-- BenjaminBordyTannenwald - 14-Jul-2011

KUCMSGroupStuff/BAnalysis

• Member list (comma-separated list):
  • Set GROUP = BenjaminBordyTannenwald, AliceBean, GemmaMariaTinti, JenniferSibille
• Persons/group who can change the list:
  • Set ALLOWTOPICCHANGE = BenjaminBordyTannenwald, AliceBean, GemmaMariaTinti, JenniferSibille

Table of Contents

Analysis Summary

This site presents the result of a Monte Carlo truth-level analysis of events where b→μD0X; D0→K&#960. A final copy of the paper written to present the analysis can be found here, and a copy of the most recent uploaded code can be found here (Last Updated: 14/7/11). The Monte Carlo sample comes from /ppMuX/Spring10-START3X_V26_S09-v1/GEN-SIM-RECO.

The algorithm developed for analysis searches an event for μ, and once a μ is found, it requires the μ to have Pt > 5 GeV and an absolute value of eta to be <= 2.4 . Once the μ passes these kinematic cuts, events are further pared down by requiring the μ to come from a direct b decay chain. After direct b μ are identified, events where a D0 shares the same b hadron parent as the μ are selected to pass further cuts. If the D0 passes kinematic cuts of Pt > 2.5 GeV and |eta| <= 2.4, decays into a K and &#960, and the K and &#960 pass final kinematic cuts of Pt > 0.75 GeV and |eta| <= 2.4, the event is classified as signal. A variety of reconstruction efficiencies are calculated in the paper, and insights into the resolution capabilities of the CMS detector are explored. The below plots present a variety of information extracted from the data in the analysis. For explicit conclusions and results, see the paper linked to above.

Plots

fPtHat

These plots show fPtHat in events where μ Pt is greater than 5/10/15 GeV

fPtHat in events where the highest Pt μ has Pt >5 GeV

fPtHat in events where the highest Pt μ has Pt >10 GeV

fPtHat in events where the highest Pt μ has Pt >15 GeV

2-D Plot comparing fPtHat on the x-axis with the Pt of the highest Pt μ (y-axis) in all μ events. The Monte Carlo generator required each event to have a μ with Pt > 5 GeV.

2-D Plot comparing fPtHat on the x-axis with the Pt of the 2nd highest Pt μ (y-axis) in all muon events. There was no explicit filter on the Pt of any μ in an event once a single μ was found to have Pt > 5 GeV.

2nd Trigger μ

These plots look only at the 2nd highest Pt μ in an event given that the highest Pt μ passes acceptance cuts ( Pt >5 GeV, |Eta| <= 2.4 ). Counts for these secondary μ are shown by Pt. Plots of all 2nd μ Pt is shown before and after the same acceptance cuts placed on the primary μ. Plots are shown from μ truth-matched to direct b parents with the full Pt spectra and Pt spectra after acceptance cuts.

Full μ Pt spectra of 2nd highest Pt μ in events where the highest Pt μ passes acceptance cuts.

Full μ Pt spectra of 2nd highest Pt μ in events where the highest Pt μ passes acceptance cuts AND the 2nd μ is truth-matched to a direct b parent.

Pt spectra of 2nd highest Pt μ in events where both the highest Pt μ and the second highest Pt μ pass acceptance cuts (Pt >5 GeV, |Eta| <= 2.4).

Pt spectra of 2nd highest Pt μ in events where both the highest Pt μ passes acceptance cuts, and the second highest Pt μ both passes acceptance cuts and is truth matched to a direct b parent.

Δ R/Eta Signal & Background

These plots show scaled, overlaid histograms comparing Signal (direct b μ) vs. Background (direct c μ, b -> c μ) events by the delta R of b/c quarks with their daughter meson/baryons in direct b events. Numbers are not shown for each histogram, but the scaling allows for a determination of appropriate cuts on delta R to discriminate signal from background in collision data.

Displays the delta R between an acceptance μ and 1) a direct b Meson/Baryon parent [white], 2) a direct c Meson/Baryon parent [green], and 3) a direct b->c parent [blue].

Displays the delta R between the meson/baryon parent of acceptance direct b/direct c μ and 1) its mother b quark [white] and 2) its mother c quark [blue].

Events by Bin

μ Acceptance: μ Pt >5 GeV, μ |Eta| <=2.4 D0 Acceptance: D0 Pt >2 GeV, K/pi Pt >.5 GeV, D0/K/pi |Eta| <=2.4

All entries denote # of events: μ

μ Pt [GeV]	All μ	b→μX μ	b→μX w/ μ Acpt.	b→μD0X w/ μ Acpt.	b→μD0X; D0→Kπ w/ μ Acpt.	b→μD0X; D0→Kπ w/ μ & D0 Acpt.
0-5	3.38631e+07	1.46498e+06	0	0	0	0
5-6	1.99876e+06	1.75463e+06	1.7516e+06	1.02907e+06	38252	17111
6-7	1.15366e+06	1.03084e+06	1.02896e+06	603037	22061	11413
7-8	650061	586161	585135	344155	12767	7278
8-9	382586	348020	347311	204524	7608	4762
9-10	234733	214526	214049	126634	4772	3117
10-15	390560	359716	358825	213255	7857	5786
15-20	69445	64676	64510	38381	1438	1204
>20	29081	27241	27160	16218	610	556

b Meson/Baryon Pt [GeV]	All b Mesons/Baryons	b→μX	b→μX w/ μ Acpt.	b→μD0X w/ μ Acpt.	b→μD0X; D0→Kπ w/ μ Acpt.	b→μD0X; D0→Kπ w/ μ & D0 Acpt.
0-2	1.18774e+06	200910	2096	0	0	0
2-4	2.12728e+06	377532	12895	0	0	0
4-6	1.8363e+06	331650	37891	11853	454	0
6-8	1.88066e+06	578736	382718	217679	8128	646
8-10	2.12775e+06	885904	755697	444609	16617	4901
10-12	1.94187e+06	874000	785383	466437	17182	7992
12-14	1.54719e+06	706015	645889	383936	14223	8316
14-16	1.1529e+06	521571	480136	287020	10612	6924
16-18	843933	374466	345676	206501	7571	5383
18-20	615364	266175	246483	147479	5444	4175
>20	1.88772e+06	733829	682678	409760	15134	12890

b→μD0X; D0→Kπ

The following plots display the majority of the efficiency and resolution results that formed the focus of the analysis. All data shown below comes from events designated as signal by the algorithm described above. Events in which b→μD0X; D0→Kπ are broken down and information extracted to create the following plots. Some plots created in the unfolding section of Lukas' dissertation work are not shown below, but links are given for their access.

Reconstruction Efficiencies (D0, μ, K, π)

Reconstruction efficiencies for D0, μ, K, and π in signal events are given below in bins of 0.5 GeV from 0-15 GeV as a function of Pt. Efficiency is found by dividing the number of reconstructed particle in a given Pt bin by the number of total particles in the same Pt bin without regard to successful reconstruction.

Shown here is the μ reconstruction efficiency as a function of Pt. The efficiency starts at 5 GeV since μ in signal events must pass acceptance cuts. The efficiency is calculated by dividing bin by bin (1 bin = 0.5 GeV) the truth Pt of properly reconstructed μ candidates in signal events by the truth Pt of all μ candidates in signal events. Out of 47615 μ candidates found at the truth level, only 47126 events are successfully reconstructed. Binned reconstruction error is also included in the plot. The truth Pt distribution can be found here and the reconstructed distribution can be found here

Shown here is the D0 reconstruction efficiency as a function of Pt. The efficiency is calculated by dividing bin by bin (1 bin = 0.5 GeV) the truth Pt of properly reconstructed D0 candidates in signal events by the truth Pt of all D0 candidates in signal events. Out of 95365 D0 candidates found at the truth level, only 32225 events are successfully reconstructed. Binned reconstruction error is also included in the plot. The truth Pt distribution can be found here and the reconstructed distribution can be found here

Shown here is the π reconstruction efficiency as a function of Pt. The efficiency is calculated by dividing bin by bin (1 bin = 0.5 GeV) the truth Pt of properly reconstructed π candidates in signal events by the truth Pt of all π candidates in signal events. Out of 47615 π candidates found at the truth level, only 45094 events are successfully reconstructed. Binned reconstruction error is also included in the plot. The truth Pt distribution can be found here and the reconstructed distribution can be found here

Shown here is the K reconstruction efficiency as a function of Pt. The efficiency is calculated by dividing bin by bin (1 bin = 0.5 GeV) the truth Pt of properly reconstructed K candidates in signal events by the truth Pt of all K candidates in signal events. Out of 47615 K candidates found at the truth level, only 41903 events are successfully reconstructed. Binned reconstruction error is also included in the plot. The truth Pt distribution can be found here and the reconstructed distribution can be found here

The reconstruction efficiencies of the μ, π, and K are nearly 100% for Pt values up to 15 GeV. The limiting factor in reconstruction efficiency in the analysis is the reconstruction of the D0 candidates. While the D0 reconstruction increases with Pt, it only achieves around 65% efficiency at the 15 GeV limit.

Reconstructed and Invariant Masses

Invariant masses are shown for the D0 in signal events where D0→ Kπ and for the b hadron when b→μD0X ; D0→Kπ. Results are shown for the invariant mass calculation from truth-level Pt and reconstructed Pt for events identified as signal.

This plot shows the truth Pt D0 invariant mass from events where D0→ Kπ. Because the both the K & π are directly accessible in the lab, there is no variance in the truth-level invariant mass.

This plot shows the reconstructed Pt D0 invariant mass from events where D0→ Kπ. The reconstructed invariant mass follows a Gaussian distribution around the accepted D0 mass of 1.865 GeV due to experimental fluctuation in the reconstruction process. The results of a Gaussian fit to the data is shown overlaid on the plot.

This plot shows the truth Pt b hadron invariant mass in signal events where b→μD0X ;D0→ Kπ. The variance in the generator level invariant mass for the b hadron is a result of the missing Pt carried away by the neutrino in these events.

This plot shows the truth Pt b hadron invariant mass in signal events where b→μD0X ;D0→ Kπ. The variance in the reconstructed invariant mass for the b hadron is again the result of the missing Pt carried away by the neutrino. The small peak at 0 is a result of improper filling of the mass of particles in the Monte Carlo generation.

ΔR Comparisons

The following plots show ΔR comparisons between the the 4 vectors of various truth-level particles. In come cases, the particles are single baryons/quarks/ μ and some are composite vectors formed from a sum of several particles. Plots used in the analysis are shown below, but additional plots recreating some of the unfolding done in Lukas' analysis were also created. Plots for the ΔR between the b muon and D0 daughter K can be found here, the ΔR distribution between the b muon and the D0 daughter π can be found here, and the ΔR between the K and π daughters of the D0 can be found here.

ΔR between the b hadron and b quark and μ in events where b→μX. ΔR is smaller, and the resolution higher, between the μ and b hadron compared to the b quark. The top statistics box describes the quark ΔR distribution.

ΔR between the b hadron and b quark and μD0 in events where b→μD0X. Again, resolution is better for the b hadron case compared to the b quark. The top statistics box describes the quark ΔR distribution.

Variance in the b→μD0X; D0→Kπ decay chain is represented by the X. One possibility is for a D* meson to decay into a π and a D0 which then decays into a K and π. By using the extra π generated in such events, it is possible to further restrict and purify signal events while ideally improving the resolution of the b hadron/quark using the extra kinematic data. Results from isolating these events and a comparison to events without the π requirement are shown below.

ΔR between the b quark/hadron and μD0π composite in events where b→μD0X and X includes a charged π. Resolution is better for the b hadron case. The top statistics box describes the quark ΔR distribution.

ΔR between the b hadron and μD0/μD0π composite. The top statistics box describes the μD0 distribution and the bottom describes the μD0π plot. While there are fewer events with a charged π, the the resolution is slightly higher.

2D Pt Unfolding

These plots follow the 2D comparisons used in the unfolding process of Lukas' dissertation. For the following plots, the x-axis shows the truth-level Pt of a μD0 composite in b→μD0X events, and the y-axis shows the truth-level Pt of the parent b hadron/ grandparent b quark in b→μD0X events. The quark distributions display much more variance in Pt possibly due to effects from the hadronization process. The plots are created once a μ has passed acceptance cuts (Pt > 5 GeV, |eta|<= 2.4) and a D0 is truth matched to the same b hadron mother as the μ.

The following three plots display the Pt comparison between the μD0 composite and the parent b hadron.

2D comparison of between the truth-level Pt of the μD0 composite and the parent b hadron in events where b→μD0X. No cuts are placed on the invariant mass of the μD0 composite.

2D comparison of between the truth-level Pt of the μD0 composite and the parent b hadron in events where b→μD0X. The invariant mass of the μD0 composite is required to be 3.0 GeV <= M(μD0) < 3.5 GeV.

2D comparison of between the truth-level Pt of the μD0 composite and the parent b hadron in events where b→μD0X. The invariant mass of the μD0 composite is required to be 4.0 GeV <= M(μD0) < 4.5 GeV.

The next three plots display the Pt comparison between the μD0 composite and the grandparent b quark.

2D comparison of between the truth-level Pt of the μD0 composite and the grandparent b quark in events where b→μD0X. No cuts are placed on the invariant mass of the μD0 composite.

2D comparison of between the truth-level Pt of the μD0 composite and the grandparent b quark in events where b→μD0X. The invariant mass of the μD0 composite is required to be 3.0 GeV <= M(μD0) < 3.5 GeV.

2D comparison of between the truth-level Pt of the μD0 composite and the grandparent b quark in events where b→μD0X. The invariant mass of the μD0 composite is required to be 4.0 GeV <= M(μD0) < 4.5 GeV.

Luminosity Summary

Summary of luminosity for different single-muon triggers in Run2010 data, calculated using lumicalc.py script.

Run2010A: 135808 <= runs < 145762 Run2010B: runs >=145762

Full data set Dec22 or Nov4 reprocessing

JSON file Cert_136033-149442_7TeV_Dec22ReReco_Collisions10_JSON.txt found here: https://cms-service-dqm.web.cern.ch/cms-service-dqm/CAF/certification/Collisions10/7TeV/Reprocessing/

Trigger	Unprescaled Luminosity (pb-1)			First run included	First prescaled run
		v3	v4
HLT_Mu5	0.2847	0.2847	0.2847	beginning	141956
HLT_Mu9	8.237	8.232	8.246	beginning	147196
HLT_Mu13_v1	9.47	9.47	9.47	147196	148819
HLT_Mu15_v1	27.9	27.9	27.9	147196	--
Total	36.145	36.133	36.178	--	--

MuonPhys data set

JSON file Cert_136033-149442_7TeV_Dec22ReReco_Collisions10_JSON_MuonPhys_v3.txt

Trigger	Unprescaled Luminosity (pb-1)	First run included	First prescaled run
HLT_Mu5	0.3004	beginning	141956
HLT_Mu9	8.636	beginning	147196
HLT_Mu13_v1	10.38	147196	148819
HLT_Mu15_v1	31.33	147196	--
Total	39.969	--	--

Review status

Reviewer/Editor and Date	Comments
BenjaminBordyTannenwald - 14 Jul 2011	Updated Code and Plots, Paper Added
JenniferSibille - 21 Feb 2011	Added lumi summary info

• GetInfoFromChain.cc: GetInfoFromChain.cc