Skip to content

CPAlgorithms: add new algorithm for fake lepton estimation with the asymptotic Matrix Method

Baptiste Ravina requested to merge ravinab/athena:FakeBkgCalculator into main

This is migrated from TopCPToolkit; credits to @nbruscin.

Adding a new algorithm ("FakeBkgCalculator" in YAML/ConfigFactory notation) to run the FakeBkgTools from IFF, see the documentation.

The code is pretty minimal for now (but will be expended on by IFF experts):

  • it can set up only the asymptotic Matrix Method, more techniques could be added in the future;
  • it allows to set up multiple estimations (each one has 1 configuration ROOT/XML file and its loose --> tight definitions);
  • it decorates each event with MM weights named weight_fake_*.

Below an example for the FakeBkgCalculator YAML block that calculates two MM weights, with different process definitions and event selections.

FakeBkgCalculator:
  - algoName: 'Test1'
    electrons: 'AnaElectrons.loose'
    electronsTarget: 'tight'
    muons: 'AnaMuons.loose'
    muonsTarget: 'tight'
    eventPreselection: 'pass_ejets_%SYS%,as_char || pass_mujets_%SYS%,as_char'
    fakeTool: 'CP::AsymptMatrixTool'
    config: 'dev/AnalysisTop/FakeBkgToolsData/Efficiency2D_Data.xml'
    definition: '1T'
    process: '1F[T]'
  - algoName: 'Test2'
    electrons: 'AnaElectrons.loose'
    electronsTarget: 'tight'
    muons: 'AnaMuons.loose'
    muonsTarget: 'tight'
    eventPreselection: ''
    fakeTool: 'CP::AsymptMatrixTool'
    config: 'dev/AnalysisTop/FakeBkgToolsData/Efficiency2D_Data.xml'
    definition: '2T'
    process: '2F[T]'

Merge request reports