-- CatherineSilvestre - 13 Mar 2014

This is AstridVauthierDailyLogBook topic.

Analysis pp 7 TeV :

xE Distribution

April 29, 2015 Decide to re-use the Twiki after a long break. The analysis is moving on. All the code is done. We are in check period.

For the moment we check if all the gamma decay contributions are compatible --> TRUE

Check if UE is negligible --> TRUE

Check if efficiency in away side and UE is compatible --> FALSE


The method has completely changed since Nicolas's thesis. Now we take into account the contamination coming from gamma decays (neutral meson) in the signal region.



Group Meetings :

May 27, 2015

April 29, 2015

31 oct 2014

27 oct 2014

Meetings :

Old Stuffs (Internship) :






18/06/2014 : Soutenance de stage









Periods QA


  • IssuesInQAPeriods: Final recap document for QA ---> To access notes about the QA download the document

  • QAPeriodsMB.pdf: QA for Min Bias data set. Less plot due to less statistic

  • IssuesInQAPeriods.pdf: Recap document for the issues in the period per period QA for MB and triggered data sets. The explanations about the issues and what we understand are available in the notes of the PDF document


  • QAPeriodsGood.pdf: Analyse de qualité des données période par période pour données triggées de niveau 1

3 periodes : LHC13d (p-Pb), LHC13e (p-Pb), LHC13f (Pb-p). LHC13f choisie comme période de référence

=> Tendance en eta

Runs QA




  • QARuns.pdf: QA run per run for 13d, e, f periods. Data triggered level 1 QA perform just for TH1F because of the statistic

  • QARunsMB.pdf: QA run per run for 13d, e, f periods. Data Min bias


ANALYSE p-Pb et pp

x_E distribution


Weird distribution for UE => I check wether we obtain the same distribution of all types of trigger particle. I don't observe that in pp but I do in pPb for isolated particles.

  • comparisonUEAllTypeTiggParticle.pdf: xE distribution for UE for different type of trigger particle (isolated and non isolated pi0 or cluster) for pp data : not the same with repeat to the trigger particle


Run the pp data on my code to crosscheck pp results : I don't have the same distributions => I do the x_E distribution for non isolated pi0 and compare to the UE distribution to see if we obtain the same shape and ratio as obtained in pp. It seems different, I have to check my code.


Version finale du rapport de stage


errors assessment


Bad formula for photons x_E distribution -----> we have to re-implement the correct formula and obtain all the plot again Stop the comparison with LHC11cd for now (report)


errors assessment


Comportement chaotique en fonction des données LHC11 que j'utilise --> le facteur 2 que l'on trouve entre pp et pPb n'est certainement pas du à la physique ! Les données sont celles données par Gustavo du 29 avril 2013 et du 18 mars 2014.


A cause des niveaux de trigger il faut faire attention lorsqu'on utilise les données des niveau L1 et L2 ensemble. On choisit de faire l'analyse avec les données L1(seuil de 12 GeV /c) pour le domaine en p_T^trigg de 12 à 25 GeV /c et avec les données L2 (seuil à 7 GeV /c) pour le range 10-11 GeV /c. La stat disponible dans L2 pour 10-11 GeV /c est 10 fois plus faible que pour L1 de 12 à 25 GeV /c ----> Barre d'erreur plus importante. Les valeurs de la pureté ont été calculées par Catherine.

  • xEPhotonL2_10_12.pdf: xE for isolated photon : the fit does not converged because of the small error bars at high x_E where the distribution is negative.

Les barres d'erreurs semblent très grandes (pas trop de stat 10 fois moins pour L2 10-12 que pour L1 12-25) mais lorsqu'on fait la propagation d'une soustraction on somme en quadrature les erreurs.

Erreur dans la propagation pour les ratios GoodRuns /AllRuns et GoodRuns /BadRuns --> problème réglé, nouveaux plots ici :


Update of x_E distribution : I did a non weighted average of each bin in [12,25] GeV /c --> it's not the proper way to do the analysis. One must scale the total distribution with the number of particles in 12-25 and not scale each bin distribution with number of particle in the bin. It doesn't change a lot but it could be interesting to do the comparison between the two methods. ---> !!!!!!!!!!!!!!! To do !!!!!!!!!!!!!!!


Check if the values for left and right side UE \Delta\Phi are OK -----> There is no hole for left side as confirmed by the fit

  • compareUEFit09.pdf: Fit of the ratio between left and right side UE with one fixed parameter = 0.9

Comparison between AllRuns data and AllGoodRuns data -----> slight difference between the two datasets -> see if we choose to remove others bad runs when we go back to QA

Comparison between L1 data and L1+L2 data : L2 has 50% of L1 statistic but for the xE histograms the biggest part of the stat is below 12 GeV /c so we don't use it for the analysis.

Comparison between the analysis for \p_{T}^{trigg} \in [10,25] GeV /c and \in [12,25] GeV /c : the purity is very low for low \p_{T}^{trig} not sure this is the best option


Do the plots with merged data <=> LHC13d, e, f without identified bad runs from QA there is 14 bins in p_T^trigg [12,25]

  • 3xE.pdf: three part of the equation for x_E distribution for isolated photon ----> Check if our x_E distribution has sense and do the ratio over f(x_E^UE) to be sure that the UE is negligible for high x_E where we don't have information on it

  • CompareUELeftRight.pdf: UE right and left comparison du to weird value in analysis code ----> we want to be sure we can use this for now

  • FitxE.pdf: fit the cluster and pi0 x_E distribution to compare with the value obtained by Nicolas ---> same order of magnitude

  • xEPhoton.pdf: x_E distribution for isolated photon with 12 bins in p_T^trigg (means still fake errors for two bins) ----> Maybe it will be interesting to search for a good binning (i.e minimise the error bars)

When we look carefully at the x_E distribution (i.e TH2F in the plots analysis) we see that it would be great to perform the analysis from 10 Gev/c to 25 Gev/c ------> we need L2 data PRELIMINARY plots

  • 3xE1_25.pdf: the three part of the isolated photon x_E distribution formula p_T^trigg in [10,25] GeV /c ----> fake value of the purity for the 2 first bins

  • xEPhoton10_25.pdf: we clearly see that we reach the high x_E region with more statistic here ! See if always true with good values of the purity

For both we have negative values for certain bins -> What do we do ? TMath::Abs() ?


Compare the three data periods ----> we want to merge data Two different plots : one with 14 bins in p_T^trigg (still a fake error on two bins for purity), and one with 1 bin in p_T^trigg

  • Compare_d_e_f_NoBin.pdf: compare the three period in order to be sure that we can merge the data with no bin in p_T^trigg



do the plot with LHC13f L1 trigg data only

  • xEPhoton.pdf: first plot of the x_E distribution for isolated photon


We try to understand how the purity estimation will affect the isolated photon x_E distribution


  • xE12BinsComparePurityMethod.pdf: Comparison of x_E for isolated photon for different purity estimation method -----> all the method are consistent => for now we can choose one method and choose further


  • ComparePurity.pdf: compare the value of the purity for different value of p_T^trigg bins

  • binOverConstantPurity.pdf: comparison of x_E for isolated photon for bin purity vs contant purity ----> better to bin but still coherent if we choose a mean value for purity





Topic attachments
I Attachment History Action Size Date Who Comment
PDFpdf IssuesInQARuns.pdf r1 manage 2431.9 K 2014-04-30 - 15:53 AstridVauthier Recap document for the issues in the run by run Quality Analysis
PDFpdf Presentation.pdf r2 r1 manage 814.0 K 2014-03-31 - 15:55 AstridVauthier Présentation Réunion de groupe 31 mars 2014
PDFpdf Presentation_15avril2014.pdf r1 manage 1182.5 K 2014-04-15 - 10:14 AstridVauthier Présentation réunion de groupe du 15 avril 2014
PDFpdf QAPeriodsGood.pdf r1 manage 21925.9 K 2014-04-02 - 13:00 AstridVauthier Analyse de qualité des données période par période
PDFpdf QAPeriodsMB.pdf r1 manage 13784.6 K 2014-04-14 - 11:08 AstridVauthier QA for Min Bias data set. Less plot due to less statistic
PDFpdf QARuns.pdf r1 manage 25045.3 K 2014-04-14 - 11:18 AstridVauthier QA run per run for 13d, e, f periods. Data triggered level 1 QA perform just for TH1F because of the statistic
PDFpdf QARunsMB.pdf r1 manage 9506.5 K 2014-04-14 - 11:21 AstridVauthier QA run per run for 13d, e, f periods. Data Min bias
PDFpdf presentation_14mars2014.pdf r1 manage 303.3 K 2014-04-01 - 10:17 AstridVauthier Présentation réunion de groupe 14 mars 2014
Edit | Attach | Watch | Print version | History: r44 < r43 < r42 < r41 < r40 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r44 - 2015-09-17 - AstridVauthier
    • 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 &© 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback