builtin_splot.py

#!/usr/bin/env python3
 
 
 
import basf2_mva
 
if __name__ == "__main__":
    variables = ['p', 'pt', 'pz', 'phi',
                 'chiProb', 'dr', 'dz', 'dphi',
                 'daughter(0, dr)', 'daughter(1, dr)', 'daughter(0, dz)', 'daughter(1, dz)',
                 'daughter(0, dphi)', 'daughter(1, dphi)',
                 'daughter(0, chiProb)', 'daughter(1, chiProb)', 'daughter(2, chiProb)', 'daughter(2, M)',
                 'daughter(0, atcPIDBelle(3,2))', 'daughter(1, atcPIDBelle(3,2))',
                 'daughter(2, daughter(0, E))', 'daughter(2, daughter(1, E))',
                 'daughter(2, daughter(0, clusterLAT))', 'daughter(2, daughter(1, clusterLAT))',
                 'daughter(2, daughter(0, clusterHighestE))', 'daughter(2, daughter(1, clusterHighestE))',
                 'daughter(2, daughter(0, clusterNHits))', 'daughter(2, daughter(1, clusterNHits))',
                 'daughter(2, daughter(0, clusterE9E25))', 'daughter(2, daughter(1, clusterE9E25))',
                 'daughter(2, daughter(0, minC2TDist))', 'daughter(2, daughter(1, minC2TDist))',
                 # We do not use kinematic variables of the daughters
                 # 'daughter(0, p)', 'daughter(0, pz)', 'daughter(0, pt)', 'daughter(0, phi)',
                 # 'daughter(1, p)', 'daughter(1, pz)', 'daughter(1, pt)', 'daughter(1, phi)',
                 # 'daughter(2, p)', 'daughter(2, pz)', 'daughter(2, pt)', 'daughter(2, phi)',
                 # 'daughterInvM(1, 2)', 'daughterInvM(0, 1)', 'daughterInvM(0, 2)',
                 # 'daughterAngle(0, 1)', 'daughterAngle(0, 2)', 'daughterAngle(1, 2)',
                 'M',
                 ]
 
    # Perform an sPlot training
    general_options = basf2_mva.GeneralOptions()
    general_options.m_datafiles = basf2_mva.vector("train_mc.root")
    general_options.m_identifier = "MVAFull"
    general_options.m_treename = "tree"
    general_options.m_variables = basf2_mva.vector(*variables)
    general_options.m_target_variable = "isSignal"
 
    fastbdt_options = basf2_mva.FastBDTOptions()
    # SPlot is more stable if one doesn't use the randRatio
    # FastBDT has a special sPlot mode, but which isn't implemented yet in the mva package
    # fastbdt_options.m_nTrees = 100
    fastbdt_options.m_randRatio = 1.0
    basf2_mva.teacher(general_options, fastbdt_options)
 
    general_options.m_identifier = "MVAOrdinary"
    general_options.m_variables = basf2_mva.vector(*variables[:-1])
    basf2_mva.teacher(general_options, fastbdt_options)
 
    meta_options = basf2_mva.MetaOptions()
    meta_options.m_use_splot = True
    meta_options.m_splot_variable = "M"
    # SPlot training assumes that the datafile given to the general options contains only data
    # It requires an additional file with MC information from which it can extract the distribution
    # of the discriminating variable (in this case M).
    # Here we use the same file
    general_options.m_datafiles = basf2_mva.vector("train_data.root")
    meta_options.m_splot_mc_files = basf2_mva.vector("train_mc.root")
 
    # First we do an ordinary sPlot training
    general_options.m_identifier = "MVASPlot"
    meta_options.m_splot_combined = False
    meta_options.m_splot_boosted = False
    basf2_mva.teacher(general_options, fastbdt_options, meta_options)
 
    # Now we combine the sPlot training with a PDF classifier for M, in one step
    general_options.m_identifier = "MVASPlotCombined"
    meta_options.m_splot_combined = True
    meta_options.m_splot_boosted = False
    basf2_mva.teacher(general_options, fastbdt_options, meta_options)
 
    # Now we use a boosted sPlot training
    general_options.m_identifier = "MVASPlotBoosted"
    meta_options.m_splot_combined = False
    meta_options.m_splot_boosted = True
    basf2_mva.teacher(general_options, fastbdt_options, meta_options)
 
    # And finally a boosted and combined training
    general_options.m_identifier = "MVASPlotCombinedBoosted"
    meta_options.m_splot_combined = True
    meta_options.m_splot_boosted = True
    basf2_mva.teacher(general_options, fastbdt_options, meta_options)
 
    # Also do a training of only the pdf classifier
    pdf_options = basf2_mva.PDFOptions()
    general_options.m_method = 'PDF'
    general_options.m_identifier = "MVAPdf"
    general_options.m_variables = basf2_mva.vector('M')
    basf2_mva.teacher(general_options, pdf_options)