How to configure the default database

Description

The first part of the database allows you to activate the analyses steps that perform conversion, skimming for data and monte carlo. Through the option "merging" and "mergingperiods" you can also create a merged sample of data and MC that includes a fraction of the different sub periods that will be used for the ML optimization.

case: LcpK0s_multiplicity_test
download:
  alice:
    activate: false
conversion:
  mc: 
    activate: true 
  data:
    activate: true
skimming: 
  mc:
    activate: true
  data:
    activate: true
merging: 
  mc:
    activate: true
  data:
    activate: true
mergingperiods: 
  mc:
    activate: true
  data:
    activate: true

In the second part of the database you can activate the ML optimization routine through the main activator "activate". If the main activation option is on, a list of other options can be enabled. This list includes: study of the training variables, training, testing, validation of the ML optimization, significance optimization and a lot more.

ml_study:   
  activate: true
  doscancuts: false
  applytodatamc: false
  doroc: true
  doboundary: false
  docorrelation: false
  docrossvalidation: false
  dogridsearch: false
  doimportance: false
  dolearningcurve: true
  dopca: false
  dosignifopt: false
  doefficiency: false
  dotesting: false
  dotraining: true

In the second last part of the database, you can activate the last step of the analysis that includes applying the model trained in the previous step to the entire dataset, apply an optimal probability cut to select the final sample of candidates. In the very last step invariant mass spectra and MC efficiency can be obtained.

analysis:
  data:
    doapply: true
    domergeapply: true
    domergeapplyperiods: true
    histomass: true
  mc:
    doapply: true
    domergeapply: true
    domergeapplyperiods: true
    histomass: true
    efficiency: true