BootstrapGeneratorConfig.py

Go to the documentation of this file.

# Copyright (C) 2002-2023 CERN for the benefit of the ATLAS collaboration
 
# AnaAlgorithm import(s):
from AnalysisAlgorithmsConfig.ConfigBlock import ConfigBlock
from AnalysisAlgorithmsConfig.ConfigAccumulator import DataType
 
class BootstrapGeneratorConfig(ConfigBlock):
    '''ConfigBlock for the bootstrap generator'''
 
    def __init__(self):
        super(BootstrapGeneratorConfig, self).__init__()
        self.addOption ('nReplicas', 1000, type=int,
            info="the number (int) of bootstrap replicas to generate. "
            "The default is 1000.")
        self.addOption ('decoration', None, type=str,
            info="the name of the output vector branch containing the "
            "bootstrapped weights. The default is bootstrapWeights.")
        self.addOption ('runOnMC', False, type=bool,
            info="toggle to force running on MC samples. The default is "
            "False, i.e. run only on data.")
    
    def makeAlgs(self, config):
        if config.dataType() is not DataType.Data and not self.runOnMC:
            print("Skipping the configuration of CP::BootstrapGeneratorAlg since we are not running on data. "
                  "Set the option 'runOnMC' to True if you want to force the bootstrapping of MC too.")
            return
        
        alg = config.createAlgorithm( 'CP::BootstrapGeneratorAlg', 'BootstrapGenerator')
        alg.nReplicas = self.nReplicas
        alg.isData = config.dataType() is DataType.Data
        if self.decoration:
            alg.decorationName = self.decoration
        else:
            alg.decorationName = "bootstrapWeights_%SYS%"
        
        config.addOutputVar ('EventInfo', alg.decorationName, alg.decorationName.split("_%SYS%")[0], noSys=True)
 
        return
 
def makeBootstrapGeneratorConfig(seq,
                                 nReplicas = None,
                                 decoration = None,
                                 runOnMC = None):
    """
    Setup a simple bootstrapping algorithm
 
    Keyword arguments:
      nReplicas -- the number of bootstrap replicas to generate
      decoration -- the name of the output vector branch containing the bootstrapped weights
      runOnMC -- toggle to force running on MC samples (default: only data)
    """
 
    config = BootstrapGeneratorConfig()
    config.setOptionValue ('nReplicas', nReplicas)
    config.setOptionValue ('decoration', decoration)
    config.setOptionValue ('runOnMC', runOnMC)
    seq.append (config)