derive_wp.yaml

wp_maker:
  _target_: derive_wp.WPMaker

  wp_definitions:
    e:
      VVTight: 0.6
      VTight: 0.7
      Tight: 0.8
      Medium: 0.9 
      Loose: 0.95
      VLoose: 0.98
      VVLoose: 0.99
      VVVLoose: 0.995
    mu:
      Tight: 0.995
      Medium: 0.998
      Loose: 0.999
      VLoose: 0.9995
    jet:
      VVTight: 0.4
      VTight: 0.5 
      Tight: 0.6
      Medium: 0.7
      Loose: 0.8
      VLoose: 0.9
      VVLoose: 0.95
      VVVLoose: 0.98
    
  tpr_step: 0.0001 # will evenly sample grid of values from 0 to 1 with step=tpr_step
  require_wp_vs_others: True # if in computation of WP thresholds for a given `vs_type` taus should pass WPs from `WPs_to_require` against remaining `vs_types` 
  WPs_to_require:
    e: VVVLoose
    mu: VLoose
    jet: VVVLoose
  
  # convergence params
  epsilon: 1e-5
  n_iterations: 100
  verbose: False
  rounding: 4

create_df:
  _target_: __main__.create_df

  # path to prediction files
  path_to_mlflow: ???
  experiment_id: ???
  run_id: ???
  path_to_preds: "${create_df.path_to_mlflow}/${create_df.experiment_id}/${create_df.run_id}/artifacts/predictions/"
  tau_type_to_select: tau
  pred_samples: # samples in `path_to_preds` to take taus from 
    2016APVUL_GluGluHToTauTau_M125: 
      filename_pattern: 'eventTuple_*.h5' # single files, list of files and "*" are supported 
      sample_lumi: 19.52
      reweight_to_lumi: 137.64 # total luminosity or `null` to skip reweighting
    2016UL_GluGluHToTauTau_M125: 
      filename_pattern: 'eventTuple_*.h5'
      sample_lumi: 16.81
      reweight_to_lumi: 137.64
    2017UL_GluGluHToTauTau_M125: 
      filename_pattern: 'eventTuple_*.h5'
      sample_lumi: 41.48
      reweight_to_lumi: 137.64
    2018UL_GluGluHToTauTau_M125: 
      filename_pattern: 'eventTuple_*.h5'
      sample_lumi: 59.83
      reweight_to_lumi: 137.64

  # naming conventions of prediction files
  pred_group_name: predictions
  pred_column_prefix: node_ # node_ // pred_
  target_group_name: targets # targets // labels
  target_column_prefix: node_ # node_ // label_

  # add columns to apply selection on
  # paths to input ROOT files will be retrieved from corresponding pred_input_filemap.json
  add_columns_from: 'inputs' # either "predictions" or "inputs"
  add_columns: [ 'tau_pt', 'tau_eta', 'tau_dz', 'tau_decayMode' ]
  selection: '(tau_pt>=30) and (tau_pt<80) and (abs(tau_eta) < 2.3) and (abs(tau_dz) < 0.2) and not (tau_decayMode in [5,6])'
  group_or_tree_name: taus # add_columns // taus -- either group name (add_columns_from=prediction) or tree name (add_columns_from=inputs)