RIVET Introduction By Mehar Ali Shah PhD Student National Centre for Physics Quaid-I-Azam University Pakistan 1.

RIVET IntroductionBy

Mehar Ali Shah

PhD Student

National Centre for Physics

Quaid-I-Azam University Pakistan

1

The Rivet project (Robust Independent Validation of Experiment and Theory) is a generator-agnostic toolkit for validation of Monte Carlo event generators.

RIVET emphasizes the avoidance of hard-coding reference data or histogram binnings in the analysis code. Rivet provides functions to extract this information from bundled data files, meaning that it is much easier to keep reference and generated data synchronised.

RIVET

2

Rivet is a library and set of programs which produce simulated distributions which can be directly compared to measured data for MC validation and tuning studies.

It can also be used without reference data to compare two or more generators to each other for regression testing or tune comparison.

The Rivet library contains the tools needed to calculate physical observables from HepMC files or objects, a large set of important experimental analyses, and histogramming tools for data

comparison and presentation.

RIVET: Software structure

3

RIVET and CMSSW

RIVET is interfaced to CMSSW. This means that through CMSSW we can run rivet standard analyses, implement new ones of our own.

We can run one or many Rivet analysis with an EDAnalyzer wrapper, that allows us to make full

advantage of the CMSSW framework. In other words we can run the generator and the Rivet analyses in a single CMSSW run, for any of the generators interfaced to CMSSW.

( It is used by both phenomenologists/generator developers and by the LHC experiments for generator tuning, analysis prototyping, and MC-data comparisons. )

4

5

SETUP

Rivet docs: online at http://projects.hepforge.org/rivet/ – PDF manual, HTML list of existing analyses.

Log in to lxplus.cern.ch or use the Rivet VM. On lxplus, if you’re not in a bash shell (echo $SHELL), then run bash to make life more pleasant: the Rivet toolkit provides contextual command line completion with bash.

On lxplus: source ~abuckley/public/setupRivetProf.sh

Test commands: rivet --help

• Separation of generator steering and event analysis• Automated binning of MC histograms from data .• A C++-replacement for HZTOOL• Data export direct from HepData • Generator-independent implementations of standard

observable calculators• Almost 100 important analyses from LEP, HERA,

RHIC, KEK-B, Tevatron Runs I & II, LHC and more• User-implemented analyses can be used as

"plugins" without modifying the Rivet code.

Some features of Rivet:

6

RIVET Projections and Analyses

If one want to use Rivet for validation and data/MC comparisons the most important things to understand are 

RivetProjections which calculate event properties in an efficient and robust way.

RivetAnalysis which book histograms from HepData files, apply projections to a HepMC event, and fill new histograms.

7

Projections

A projection is an object which calculates a property or properties from an Event. Such properties can be

◮ A number (such as Q2)

◮ A set of final-state Rivet::Particles

◮ Jets

They are just observable calculators: given an Event object, they project out physical observables.

They also automatically cache themselves, to avoid recomputation: this leads to the most unintuitive code structures in Rivet.

8

• FinalState - projection finds all final state particles in

a given range of eta and a given pT cutoff.

• Subclasses:

– ChargedFinalState and NeutralFinalState have the predictable effect!

– IdentifiedFinalState can be used to find particular particle species.

– VetoedFinalState finds particles other than specified.

– VisibleFinalState excludes invisible particles like Neutrinos etc.

Examples of projections

9

Histogram autobooking

This is a means for getting your Rivet histograms binned with the same bin edges as used in the experimental data that you’ll be comparing to.

To use autobooking, just call the booking helper function with only the histogram name (check that this matches the name in the reference .aida file), e.g.

_hist1 = bookHistogram1D("d01-x01-y01")

10

Sorry experimentalists no ROOT:

No Worry !!

You can convert the Rivet output with the aida2root script. . .

The plots are written out as dataPointSet objects in AIDA XML format.

Aida

11

Central Track Multiplicities and Forward Energy Flow in W/Z Boson Events

Summer Student

2011 at CERNWith

Forward Physics Group

1212

My Experience with RIVET

Introduction

Study the energy flow in the forward detectors, and the correlations of the forward energy flow with the trackmultiplicity in the central detector using events with centrally produced W and Z bosons, identified with their leptonic decays

(This analysis of W and Z events shows that the underlying event structure and the observed correlations are not described by the

available non-diffractive soft hadron production models)(The analysis of W and Z events shows that the underlying event structure and the observed correlations are not described by the

available non-diffractive soft hadron production models)

1313

Introduction

The analysis of the underlying event structure and energy flow correlation in processes with

colorless final states, such as

pp → W(Z)X →ℓᶹ(ℓℓ)X

can provide new insights into so far unexplored aspects of multi-parton interactions. W(Z) events with high track multiplicities in the central detector indicate a possibly large contribution from multi-parton interactions.

1414

Reference Paper and Data Samples

“Forward Energy Flow, Central Track Multiplicities and Large Rapidity Gaps in W and Z Boson Events at 7 TeV pp Collisions”

(CMS PAS FWD-10-008)

7 TeV pp collision data sample of about 36 pb-1, recorded during the 2010 LHC operation is used for comparison.

Pythia 6 z2 tune (describe the pp -> X interactions at high energies by a perturbative hard interaction of two partons within the protons combined with a non-perturbative multi-parton interaction component, existing in many variations, between the proton remnants)

1515

Event Selection

measured W/Z with

lepton in |η| < 1.4

pT electron > 25 GeV Missing Et > 30 GeV

Transverse mass of lepton and neutrino > 60 GeV

1616

Peaked at

pT = mW/2 = 40 GeV

17

Aida to Root

17

Example plots from RIVET

W → ℓᶹ -2.5 < η < 2.5 pT > 1 GeV

Central track multiplicity

1818

W → ℓᶹ ( -2.5 < η < 2.5) 3 < |η| < 4.9 pT > 1 GeV

20 < ∑HF- < 100 200 < ∑HF- < 400 ∑HF- > 500

Normalized Track Multiplicity Distributions

1919

W →eX 3 < |η| < 4.9

W Energy flow measurement

2020

W →eX 3 < |η| < 4.9

E(each particle) > 0, 2, 4, 6 (GeV)

W Energy flow measurement 0, 2, 4 ,6 GeV

2121

Correction factors from detector to hadron level

2222

Forward energy flow in the acceptance of HF

(CMS AN AN-10-331)

Correction factors from detector to hadron level

CFs 3 < η < 3.5 3.5 < η < 4.5 4.5 < η < 4.9

C1: dijets 1.4 1.25 2.2

C2: Minbias 1.9 1.65 2.5

C3 = 1.5C1 2.10 1.87 3.30

C4 = 2C1 2.8 2.50 4.40

2323

All Correction Factors with 0 (GeV) Cut

2424

All Correction Factors with 2 (GeV) cut

2525

All Correction Factors with 4 (GeV) cut

2626

All Correction Factors with 6 (GeV) cut

2727

HF Energy measurement with all Cfs

2828

Energy Distributions in HF+ with cuts in HF-

20 < ∑HF- < 100

2929

20 < ∑HF- < 100 200 < ∑HF- < 400

Energy Distributions in HF+

30

4(GeV)

6GeV

30

RIVET Beauty

31

Future Plan

RIVET people are much cooperative and helpful. They respond immediatly whenever help needed.

I am going to start my PhD research. I would love to continue with RIVET.

Now I am looking for an interesting topic. One option could be to choose a top paper and get the same results using RIVET and compare it with different generators. There could be performed so many other interesting studies by making use of RIVET.

Conclusion

Lots of standard analyses built in, including key ones for pQCD and MPI model testing.

New analyses can be picked up at runtime: nice API with lots of tools to make this as simple and pleasant as we can.

Computations automatically cached. Histograms automatically synchronised. Satisfaction automatically guaranteed

Simply RIVET makes life very easy !32

Thanks

33INSC 201233