( call PIDLIB each time - YES )

To Do Lists

  • Gunar - fixes to GMC trans
    • use for 4pi xsec, truncated smearing matrix, background model systematic tests
  • trigger efficiency studies for 06 and 07 ??
  • FIX DIS xsec for data norm, get from disNG, not PYTHIA
  • XC everything
  • eliminate 2000??
  • study theta_qh cut
  • correction for low statistics (see Paul's thesis p 65) ==> Accounted for with PEPSI challenge?
  • correction for underestimate in fitting (see Uli's thesis p89) ==> Accounted for with PEPSI challenge?
  • Lagrange multipliers to constrain the fit to <1 ==> Possible??
  • cos(phi) cos(2phi) cross talk ==> additional moment studies??
  • Zaven's beam correction ==> Taken care of with HTC???
  • Test "Patty cuts" for wrong charge tracks (see her thesis ~p28) ==> Sylvester sees no effect
  • Calculate systematics from
    • Smearing matrix statistics
    • Smearing matrix generator (background)
    • 4pi xsection for unfolding
    • Other systematics included in hadron release ??
    • Misalignment
    • Additional moments in the Fit function
    • Calculate systematics in the 500 bins!!!!
    • Recalculate RICH systematics (PEPSI challenge)
    • Final PEPSI challenge (merge with final PEPSI)
    • error on cherry picking using param


  • Do we really need these?
    • cos(phi) cos(2phi) cross talk
    • add other years??
    • unfold in xF and/or Q^2 ?
    • unfold pi+ - pi- ???
    • use maximum likelihood fitting in each kinematic bin ???
    • use Steve's unbinned unfolding
    • Lagrange multipliers to constrain the fit to <1
    • QED radiation without peaking approximation
    • Trigger efficiency correction? systematic? But Bkg subtraction?
    • cos(nphi) spin states parallel/anti (A_LL) and beam-spin (A_LU)
    • pi0
    • nuclear targets?

Francesca:

  • start MC production
  • change bin #s and sizes
  • redo all sys checks after binning definition
  • implement new bining & cherry picking
  • implement prefactor elimination
  • fit all years together
  • draft!!!!

Rebecca

  • Draft / Report text
  • compare 1Gev and 2Gev results (BBs, moments)
  • Vary # phi bins from 8+
    • Compare to a PEPSI challenge
    • where does it break down (min/max)
  • HTC test
    • compare to HRC
    • probability cuts
    • misalignment (top-bottom, left-right)
    • ( check effect on DIS number )
  • Year dependence tests
    • 00 misalignment? do I see it?
    • t-test for 06e- vs 06e+ ? with HTC?
    • quadrant studies
  • Code for systematics in 500 bins and projecting systematics
  • Steve's method for evaluating smearing matrix sys. error (see "smearing error stuff" in DRAFTS folder)
  • RICH PEPSI - do separately, then incorporate into final PEPSI (with Rewt pythia from smearing production)
  • RICH systematic
  • Email Alessandro about PF
  • Compare to COMPASS point w/o PF
  • disNG for DIS xsec

MC productions

See the release report page 6

And also productions 5-11 on the MC Productions page

MC productions needed

Format: Geometry Beam Target 

Full spectrometer simulated:

  geometry         tracking        generated events (Millions)
 2000 proton         hrc               240                                                            DONE!
 2000 neutron        hrc               120                                                            DONE!
             
 2005 proton         htc               120                                                            DONE!
 2005 neutron        htc               120                                                            DONE!
             
 
2006 e- proton       htc               240                                                          In progress....
2006 e+ proton       htc               720       (can be used for 2006 and 2007 e+ data)            In progress....
2006 e+ neutron      htc               120       (can be used for 2006 and 2007 e+ data)            In progress....

 
Productions with CALO ON (for pi0) && RICH ON

Born productions (Born does not depend on cell geometry or beam charge):

 geometry used       generated events (Millions)
2000 e+ proton           345                                                                         DONE!
2000 e+ neutron         172,5                                                                        DONE!


  • increased statistics for Pmatrix center tune ??
  • Xsec (no seclector) productions --> must be disNG!!
    • 06 e+ H (for 06, 07 data)
    • 06 e+ D (for 06, 07 data)
    • 06 e- H (for 06 data)
    • 00 e+ H (for 00 data) >> we have already
    • 00 e+ D (for 00 data) >> we have already
    • 00 e- D (for 05 data) >> to be checked

Data Statistics

  • 00H 5.1M DIS
    • 723,078 h+
      • 518,937 pi+
      • 90,932 k+
    • 411,691 h-
      • 360,062 pi-
      • 31,251 k-
  • 00D 6.4M DIS
    • 927,435 h+
      • 666,364 pi+
      • 114,840 k+
    • 606,733 h-
      • 532,020 pi-
      • 45,410 k-
  • 05D 4.2M DIS
    • 602,337 h+
      • 431,050 pi+
      • 73,130 k+
    • 393,002 h-
      • 343,021 pi-
      • 29,163 k-
  • 06H 5.1M DIS
    • 734,657 h+
      • 524,977 pi+
      • 92,514 k+
    • 412,451 h-
      • 358,385 pi-
      • 31,783 k-
  • Estimated:
    • 06He+ 23.2M DIS
    • 06De+ 5.6M DIS
    • 07H 7.1M DIS
    • 07D 26.1M DIS

XC status Dec 09, 2trk HTC

D

  • 2000
    • Data
      • pi+ 96 diff
    • disNG
      • 2595540.500000 DIS events
      • IEVGEN_tot = 69189498
      • extraweight_avg = 1.000000


  • 2005
    • Data
      • pi+ 238 diff
    • disNG
      • 2450084.750000 DIS events
      • IEVGEN_tot = 64306777
      • extraweight_avg = 1.000000


  • 2006 e+
    • Data
      • pi+ 13 diff
    • disNG
      • 2564853.000000 DIS events
      • IEVGEN_tot = 66313166
      • extraweight_avg = 1.000000
  • 2007

H

  • 2000
    • disNG
      • 2944201.750000 DIS events
      • IEVGEN_tot = 69208197
      • extraweight_avg = 1.000000
  • 2006 e-
    • disNG
      • 2940749.250000 DIS events
      • IEVGEN_tot = 66993839
      • extraweight_avg = 1.000000
  • 2006 e+
    • disNG
      • 2909171.000000 DIS events
      • IEVGEN_tot = 66324755
      • extraweight_avg = 1.000000
  • 2007


XC status Oct 09, HTC

Not perfect but ok. See plots

Also need to XC with 1GeV hadron cut

06H positron MC

  • Rebecca:
    • IEVGEN_tot: 1768006185
    • extraweight_avg: 0.185854
  • Francesca agrees

pi+

  • All elements:
    • 6047 36282 244333 MIGMAT_06e+/pp+_all_1%
  • Diagonal elements:
    • 51 306 2180 MIGMAT_06e+/pp+_diag_1%
  • Off-diag elements:
    • 5996 35976 242153 MIGMAT_06e+/pp+_off_1%
  • 6047 diff>1%
  • 51 on the diagonal elements, the other off-diag

pi-

  • Fr has 1011 more events than Re
  • 4692 diff > 1%
  • 1 diff of 6
  • 6 diff of 5
  • 15 diff of 4

07c2 H data

  • DIS disagree as of Oct 12 2009
    • 16777200 Rebecca
    • 17952716 Francesca

but only one diff in the small test list /user/giordano/ANALISI/test07.list

Rebecca sees the same DIS # in hrc and htc for the full 07 but for a short test run of 10 runs htc is 10% less :(

pi+

  • Fr has ~12 more
  • 206 diff
    • 2 diff of 4
    • 2 diff of 3

pi-

  • Fr has ~28 more
  • 173 diff
    • 1 diff of 3

Born

  • Rebecca
    • IEVGEN_tot: 345000000
    • extraweight_avg: 0.16541
  • Francesca agrees

pi+

  • 1%    25 diff

pi-

  • 1%    13 diff

XC status Jul09

pi+

  • 00H
    • Data
    • Born
    • Mig Matr
    • Dcos=, Dcos2=, cos=, cos2=
    • Cov:
      • off diag #=
      • diag cos #=
      • diag cos2 #=

XC status Feb09

Plots are here

pi+

  • 00H
    • Data 390 diff/TOT= 0.75E-03 --> 20
    • Born 5 diff/TOT= 0.11E-05
    • Mig Matr 365 diff/TOT= 0.92E-04
    • Dcos=7, Dcos2=7, cos=60, cos2=81
    • Cov:
      • off diag #=91546
      • diag cos #=12
      • diag cos2 #=13
  • 06H
    • Data 369 diff/TOT= 0.70E-03 --> 23
    • Born 7 diff/TOT= 0.15E-05
    • Mig Matr 180 diff/TOT= 0.89E-04
    • Dcos=9, Dcos2=7, cos=91, cos2=107
    • Cov:
      • off diag #=113434
      • diag cos #=20
      • diag cos2 #=18
  • H avg(00,06)
    • Dcos=3, Dcos2=5, cos=86, cos2=113
    • Cov
      • off diag #=103208
      • diag cos #=10
      • diag cos2 #=9
  • 00D
    • Data 426 diff/TOT= 0.64E-03 --> 26
    • Born 10 diff/TOT= 0.23E-05
    • Mig Matr 377 diff/TOT= 0.85E-04
    • Dcos=7, Dcos2=8, cos=79, cos2=105
    • Cov:
      • off diag #=113256
      • diag cos #=16
      • diag cos2 #=16
  • 05D
    • Data 347 diff/TOT= 0.81E-03 --> 14
    • Born 3 diff/TOT= 0.68E-06
    • Mig Matr 339 diff/TOT= 0.89E-04
    • Dcos=4, Dcos2=7, cos=63, cos2=75
    • Cov:
      • off diag #=108772
      • diag cos #=18
      • diag cos2 #=19
  • D avg(00,05)
    • Dcos=2, Dcos2=4, cos=72, cos2=97
    • Cov
      • off diag #=106804
      • diag cos #=13
      • diag cos2 #=12

pi-

  • 00H
    • Data 352 diff/TOT= 0.98E-03 --> 16
    • Born 8 diff/TOT= 0.23E-05
    • Mig Matr 250 diff/TOT= 0.83E-04
    • Dcos=7, Dcos2=8, cos=62, cos2=71
    • Cov:
      • off diag #=83406
      • diag cos #=19
      • diag cos2 #=16
  • 06H
    • Data 419 diff/TOT= 0.12E-02 --> 31
    • Born 9 diff/TOT= 0.26E-05
    • Mig Matr 182 diff/TOT= 0.12E-03
    • Dcos=13, Dcos2=12, cos=101, cos2=118
    • Cov:
      • off diag #=106094
      • diag cos #=25
      • diag cos2 #=24
  • H avg(00,06)
    • Dcos=5, Dcos2=6, cos=92, cos2=107
    • Cov
      • off diag #=97680
      • diag cos #=10
      • diag cos2 #=16
  • 00D
    • Data 397 diff/TOT= 0.75E-03 --> 15
    • Born 19 diff/TOT= 0.52E-05
    • Mig Matr 270 diff/TOT= 0.85E-04
    • Dcos=6, Dcos2=4, cos=77, cos2=81
    • Cov:
      • off diag #=88744
      • diag cos #=13
      • diag cos2 #=18
  • 05D
    • Data 404 diff/TOT= 0.12E-02 --> 16
    • Born 7 diff/TOT= 0.19E-05
    • Mig Matr 284 diff/TOT= 0.90E-04
    • Dcos=4, Dcos2=7, cos=63, cos2=75
    • Cov:
      • off diag #=85004
      • diag cos #=12
      • diag cos2 #=11
  • D avg(00,05)
    • Dcos=2, Dcos2=1, cos=80, cos2=81
    • Cov:
      • off diag #=83296
      • diag cos #=11
      • diag cos2 #=10

K+

  • 00H
    • Data 190 diff/TOT= 0.21E-02 -->16
    • Born 10 diff/TOT= 0.12E-04
    • Mig Matr 120 diff/TOT= 0.17E-03
    • Dcos=4, Dcos2=4, cos=45, cos2=47
    • Cov:
      • off diag #=41846
      • diag cos #=10
      • diag cos2 #=12
  • 06H
    • Data 194 diff/TOT= 0.21E-02 --> 21
    • Born 17 diff/TOT= 0.20E-04
    • Mig Matr 75 diff/TOT= 0.21E-03
    • Dcos=6, Dcos2=4, cos=47, cos2=54
    • Cov:
      • off diag #=38464
      • diag cos #=12
      • diag cos2 #=9
  • H avg(00,06)
    • Dcos=4, Dcos2=4, cos=48, cos2=59
    • Cov
      • off diag #=44026
      • diag cos #=10
      • diag cos2 #=8
  • 00D
    • Data 238 diff/TOT= 0.21E-02 --> 15
    • Born 12 diff/TOT= 0.15E-04
    • Mig Matr 98 diff/TOT= 0.14E-03
    • Dcos=6, Dcos2=5, cos=43, cos2=46
    • Cov:
      • off diag #=50278
      • diag cos #=10
      • diag cos2 #=14
  • 05D
    • Data 193 diff/TOT= 0.26E-02 --> 6
    • Born 16 diff/TOT= 0.20E-04
    • Mig Matr 59 diff/TOT= 0.87E-04
    • Dcos=1, Dcos2=2, cos=20, cos2=29
    • Cov:
      • off diag #=24478
      • diag cos #=5
      • diag cos2 #=8
  • D avg(00,05)
    • Dcos=3, Dcos2=2, cos=36, cos2=38
    • Cov:
      • off diag #=46028
      • diag cos #=8
      • diag cos2 #=6

K-

  • 00H
    • Data 96 diff/TOT= 0.31E-02 --> 6
    • Born 23 diff/TOT= 0.25E-04 --> 12
    • Mig Matr 55 diff/TOT= 0.14E-03
    • Dcos=2, Dcos2=3, cos=11, cos2=13
    • Cov:
      • off diag #=6942
      • diag cos #=3
      • diag cos2 #=3
  • 06H
    • Data 137 diff/TOT= 0.43E-02 --> 13
    • Born 8 diff/TOT= 0.17E-04
    • Mig Matr 37 diff/TOT= 0.19E-03
    • Dcos=4, Dcos2=4, cos=22, cos2=22
    • Cov:
      • off diag #=10512
      • diag cos #=7
      • diag cos2 #=8
  • H avg(00,06)
    • Dcos=1, Dcos2=1, cos=23, cos2=21
    • Cov
      • off diag #=9146
      • diag cos #=2
      • diag cos2 #=4
  • 00D
    • Data 144 diff/TOT= 0.32E-02 --> 7
    • Born 10 diff/TOT= 0.20E-04
    • Mig Matr 53 diff/TOT= 0.13E-03
    • Dcos=0, Dcos2=0, cos=15, cos2=22
    • Cov:
      • off diag #=11244
      • diag cos #=0
      • diag cos2 #=1
  • 05D
    • Data 110 diff/TOT= 0.38E-02 --> 3
    • Born 7 diff/TOT= 0.14E-04
    • Mig Matr 52 diff/TOT= 0.13E-03
    • Dcos=2, Dcos2=3, cos=10, cos2=4
    • Cov:
      • off diag #=6330
      • diag cos #=2
      • diag cos2 #=3
  • D avg(00,05)
    • Dcos=0, Dcos2=0, cos=14, cos2=14
    • Cov:
      • off diag #=9862
      • diag cos #=0
      • diag cos2 #=0

Francesca vs Rebecca, both with Rebecca's input files (last)

  • 00H
    • Dcos= Dcos2= cos= cos2= 0
    • COV
      • off diag #=448
      • diag cos #=0
      • diag cos2 #=0
  • 06H
    • Dcos= Dcos2= cos= cos2= 0
    • COV
      • off diag #=80
      • diag cos #=0
      • diag cos2 #=0
  • Average 00H+06H
    • Dcos= Dcos2= cos= cos2= 0
    • COV
      • off diag #=368
      • diag cos #=0
      • diag cos2 #=0


XC status

  • Number of bins with differences >1%
  • Rebecca's files are in /user/rlamb/analysis/xcheck/Feb2/

Raw input Francesca vs Rebecca

  • 00H
    • Raw data = 1
    • Raw mig = 366
    • Raw born = 6
  • 06H
    • Raw data = 2
    • Raw mig = 181
    • Raw born = 7

Francesca vs Rebecca, both with Rebecca's input files

  • 00H
    • Mom = 0
    • Cov = 5921
  • 06H
    • XCX_josh = 520 (16 real diff, the rest Fran=1)
    • XCX_rebe(full) = 1026 (19 real, the rest Rebe=0)
    • Mom = Dcos2=1, cos=1
    • Cov = 6571
  • H avg(00,06)
    • Mom = 227 Dcos=101, Dcos2=99, cos=3, cos2=14
    • Cov = 431,241

Francesca vs Rebecca, both with Francesca's input files

  • 00H
    • Mom = 0
    • Cov =6049 (582 >10% diff)
  • 06H
    • Mom= 3, cos=1, cos2=1, Dcos2=1
    • Cov = 6745 (612 >10% diff)
  • Havg(00,06)
    • Mom = 16, Dcos2=1, cos=2, cos2=13
    • Cov = 20143 (1925 >10%)

Francesca vs Rebecca, full XC

  • 00H
    • Mom = Dcos=7, Dcos2=7, cos=61, cos2=82
    • Cov = 94072
  • 06H
    • Mom = Dcos=8, Dcos2=9, cos=93, cos2=105
    • Cov = 115,800
  • H avg(00,06)
    • Mom = 212 Dcos=3, Dcos2=6, cos=85, cos2=117
    • Cov = 108909

Average code check

  • Francesca + Rebecca input + Rebecca 00 covariance + Rebecca 06 covariance
    • Cos = 0 Dcos=0, cos2=2, Dcos2=0
    • Cov= 20 off-diagonal elements


Relevant Papers

  • On Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle h_1^\perp ~ k_T/k f_(1T)^\perp}
    • M.Burkardt Phys.Rev. d72:094020, 2005 in particular at page 9 few lines just before "summary"
    • in M.Goeckeler, Phys.Rev. Lett. 98:222001, 2007 there are the calculations of the factor k_T and k in lattice
    • Alexei and company's paper
    • Daniels calculations for cos2\phi @tevatron arXiv:0712.0777 where he uses GRV98 and factorizes the pT dependence.
Abgerufen von „https://hermes-wiki.desy.de/index.php?title=CosPhi_Status&oldid=12332“