HGGfitter Tutorials

Introduction

The HGGfitter package is a powerful tool that contains various scripts to perform statistical analysis. In the following two topics will be discussed: extraction of the Signal Model and statistical analysis of data using a complete, signal+background, model.

In the tutorials below, it is assumed that you have downloaded and compiled Hfitter and HGGfitter, following the intsructions given here.

Signal Parametrisation

This tutorial is based on the examples given in folders named MC165501 either in HGGfitter/Tools/SignalParametrizationRun2/ or in HGGfitter/datacards/. Do not modify any of the folders in the trunk, so that this tutorial remains valid. Instead, create your own copy.

Obtain detector resolution = Narrow Width Approximation (NWA).

In following, an example to extract the signal parametrisation from nominal samples is given. In Step_6, SP for systematics variation are described. In Step_7 clarifications about other production modes are given.

Step 1. Create HFitter suitable files

1.1 Create a folder in HGGfitter/Tools/SignalParametrizationRun2/MCXXXX with a meaningful name related to the type of MC data you want to extract the parametrisation from. Note that the name must start with "MC", followed by digits!

1.2 Convert MxAODs to Hfitter format files with macro HGGfitter/Tools/CreateSamples/Inclusive/HftSamp_MGScalar.C. Edit the macro:

  • Update the input file path.
  • Change the output file path (now it is /afs/cern.ch/work/k/kgrevtso/public/HighLowMassHgg_files), so that it points to the MCXXXX directory you have created at step 1.1.
  • Change the root of the output file name in out_name_fin to something meaningful (now it is mc15_13TeV_MG).

1.3 Run loop.sh in HGGfitter/Tools/CreateSamples/Inclusive/loop.sh over eos files to create Hfitter format files.

  • Change the eos path to be the same as in HftSamp_MGScalar.C)
  • Run the script: source loop.sh
  • Output files are in: HGGfitter/Tools/SignalParametrizationRun2/MCXXXX/mc15_13TeV_MGYYY.root
where YYY available mass points.

Step 2. Prepare structure for signal model

2.1 In HGGfitter/datacards/ choose a folder myAnalysisVersion corresponding to your analysis version (LowHigMassRun2, LowMassRun2,...).

  • If it does not exist yet, create a folder SP.
  • In SP, create a folder called MCXXXX as in step 1.1.
  • In MCXXXX create the following subfolders:
           MCXXXX/output
           MCXXXX/plots
           MCXXXX/Systematics/output           
           MCXXXX/Systematics/plots
  • For further iterations, clean the previous results in the folders above.

Step 3. Create combined multiple file sample

3.1 Copy the files contained in HGGfitter/Tools/SignalParametrizationRun2/MC165501 into HGGfitter/Tools/SignalParametrizationRun2/MCfiles

3.2 Edit the macro mergeMC_sep.C and update (if necessary) the list of mass points according to the list of input files created in Step 1 and stored in Tools/SignalParametrizationRun2/MCfiles:

  • the production mode list itself: massPoints.push_back(XXX);
  • the category name list: output.AddDouble("mgg_mXXX");
  • the filling of the categories:
    output.Double("mgg_mXXX")->SetValue(i == 0 ? input.Float(mymass)->GetValue() : XXX);

Also:

  • update the path to input files BaseFileName.
  • Change the root of the output file name in outputFileName to something meaningful (now it is gen15_13TeV_).

3.3 There are a couple of other files that might need to be modified:

  • HGGfitter/Tools/SignalParametrizationRun2/MCfiles/MGMC15.txt contains the list of mass points. Change its name according to the the samples you are running on: "prod_mode"+"mc_type".txt. Here "prod_mode" is =MG and "mc_type" is MC15.
  • HGGfitter/datacards/myAnalysisVersion/SP/MCfiles/multiple_input_MG.dat - here need to update points in category mHCat and category list twice: in definition and in range. Make diff multiple_input_MG.dat multiple_input_MG_change.dat to get a feeling of difference.

3.4 Create a folder merged in HGGfitter/Tools/SignalParametrizationRun2/MCfiles.

3.5 Run the macro from HGGfitter/run (specifying options if systematics, see step 6): root -l '../Tools/SignalParametrizationRun2/MCfiles/mergeMC_sep.C'. This will create one mc file for the multiple fit (category structure) in: HGGfitter/Tools/SignalParametrizationRun2/MCfiles/merged.

Step 4. Run fits to obtain signal parameterization

There are two main parts: multiple fit and single point fits. The multiple fit provides coefficients of the functions which are used to describe each of the 6 parameters of the DSCB: delta mX (pol2), sigmaCB (pol1), alphaLow/High (pol2), nLow/High (const). The single fit provides the values of the 6 parameters of the DSCB at each mass point, independently.

-- LydiaRoos - 2018-05-02

Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r3 - 2018-05-04 - LydiaRoos
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Sandbox 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