TWiki> Main Web>TWikiUsers>ShingoSakai>Analysis>02012012 (revision 14)EditAttachPDF

Execute Summary

  • This is a page for the execute summary of 01272012
  • The current understanding is:
    • The trivial observation is that the small stat sample causes difficulties to draw strong conclusions - these remain rather conservative
    • The estimations of hadron contamination of the electron sample after E/p cut is ~7-9% with a generous uncertainty of 4%.
    • The 10-20% difference between the efficiencies derived from MC and V0 is caused by the efficiency in cluster-track matching with some contribution from E/p cut - both of these are sensitive to hadron contamination. By comparison of the cluster-track matching efficiencies with another data sample - LHC11a LHC11a (in the pT of interest - where the plateau is reached - easiest to conclude) we see that:
      • the V0 efficiency from data (c-t matching) shows a dependence as for inclusive track-matching (any particle - so in electron jargon : hadron contaminated) - somewhat below 0.8
      • whereas the efficiency from MC shows values typical for purified electron candidates - above 0.8
    • The above observation points to MC efficiency is to be used, however, of course the efficiency for V0s in MC will be investigated and compared to data - to be done right away. Also the closer investigation of at the E/p within the V0 sample in data will be made - this would be a good consistency check for the contamination of the electron sample in V0 selections.
  • The resulting x-section is shown.

Hadron contamination & electron yields from E/p distribution

The relative ratios of integrals in the upward vs downward of the signal within the E/p peak

  • Integrated E/p distribution from -n sigma to mean (downward) and from mean to n sigma (upward), and calculate the ratio downward / upward
  • The ratio deviating from 1 indicated asymmetry of the integrals:
    • low-pT: hadron contamination seen in the n-sigma selection - directly seen in the E/p distributions - the hadron tail under the gaussian peak (see E/p fits)
    • at high-pT the extracted ratio is consistent with 1 within the stat error bars but the deviation from 1. - especially for the 3sigma selection does indicate a systematic effect - but within the taken into account in the contamination estimation and uncertainty


Raw yield extraction from E/p cut based on n sigma cut

  • Compared electron yields between n sigma cut (-3 sigma < E/p < 3 sigma) & fix cut (0.8 < E/p < 1.3)
    • denominator ; electron yields from 0.8 <E/p <1.3 / numerator ; electron yields from nSIgma cut
  • Expected difference between 2 sigma cut and 3 sigma cut is ~ 4 % - this plot below is consistent within 1-2% with this limit and that points to a systematic uncertainty on the signal extraction of this order. The low pT (3sigma selection) deviation from 1 is still caused by the hadron tail in E/p. The high pT deviation from 1 is on the order of 10% but consistent within the stat. error bars - one could consider assigning an uncertainty of few %.


Efficiency for TPC nSigma dE/dx

  • the nSIgma efficiency (-1.5 < nSigma < 3.0) extracted from the fits is constant at 93% (for all pT's)
    • expected from a Gaussian distribution


Efficiency for EMCal from V0 and MC

Efficiency from E/p and track matching from V0 and MC

  • Note:
    • The track matching efficiency is purely track-by-cluster (geometry, detector perf., cluster cuts etc)
    • The E/p efficiency is efficiency of the E/p cut (after the track matching)
    • Both are sensitive to hadron contamination (and purity of the electron sample)


Pseudorapidity systematics: efficiency of E/p and matching from V0



  • No eta dependence both Track matching and E/p cut at low pT.
  • At high pT, some discrepancy was observed but the discrepancy is within error

Pseudorapidity systematics: efficiency of E/p and matching from MC



  • No eta dependence both Track matching and E/p cut


HFE spectrum (corrected with 10f6a MB MC)


Further checks

  • The comparison of efficiency (E/p and matching) from V0 and MC show the discrepancy at high pT. Following can help to understand the discrepancy:
    • Compare V0 efficiency in data and MC
    • Check E/p in V0 sample
Topic attachments
I Attachment History Action Size Date Who Comment
PDFpdf EMCalSummary.pdf r1 manage 454.2 K 2012-02-02 - 11:26 Ssakai  
PNGpng PWG3-Preliminary-029.png r2 r1 manage 17.7 K 2012-02-02 - 13:53 Ssakai  
Edit | Attach | Watch | Print version | History: r18 | r16 < r15 < r14 < r13 | Backlinks | Raw View | Raw edit | More topic actions...
Topic revision: r14 - 2012-02-02 - Ssakai
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback