#alp_analysis

Repo to create skimmed ntuples from the output of ALPHA framework. Main features: trigger, kinematics, btagging selection; weights computation/handling; plot creation; higher level variables creation.

CODE STRUCTURE

alp_analysis inherits classes from ALPHA framework. The latter has to be installed before, together with a CMSSW release. Execution from python scripts which call operators defined in interface folder.

TO DEBUG:

Look at issue on GitHub and use this chat for discuss about the code:

TO INSTALL:

• Prerequisites git account, git environment set, CERN account.

• Git instructions 0. In your working area, first set up the CMSSW release: ``` cmsrel CMSSW_8_0_12

    cd CMSSW_8_0_12/src/
  
    cmsenv
  
    git cms-init
    ```
  

  1. Install packages needed by ALPHA: see README from ALPHA repo [https://github.com/cms-hh-pd/ALPHA#alpha]

  2. Clone ALPHA and setup it: see README from ALPHA repo [https://github.com/cms-hh-pd/ALPHA#alpha] NOTE: clone it from cms-hh-pd repo: bash git clone [email protected]:cms-hh-pd/ALPHA.git

  3. Clone the alp_analysis git repository:

     cd $CMSSW_BASE/src/Analysis
     
     git clone [email protected]:cms-hh-pd/alp_analysis.git

TO RUN:

Compile the code:

cd $CMSSW_BASE/src/
scram b -j 8

and run it:

cd $CMSSW_BASE/src/Analysis/alp_analysis
python scripts/Selector.py

If you do not have ALPHA compiled do:

cd Analysis/alp_analysis/src/ root -l

.L ../src/alp_objects.h++

SCRIPT DESCRIPTION:

• BaselineSelector:

  to apply object/event selection. it has option to get antitag CR and to run over mixed events. event tree and hemisphere tree are saved by default

  python scripts/BaselineSelector.py -s signals -o def_cmva -i v2_20170606 (-a -m)

• MixingSelector:

  to create mixed events (baseline selector needs to be run after, to get standard ntuple structure). !!remember to use -a option to run in antitag CR!!

  python scripts/MixingSelector.py -s Data -i def_cmva --comb appl (-a)

• ClfSelector:

  to select event in a specif range of the classifier. !!additional sample needed with classifier, run columns2tree from hh2bbbb_limit before!!

  python scripts/ClfSelector.py -s Data_ -i 20171205 -o test

• TrgEffStudies:

  to get event selection for trigger efficiency studies.

• MCTruthSelector.py:

  to select only jets matched with gen

• TriggerSelector:

  to apply trigger and get plots.

• NoCutSelector.py:

  simple processing of the input data. No selection/object creation applied.

• comp_boost:

  to count fraction of common events between two samples (for comparison with boosted analysis)

• tkTDRSelector:

  to apply default selection to ntuple with different pu (for tracker tdr studies)

• ttHSelector:

  to apply a baseline selection for ttH searches. output is a 'plain tree' with no structures.

• AddPlots:

  to add plots (jets, dijets, dihiggs) on top of the input file

• GetPlainTree:

  to get simple tree with one variable per branch. readable with a simple ROOT macro - no classes needed. Useful for students.

• ReWeighting.py:

  to re-weight input file not maintained

• ToyDatasetCreator:

  to create a pseudo-data with signal injected. not maintained

• HistFromPangea.py:

  to get histogram comparison - pangea vs different benchmarks not maintained

• DiHiggsSelector:

  to test different higgs mass selection not maintained

• ComposableSelector:

  bare example of how to use operators.

INPUT FOLDERS FROM ALPHA_NTUPLES:

• v2_20170222-trg -> TT

• v2_20170405 -> ttHbb

• v2_20170222 -> Data

• v2_20170606 -> signals