1. Samples

/PhotonJet_Pt15/Spring10-START3X_V26_S09_preproduction-v1/GEN-SIM-RECO; contains 1221770 events; $ \sigma = 192.2 nb $

2. Events Selection

Reconstruction of objects: $ \gamma $, Probe Jet

  • $ \gamma-candidate $ with highest Pt is selected

  • Isolation criteria are applied to selected $ \gamma-candidate $; Isolation criteria are taken from http://cms.cern.ch/iCMS/jsp/openfile.jsp?tp=draft&files=AN2009_012_v5.pdf ( CMS AN-2009/012 "Jet energy calibration with photon+jet events" ) , page 11, table 6. "Tight" selection is used. Cuts are applied to HCAL iso, ECAL iso, $ N_{tracks}$ and $ p_T^{sum, track} $ variables. Cuts on cluster profile variables (cluster major, cluster minor) are not considered.

  • All reconstructed jets inside cone 0.1 around $ \gamma $ direction are excluded from consideration. It's done because reconstructed gammas appear both in jet and gamma listings. Among the rest of the jets, the jet with highest Pt is selected. It's called "Probe Jet". Further in this discription I'll refer to it simply as to the Jet.

Cuts applied:

  • Photon: $  | \eta_{\gamma}| < 2, Pt_{\gamma}>20GeV $;    Probe Jet: $ \Delta_{\gamma, j} > 150^o, 4.8>|\eta_j|>1.4, targetE>10GeV, Em_{fraction}<0.4 $;     Third Jet: $ Pt<5GeV $

3. Some Distributions

There is some distributions for events selected using criteria described above.

3.1 These are ratios of targetE to hadronic energy of the Probe Jet for different $ \eta $ rings:

hfm_21.gif hfp_21.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_22.gif hfp_22.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_23.gif hfp_23.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_24.gif hfp_24.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_25.gif hfp_25.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_26.gif hfp_26.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_27.gif hfp_27.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_28.gif hfp_28.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_29.gif hfp_29.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_30.gif hfp_30.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_31.gif hfp_31.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_32.gif hfp_32.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_33.gif hfp_33.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_34.gif hfp_34.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_35.gif hfp_35.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_36.gif hfp_36.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_37.gif hfp_37.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_38.gif hfp_38.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_39.gif hfp_39.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_40.gif hfp_40.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_41.gif hfp_41.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

4 Calibration vs. different cuts variation

4.1 cut on $ Pt_\gamma $

        Photon: $  | \eta_{gamma}| < 2 $;    Probe Jet: $ \Delta_{\gamma, j} > 150^o; |\eta_j|<4.8 $; targetE>10GeV;     Third Jet: Pt<5GeV

pt1.gif
Fig.1 'calibration vs. $ \eta_{index} $' for different values of the cut on $ Pt_\gamma $;



4.2 cut on $ \eta_\gamma $

        Probe Jet: $ \Delta_{\gamma, j} > 150^o; |\eta_j|<4.8 $; targetE>10GeV;     Third Jet: Pt<5GeV

pt_15_eta.gif pt_20_eta.gif
Fig.2 'calibration vs. $ \eta_{index} $' for different values of the cut on $ \eta_\gamma $:    a) $ Pt_\gamma>15GeV $    b) $ Pt_\gamma>20GeV $



4.3 cut on Pt of the Third Jet

        Photon: $ | \eta_{gamma}| < 2; |Pt_{gamma}|>20GeV $;    Probe Jet: $ \Delta_{\gamma, j} > 150^o; |\eta_j|<4.8 $; targetE>10GeV

pt_third.gif
Fig.3 'calibration vs. $ \eta_{index} $' for different values of the cut on Pt of the Third Jet;

5. L3 minimization

There is a description of L3 minimization algorithm:

The calibration coefficient $ C^j $ for $ j^{th} $ cell is calculated as follows:

$ C^j = \frac{w_1^j*E_1^{target} / HAD_1 + w_2^j*E_2^{target} / HAD_2 + ... + w_n^j*E_n^{target} / HAD_n } {w_1^j + w_2^j + ... + w_n^j} $,

where summing is done over n events, $ HAD_i $ is a hadronic energy of the jet in $ i^{th} $ event; $ E_i^{target} $ is an expected response of HCAL to the jet; $ w_i$ is an event weight of $ i^{th} $ event calculated as follows:

$ w_i = \frac{e^j}{HAD_i} * (1-|1-\frac{HAD_i}{E_i^{target}}|)^\alpha $, where $ \alpha = 0, 1, 2, ... $

$ e^j $ is an energy deposited at $ j^{th} $ cell

6. Some thoughts on calibration with L3 minimization

L3 procedure produces some minimization, but it works like a blackbox. We can influence on the result only changing the values of input parameters and the number of iterations (nIter) inside algorith itself.

The question is whether we can control this minimization somehow? What is the output of L3 algorithm? Please look at Section 3 of this page. The plots present there show the ratio of targetE to hadron energy deposition inside the probe jet for all jets which have at least one tower with a given $ \eta $. The average values of this prots give a crude approximation to the calibration coefficient for given $ \eta $. If we want to visualize L3 algorithm in terms of histograms we need to multiply each particular entry of histograms of Section 3 by weight which is the ratio of energy deposition inside cell with a given $ \eta $ to the overall hadron energy deposition inside the jet. This procedure result in some histograms, with the shape similar to histograms of Section 3.

Now let us look to these new histograms (given below). They are shown in blue and are superimposed on the histograms of Section 3 (ratio of targetE to hadron energy deposition inside the probe jet). These new histograms have a narrow peak with a long tail. L3 minimization takes mean values of these histograms. The erros of mean values are RMS/$ \sqrt N$. We can try to fit the peaks and take mean values of the fit. This result in much more precise mean values. The question is how "physical" is such a procedure. It depends on the nature of the tails. If they are due to HCAL energy resolution probably it's not quite good. If it's because of leakage of energy to the tube, this procedure is Ok. It allows us to select event with a strict energy balance between $ \gamma $ and probe jet.

hfm_21w.gif hfp_21w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_22w.gif hfp_22w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_23w.gif hfp_23w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_24w.gif hfp_24w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_25w.gif hfp_25w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_26w.gif hfp_26w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_27w.gif hfp_27w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_28w.gif hfp_28w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_29w.gif hfp_29w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_30w.gif hfp_30w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_31w.gif hfp_31w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_32w.gif hfp_32w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_33w.gif hfp_33w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_34w.gif hfp_34w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_35w.gif hfp_35w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_36w.gif hfp_36w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_37w.gif hfp_37w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_38w.gif hfp_38w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_39w.gif hfp_39w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_40w.gif hfp_40w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

hfm_41w.gif hfp_41w.gif
'N entries vs. targetE/$E_{Probe Jet}  $'   for:    a) HF-    b) HF+

7. Small Ntuples

Small Gamma+Jet ntuples can be found at /afs/cern.ch/cms/CAF/CMSALCA/ALCA_HCALCALIB/data/smallTestNtuples/GammaJet/ As I mentioned at Section 1 they were obtained from 1221770 PhotonJet_Pt15 events.

The format of the ntuples are the same as for dijets. Variables of tagJetP4 are filled with information about gamma.

-- AndreiKrokhotin - 23-Apr-2010

Topic attachments
I Attachment History Action Size Date Who Comment
GIFgif etPhot_etJet.gif r1 manage 5.5 K 2010-04-29 - 12:38 AndreiKrokhotin  
GIFgif hfm_21.gif r1 manage 9.4 K 2010-05-05 - 18:02 AndreiKrokhotin  
GIFgif hfm_21w.gif r1 manage 13.6 K 2010-07-07 - 10:55 AndreiKrokhotin  
GIFgif hfm_22.gif r1 manage 9.3 K 2010-05-05 - 18:02 AndreiKrokhotin  
GIFgif hfm_22w.gif r1 manage 13.4 K 2010-07-07 - 10:56 AndreiKrokhotin  
GIFgif hfm_23.gif r1 manage 9.4 K 2010-05-05 - 18:02 AndreiKrokhotin  
GIFgif hfm_23w.gif r1 manage 13.5 K 2010-07-07 - 10:56 AndreiKrokhotin  
GIFgif hfm_24.gif r1 manage 9.2 K 2010-05-05 - 18:03 AndreiKrokhotin  
GIFgif hfm_24w.gif r1 manage 13.3 K 2010-07-07 - 10:56 AndreiKrokhotin  
GIFgif hfm_25.gif r1 manage 9.4 K 2010-05-05 - 18:03 AndreiKrokhotin  
GIFgif hfm_25w.gif r1 manage 13.6 K 2010-07-07 - 10:56 AndreiKrokhotin  
GIFgif hfm_26.gif r1 manage 9.2 K 2010-05-05 - 18:03 AndreiKrokhotin  
GIFgif hfm_26w.gif r1 manage 13.3 K 2010-07-07 - 10:57 AndreiKrokhotin  
GIFgif hfm_27.gif r1 manage 8.8 K 2010-05-05 - 18:04 AndreiKrokhotin  
GIFgif hfm_27w.gif r1 manage 12.9 K 2010-07-07 - 10:57 AndreiKrokhotin  
GIFgif hfm_28.gif r1 manage 9.0 K 2010-05-05 - 18:04 AndreiKrokhotin  
GIFgif hfm_28w.gif r1 manage 13.3 K 2010-07-07 - 10:57 AndreiKrokhotin  
GIFgif hfm_29.gif r1 manage 9.3 K 2010-05-05 - 18:04 AndreiKrokhotin  
GIFgif hfm_29w.gif r1 manage 12.6 K 2010-07-07 - 10:57 AndreiKrokhotin  
GIFgif hfm_30.gif r1 manage 9.0 K 2010-05-05 - 18:05 AndreiKrokhotin  
GIFgif hfm_30w.gif r1 manage 13.3 K 2010-07-07 - 10:57 AndreiKrokhotin  
GIFgif hfm_31.gif r1 manage 8.7 K 2010-05-05 - 18:05 AndreiKrokhotin  
GIFgif hfm_31w.gif r1 manage 12.7 K 2010-07-07 - 10:58 AndreiKrokhotin  
GIFgif hfm_32.gif r1 manage 8.7 K 2010-05-05 - 18:05 AndreiKrokhotin  
GIFgif hfm_32w.gif r1 manage 12.6 K 2010-07-07 - 10:58 AndreiKrokhotin  
GIFgif hfm_33.gif r1 manage 8.9 K 2010-05-05 - 18:06 AndreiKrokhotin  
GIFgif hfm_33w.gif r1 manage 12.7 K 2010-07-07 - 10:58 AndreiKrokhotin  
GIFgif hfm_34.gif r1 manage 8.9 K 2010-05-05 - 18:06 AndreiKrokhotin  
GIFgif hfm_34w.gif r1 manage 12.9 K 2010-07-07 - 10:58 AndreiKrokhotin  
GIFgif hfm_35.gif r1 manage 9.0 K 2010-05-05 - 18:06 AndreiKrokhotin  
GIFgif hfm_35w.gif r1 manage 12.7 K 2010-07-07 - 10:58 AndreiKrokhotin  
GIFgif hfm_36.gif r1 manage 9.3 K 2010-05-05 - 18:07 AndreiKrokhotin  
GIFgif hfm_36w.gif r1 manage 13.9 K 2010-07-07 - 10:59 AndreiKrokhotin  
GIFgif hfm_37.gif r1 manage 9.7 K 2010-05-05 - 18:07 AndreiKrokhotin  
GIFgif hfm_37w.gif r1 manage 14.2 K 2010-07-07 - 10:59 AndreiKrokhotin  
GIFgif hfm_38.gif r1 manage 10.2 K 2010-05-05 - 18:07 AndreiKrokhotin  
GIFgif hfm_38w.gif r1 manage 14.9 K 2010-07-07 - 10:59 AndreiKrokhotin  
GIFgif hfm_39.gif r1 manage 10.6 K 2010-05-05 - 18:08 AndreiKrokhotin  
GIFgif hfm_39w.gif r1 manage 15.4 K 2010-07-07 - 10:59 AndreiKrokhotin  
GIFgif hfm_40.gif r1 manage 10.7 K 2010-05-05 - 18:08 AndreiKrokhotin  
GIFgif hfm_40w.gif r1 manage 15.2 K 2010-07-07 - 10:59 AndreiKrokhotin  
GIFgif hfm_41.gif r1 manage 10.2 K 2010-05-05 - 18:08 AndreiKrokhotin  
GIFgif hfm_41w.gif r1 manage 13.8 K 2010-07-07 - 11:00 AndreiKrokhotin  
GIFgif hfp_21.gif r1 manage 9.3 K 2010-05-05 - 18:08 AndreiKrokhotin  
GIFgif hfp_21w.gif r1 manage 13.4 K 2010-07-07 - 11:00 AndreiKrokhotin  
GIFgif hfp_22.gif r1 manage 9.2 K 2010-05-05 - 18:09 AndreiKrokhotin  
GIFgif hfp_22w.gif r1 manage 13.7 K 2010-07-07 - 11:00 AndreiKrokhotin  
GIFgif hfp_23.gif r1 manage 9.0 K 2010-05-05 - 18:09 AndreiKrokhotin  
GIFgif hfp_23w.gif r1 manage 13.4 K 2010-07-07 - 11:01 AndreiKrokhotin  
GIFgif hfp_24.gif r1 manage 9.2 K 2010-05-05 - 18:10 AndreiKrokhotin  
GIFgif hfp_24w.gif r1 manage 13.5 K 2010-07-07 - 11:01 AndreiKrokhotin  
GIFgif hfp_25.gif r1 manage 9.3 K 2010-05-05 - 18:10 AndreiKrokhotin  
GIFgif hfp_25w.gif r1 manage 13.8 K 2010-07-07 - 11:01 AndreiKrokhotin  
GIFgif hfp_26.gif r1 manage 9.3 K 2010-05-05 - 18:13 AndreiKrokhotin  
GIFgif hfp_26w.gif r1 manage 13.8 K 2010-07-07 - 11:01 AndreiKrokhotin  
GIFgif hfp_27.gif r1 manage 9.2 K 2010-05-05 - 18:13 AndreiKrokhotin  
GIFgif hfp_27w.gif r1 manage 13.0 K 2010-07-07 - 11:02 AndreiKrokhotin  
GIFgif hfp_28.gif r1 manage 9.0 K 2010-05-05 - 18:14 AndreiKrokhotin  
GIFgif hfp_28w.gif r1 manage 13.2 K 2010-07-07 - 11:02 AndreiKrokhotin  
GIFgif hfp_29.gif r1 manage 8.8 K 2010-05-05 - 18:14 AndreiKrokhotin  
GIFgif hfp_29w.gif r1 manage 12.3 K 2010-07-07 - 11:02 AndreiKrokhotin  
GIFgif hfp_30.gif r1 manage 8.8 K 2010-05-05 - 18:15 AndreiKrokhotin  
GIFgif hfp_30w.gif r1 manage 12.8 K 2010-07-07 - 11:02 AndreiKrokhotin  
GIFgif hfp_31.gif r1 manage 8.8 K 2010-05-05 - 18:15 AndreiKrokhotin  
GIFgif hfp_31w.gif r1 manage 12.6 K 2010-07-07 - 11:03 AndreiKrokhotin  
GIFgif hfp_32.gif r1 manage 8.8 K 2010-05-05 - 18:15 AndreiKrokhotin  
GIFgif hfp_32w.gif r1 manage 12.5 K 2010-07-07 - 11:03 AndreiKrokhotin  
GIFgif hfp_33.gif r1 manage 9.2 K 2010-05-05 - 18:16 AndreiKrokhotin  
GIFgif hfp_33w.gif r1 manage 12.7 K 2010-07-07 - 11:03 AndreiKrokhotin  
GIFgif hfp_34.gif r1 manage 9.1 K 2010-05-05 - 18:16 AndreiKrokhotin  
GIFgif hfp_34w.gif r1 manage 12.9 K 2010-07-07 - 11:04 AndreiKrokhotin  
GIFgif hfp_35.gif r1 manage 9.2 K 2010-05-05 - 18:17 AndreiKrokhotin  
GIFgif hfp_35w.gif r1 manage 13.3 K 2010-07-07 - 11:04 AndreiKrokhotin  
GIFgif hfp_36.gif r1 manage 10.0 K 2010-05-05 - 18:17 AndreiKrokhotin  
GIFgif hfp_36w.gif r1 manage 13.7 K 2010-07-07 - 11:04 AndreiKrokhotin  
GIFgif hfp_37.gif r1 manage 9.7 K 2010-05-05 - 18:17 AndreiKrokhotin  
GIFgif hfp_37w.gif r1 manage 14.0 K 2010-07-07 - 11:05 AndreiKrokhotin  
GIFgif hfp_38.gif r1 manage 10.0 K 2010-05-05 - 18:18 AndreiKrokhotin  
GIFgif hfp_38w.gif r1 manage 14.6 K 2010-07-07 - 11:05 AndreiKrokhotin  
GIFgif hfp_39.gif r1 manage 10.2 K 2010-05-05 - 18:18 AndreiKrokhotin  
GIFgif hfp_39w.gif r1 manage 14.2 K 2010-07-07 - 11:05 AndreiKrokhotin  
GIFgif hfp_40.gif r1 manage 11.4 K 2010-05-05 - 18:18 AndreiKrokhotin  
GIFgif hfp_40w.gif r1 manage 16.0 K 2010-07-07 - 11:05 AndreiKrokhotin  
GIFgif hfp_41.gif r1 manage 11.6 K 2010-05-05 - 18:19 AndreiKrokhotin  
GIFgif hfp_41w.gif r1 manage 15.7 K 2010-07-07 - 11:05 AndreiKrokhotin  
GIFgif pt1.gif r1 manage 9.1 K 2010-04-29 - 11:57 AndreiKrokhotin  
GIFgif pt_15_eta.gif r1 manage 9.8 K 2010-04-29 - 11:58 AndreiKrokhotin  
GIFgif pt_20_eta.gif r1 manage 9.6 K 2010-04-29 - 11:58 AndreiKrokhotin  
GIFgif pt_third.gif r1 manage 9.2 K 2010-04-29 - 11:58 AndreiKrokhotin  
Edit | Attach | Watch | Print version | History: r9 < r8 < r7 < r6 < r5 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r9 - 2010-07-07 - AndreiKrokhotin
 
    • 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