create_dataset.yaml

defaults:
  - data@_global_: features
  - data@gen_cfg: gen_config
  - _self_

verbose: False

# inputs 
tree_name: taus
step_size: "1 GB"
input_data: # "sample_alias": [list of strings with tau types to be selected]
  train:
    recompute_tau_type: False
    add_columns: null
    files:
     - "../data/ShuffleMergeSpectral_trainingSamples-2_rerun/ShuffleMergeSpectral_*.root"
    #  - "root://eosuser.cern.ch///eos/cms/store/group/phys_tau/TauML/prod_2018_v2/ShuffleMergeSpectral_trainingSamples-2_rerun/" # for xrootd, directory to read all ROOT files from
  
  val:
    recompute_tau_type: False
    add_columns: null
    files: 
      - "../data/ShuffleMergeSpectral_trainingSamples-2_rerun/ShuffleMergeSpectral_450.root"
  
  # ### for evaluation
  # test:
  #   recompute_tau_type: True
  #   add_columns:  # will return them in this order; set to null to skip returning
  #     - tau_byDeepTau2017v2p1VSeraw
  #     - tau_byDeepTau2017v2p1VSmuraw
  #     - tau_byDeepTau2017v2p1VSjetraw
  #     - tau_pt
  #     - tau_eta
  #     - tau_decayMode
  #     - tau_dz
  #   files:
  #     "../data/GluGluHToTauTau_M125/eventTuple_*": ["tau"]

# outputs
dataset_name: ???
path_to_dataset: ??? # directory to store the datasets