Collaboration diagram for ShowerShapeRegression.ShowerShapeRegressor:

Public Member Functions
def	__init__ (self, root_file_name='ISF_HitAnalysispion.root', base_output_folder='RegressionOutput')

def	initialize_cumulative_trees (self)

def	fill_regression_tree (self, cumulative_histogram)

def	run_regression (self)

def	run (self)

def	process_entry (self, chain, event_number)

def	take_event_into_account (self, chain, event_number)

def	create_output (self)

def	check_output_folders (self)

def	obtain_output_names (self)

def	create_eta_phi_histogram (self, name, title)

def	shower_test_plot_getrandom2 (self, histogram)

def	shower_test_plot_tmva (self, reference, randomize_pos=True)

def	plot_regression_control_plots (self, file_name=None, show_target=False, compare_to_train=False)

def	difference_to_reference (self, reference, histogram, output_name, x_label, y_label, inverted_ratio=False)

def	get_layer_name (self, layer_id)

Static Public Member Functions
def	check_folder (folder)

def	plot_root_hist2d (histogram, file_name='root_hist2d.pdf', x_label='X-Axis', y_label='Y-Axis', z_label='Probability', logscale=False, normalize=False, root_file=None)

def	phi_mpi_pi (phi)

Detailed Description

Definition at line 6 of file ShowerShapeRegression.py.

Constructor & Destructor Documentation

◆ init()

def ShowerShapeRegression.ShowerShapeRegressor.__init__	(	self,
		root_file_name = `'ISF_HitAnalysispion.root'`,
		base_output_folder = `'RegressionOutput'`
	)

Most of the important settings are configured/set within this function.

TODO: Make these options configurable via an external options file to ease the use of the script.
TODO: Make remaining hardcoded options configurable here.

Definition at line 7 of file ShowerShapeRegression.py.

     def __init__(self, root_file_name='ISF_HitAnalysispion.root', base_output_folder='RegressionOutput'):
         """ Most of the important settings are configured/set within this function.
  
         TODO: Make these options configurable via an external options file to ease the use of the script.
         TODO: Make remaining hardcoded options configurable here.
         """
  
         self.neuron_type = 'tanh'   # linear does not work when fitting the efrac without log, seems to cause huge troubles somewhere
  
         self.log_efrac = False
  
         self.num_neurons = 10
         self.root_file_name = root_file_name
         self.base_output_folder = base_output_folder
  
         self.truth_energy_threshold = 40000
  
         self.check_output_folders()
  
         self.etaphi_nbins = [400, 400]
         self.etaphi_xrange = [-0.4, 0.4]
         self.etaphi_yrange = [-0.4, 0.4]
  
         self.cumulative_etaphi_nbins = [400, 400]
         self.cumulative_etaphi_xrange = [-0.4, 0.4]
         self.cumulative_etaphi_yrange = [-0.4, 0.4]
  
         self.equidistant_binning = False
  
         # self.etaphi_nbins = [200, 200]
         # self.etaphi_xrange = [-0.2, 0.2]
         # self.etaphi_yrange = [-0.2, 0.2]
         #
         # self.cumulative_etaphi_nbins = [200, 200]
         # self.cumulative_etaphi_xrange = [-0.2, 0.2]
         # self.cumulative_etaphi_yrange = [-0.2, 0.2]
  
         self.eta_phi_efrac_hists = []
  
         self.cumulated_events = 0
         self.cumulated_events_threshold = 1000
  
         self.layer_names = ["PreSamplerB", "EMB1", "EMB2", "EMB3",
                             "PreSamplerE", "EME1", "EME2", "EME3",
                             "HEC0", "HEC1", "HEC2", "HEC3",
                             "TileBar0", "TileBar1", "TileBar2",
                             "TileGap1", "TileGap2", "TileGap3",
                             "TileExt0", "TileExt1", "TileExt2",
                             "FCAL0", "FCAL1", "FCAL2"]
  
         self.selected_layer = 13
  
         # Trees needed for regression
         self.cum_distribution_trees = None
  
         from array import array
         self.cum_eta_array = array('f', [0.])
         self.cum_phi_array = array('f', [0.])
         self.cum_r_array = array('f', [0.])
         self.cum_efrac_array = array('f', [0.])
  
         self.regression_method = 'MLP'
         #self.regression_method = 'FDA_GA'
  
         self.base_file_name = 'no_name'
         self.mlp_base_file_name = 'no_name'
         self.obtain_output_names()
  
         self.do_regression = True
  

Member Function Documentation

◆ check_folder()

def ShowerShapeRegression.ShowerShapeRegressor.check_folder ( folder )

static

Definition at line 350 of file ShowerShapeRegression.py.

     def check_folder(folder):
         import os
         from os import path
         if not path.exists(folder):
             print('Output folder %s does not exist, will try to create it.' % folder)
             os.makedirs(folder)
             if not path.exists(folder):
                 raise Exception('ERROR: Could not create output folder!')
             else:
                 print('Basic output folder successfully created :)')
  

◆ check_output_folders()

def ShowerShapeRegression.ShowerShapeRegressor.check_output_folders ( self )

Definition at line 361 of file ShowerShapeRegression.py.

     def check_output_folders(self):
         from os import path
         self.check_folder(self.base_output_folder)
         self.check_folder(path.join(self.base_output_folder, 'cumulative_eta_phi'))
         pass
  

◆ create_eta_phi_histogram()

def ShowerShapeRegression.ShowerShapeRegressor.create_eta_phi_histogram	(	self,
		name,
		title
	)

This is a helper function to create the eta phi histograms either with equidistant or non-equidistant
binning depending on the settings. Please use this function to avoid code duplication.

Definition at line 389 of file ShowerShapeRegression.py.

     def create_eta_phi_histogram(self, name, title):
         """ This is a helper function to create the eta phi histograms either with equidistant or non-equidistant
         binning depending on the settings. Please use this function to avoid code duplication.
         """
  
         import ROOT
         if self.equidistant_binning:
             eta_phi_histogram = ROOT.TH2D(name, title,
                                           self.cumulative_etaphi_nbins[0], self.cumulative_etaphi_xrange[0],
                                           self.cumulative_etaphi_xrange[1], self.cumulative_etaphi_nbins[1],
                                           self.cumulative_etaphi_yrange[0], self.cumulative_etaphi_yrange[1])
         else:
             import array
             x_binning = [-0.4, -0.1, -0.05, -0.04, -0.03, -0.02, -0.016, -0.014, -0.012, -0.01, -0.008, -0.006,
                          -0.004, -0.002, 0, 0.002, 0.004, 0.006, 0.008, 0.01, 0.012, 0.014, 0.016, 0.02, 0.03, 0.04,
                          0.05, 0.1, 0.4]
             y_binning = x_binning
  
             eta_phi_histogram = ROOT.TH2D(name, title,
                                           len(x_binning) - 1, array.array('d', x_binning),
                                           len(y_binning) - 1, array.array('d', y_binning))
  
         return eta_phi_histogram
  

◆ create_output()

def ShowerShapeRegression.ShowerShapeRegressor.create_output ( self )

This function creates the average shower shape from all the energy fraction distributions and plots it. In
case self.do_regression is set to True, it will initialize the input trees used by TMVA (using
self.initialize_cumulative_trees), it will fill them (using self.fill_regression_tree) and run the regression
by calling self.run_regression. Once TMVA finished, regression test/control plots and our shower shape control
plots will be created.

Definition at line 259 of file ShowerShapeRegression.py.

     def create_output(self):
         """ This function creates the average shower shape from all the energy fraction distributions and plots it. In
         case self.do_regression is set to True, it will initialize the input trees used by TMVA (using
         self.initialize_cumulative_trees), it will fill them (using self.fill_regression_tree) and run the regression
         by calling self.run_regression. Once TMVA finished, regression test/control plots and our shower shape control
         plots will be created.
         """
         import ROOT
  
         cumulative_histogram = self.create_eta_phi_histogram(self.base_file_name, 'cumulative eta vs phi vs efrac')
  
         if self.equidistant_binning:
             for hist in self.eta_phi_efrac_hists:
                 cumulative_histogram.Add(hist)
  
             if cumulative_histogram.Integral() > 0:
                 cumulative_histogram.Scale(1.0/cumulative_histogram.Integral())
         else:
             # first fill the cumulative histogram
             x_axis = self.eta_phi_efrac_hists[0].GetXaxis()
             n_bins_x = x_axis.GetNbins()
  
             y_axis = self.eta_phi_efrac_hists[0].GetYaxis()
             n_bins_y = y_axis.GetNbins()
             for x_bin in range(n_bins_x + 1):
                 x_center = x_axis.GetBinCenter(x_bin)
                 for y_bin in range(n_bins_y + 1):
                     y_center = x_axis.GetBinCenter(y_bin)
                     for hist in self.eta_phi_efrac_hists:
                         value = hist.GetBinContent(x_bin, y_bin)
                         cumulative_histogram.Fill(x_center, y_center, value)
  
             # now calculate the PDF
             x_axis = cumulative_histogram.GetXaxis()
             n_bins_x = x_axis.GetNbins()
  
             y_axis = cumulative_histogram.GetYaxis()
             n_bins_y = y_axis.GetNbins()
             for x_bin in range(n_bins_x + 1):
                 x_bin_width = x_axis.GetBinWidth(x_bin)
                 for y_bin in range(n_bins_y + 1):
                     y_bin_width = y_axis.GetBinWidth(y_bin)
                     area = x_bin_width*y_bin_width
  
                     cumulative_content = cumulative_histogram.GetBinContent(x_bin, y_bin) / area
                     cumulative_histogram.SetBinContent(x_bin, y_bin, cumulative_content)
  
             if cumulative_histogram.Integral() > 0:
                 cumulative_histogram.Scale(1.0/cumulative_histogram.Integral())
  
         from os import path
         cumulative_etaphi_file_name = path.join(self.base_output_folder, 'cumulative_eta_phi',
                                                 'average_shower_%s.pdf' % self.base_file_name)
         cumulative_etaphi_root_file_name = path.join(self.base_output_folder, 'cumulative_eta_phi',
                                                      '%s.root' % self.base_file_name)
  
         root_file = ROOT.TFile(cumulative_etaphi_root_file_name, 'RECREATE')
         root_file.cd()
         self.plot_root_hist2d(cumulative_histogram.Clone(), cumulative_etaphi_file_name, '#eta', '#phi',
                               root_file=root_file)
         root_file.Write()
         root_file.Close()
  
         if self.do_regression:
             if cumulative_histogram.Integral() > 0:
                 cumulative_histogram.Scale(1.0/cumulative_histogram.Integral())
  
             self.initialize_cumulative_trees()
             self.fill_regression_tree(cumulative_histogram)
  
             self.run_regression()
             self.plot_regression_control_plots()
             self.shower_test_plot_tmva(cumulative_histogram)
             self.shower_test_plot_tmva(cumulative_histogram, False)
             self.shower_test_plot_getrandom2(cumulative_histogram)
  
         for hist in self.eta_phi_efrac_hists:
             hist.Delete()
  
         self.eta_phi_efrac_hists = []
         self.cumulated_events = 0
  
         return
  

◆ difference_to_reference()

def ShowerShapeRegression.ShowerShapeRegressor.difference_to_reference	(	self,
		reference,
		histogram,
		output_name,
		x_label,
		y_label,
		inverted_ratio = `False`
	)

Definition at line 685 of file ShowerShapeRegression.py.

     def difference_to_reference(self, reference, histogram, output_name, x_label, y_label, inverted_ratio=False):
         if inverted_ratio:
             clone = reference.Clone(reference.GetName() + 'diff_to_ref_clone')
             base = histogram
         else:
             clone = histogram.Clone(histogram.GetName() + 'diff_to_ref_clone')
             base = reference
  
         clone.Add(base, -1.0)
         clone.Divide(base)
  
         if inverted_ratio:
             z_label = 'reference - output / output'
         else:
             z_label = 'output - reference / reference'
  
         self.plot_root_hist2d(clone, output_name, x_label, y_label, z_label, logscale=True)
  
         pass
  

◆ fill_regression_tree()

def ShowerShapeRegression.ShowerShapeRegressor.fill_regression_tree	(	self,
		cumulative_histogram
	)

Uses the supplied histogram to fill the regression tree with values. Depending on self.log_efrac the trees
are either filled with the logarithm of the energy fraction or with just the energy fraction.

Definition at line 89 of file ShowerShapeRegression.py.

     def fill_regression_tree(self, cumulative_histogram):
         """ Uses the supplied histogram to fill the regression tree with values. Depending on self.log_efrac the trees
         are either filled with the logarithm of the energy fraction or with just the energy fraction.
         """
         import math
  
         n_bins_x = cumulative_histogram.GetNbinsX()
         n_bins_y = cumulative_histogram.GetNbinsY()
         from past.builtins import xrange # Temporary workaround for python2/3 compatibility - should really use range for python3
         for x in xrange(n_bins_x + 1):
             eta_value = cumulative_histogram.GetXaxis().GetBinCenter(x)
             for y in xrange(n_bins_y + 1):
                 value = cumulative_histogram.GetBinContent(x, y)
                 if value <= 0:
                     continue
  
                 phi_value = cumulative_histogram.GetYaxis().GetBinCenter(y)
                 r_value = math.sqrt(phi_value*phi_value + eta_value*eta_value)
  
                 if self.log_efrac:
                     self.cum_efrac_array[0] = math.log(value)
                 else:
                     self.cum_efrac_array[0] = value
  
                 self.cum_eta_array[0] = eta_value
                 self.cum_phi_array[0] = phi_value
                 self.cum_r_array[0] = r_value
  
                 self.cum_distribution_trees.Fill()
  
         pass
  

◆ get_layer_name()

def ShowerShapeRegression.ShowerShapeRegressor.get_layer_name	(	self,
		layer_id
	)

Definition at line 705 of file ShowerShapeRegression.py.

     def get_layer_name(self, layer_id):
         if layer_id < 0 or layer_id >= len(self.layer_names):
             return 'Unknown'
         else:
             return self.layer_names[layer_id]
  

◆ initialize_cumulative_trees()

def ShowerShapeRegression.ShowerShapeRegressor.initialize_cumulative_trees ( self )

Definition at line 77 of file ShowerShapeRegression.py.

     def initialize_cumulative_trees(self):
         import ROOT
         self.cum_distribution_trees = ROOT.TTree('cumulative_distributions',
                                                  'Cumulative eta phi energy fraction distributions')
  
         self.cum_distribution_trees.Branch('d_eta', self.cum_eta_array, 'd_eta/F')
         self.cum_distribution_trees.Branch('d_phi', self.cum_phi_array, 'd_phi/F')
         self.cum_distribution_trees.Branch('r', self.cum_r_array, 'r/F')
         self.cum_distribution_trees.Branch('energy_fraction', self.cum_efrac_array, 'energy_fraction/F')
  
         pass
  

◆ obtain_output_names()

def ShowerShapeRegression.ShowerShapeRegressor.obtain_output_names ( self )

Definition at line 367 of file ShowerShapeRegression.py.

     def obtain_output_names(self):
         ext = ''
         if self.log_efrac:
             ext = '_logEfrac'
  
         eq_bin = ''
         if self.equidistant_binning:
             eq_bin = '_eqbin'
  
         self.base_file_name = 'Evts%i_Lay%i_E%i_eta%.2f_PID%i%s' \
                               % (self.cumulated_events_threshold, self.selected_layer, 50, 0.20, 211, eq_bin)
         self.mlp_base_file_name = 'NN_reg%s_%s_NNeur%i_NT%s_%s' % (ext, self.regression_method, self.num_neurons,
                                                                    self.neuron_type, self.base_file_name)
  

◆ phi_mpi_pi()

def ShowerShapeRegression.ShowerShapeRegressor.phi_mpi_pi ( phi )

static

Definition at line 712 of file ShowerShapeRegression.py.

     def phi_mpi_pi(phi):
         import ROOT
         pi = ROOT.TMath.Pi()
         while phi >= pi:
             phi -= pi
         while phi < -pi:
             phi += pi
         return phi
  
  

◆ plot_regression_control_plots()

def ShowerShapeRegression.ShowerShapeRegressor.plot_regression_control_plots	(	self,
		file_name = `None`,
		show_target = `False`,
		compare_to_train = `False`
	)

This is a pythonic translation of the deviations function found in the TMVA examples.

Definition at line 541 of file ShowerShapeRegression.py.

     def plot_regression_control_plots(self, file_name=None, show_target=False, compare_to_train=False):
         """ This is a pythonic translation of the deviations function found in the TMVA examples.
         """
         import ROOT
         ROOT.gROOT.SetBatch(True)
  
         #ROOT.TMVA.TMVAGlob.Initialize(True)
         ROOT.gStyle.SetNumberContours(999)
  
         if not file_name:
             file_name = '%s.root' % self.mlp_base_file_name
         input_file = ROOT.TFile(file_name)
  
         current_canvas = 0
         for key in input_file.GetListOfKeys():
             print('Key: %s' % key.GetName())
  
             if 'Method_' not in key.GetName():
                 continue
             if not ROOT.gROOT.GetClass(key.GetClassName()).InheritsFrom('TDirectory'):
                 continue
  
             method_name = key.GetName().replace('Method_', '')
  
             print('Now plotting control plots for method %s' % method_name)
  
             directory = key.ReadObj()
  
             for sub_method in directory.GetListOfKeys():
                 print('sub_method: %s' % sub_method.GetName())
  
                 if not ROOT.gROOT.GetClass(sub_method.GetClassName()).InheritsFrom('TDirectory'):
                     continue
  
                 sub_method_name = sub_method.GetName()
  
                 print('Now plotting control plots for sub-method %s' % sub_method_name)
  
                 sub_directory = sub_method.ReadObj()
  
                 for plot_key in sub_directory.GetListOfKeys():
                     print('plot_key: %s' % plot_key.GetName())
  
                     plot_obj = plot_key.ReadObj()
                     if isinstance(plot_obj, ROOT.TH2F):
                         plot_name = plot_key.GetName()
                         if '_reg_' not in plot_name:
                             continue
  
                         if not ((show_target and '_tgt' in plot_name) or
                                 (not show_target and ('_tgt' not in plot_name))):
                             continue
  
                         if not ((compare_to_train and 'train' in plot_name) or
                                 (not compare_to_train and 'test' in plot_name)):
                             continue
  
                         inortarg = 'input'
                         type_string = 'test'
                         if show_target:
                             inortarg = 'target'
  
                         if compare_to_train:
                             type_string = 'training'
  
                         canvas = ROOT.TCanvas('canvas_%i' % current_canvas,
                                               'Regression output deviation versus %s for method %s' % (inortarg,
                                               method_name), 800, 600)
  
                         canvas.SetRightMargin(0.10)
  
                         plot_obj.SetTitle('Output deviation for method: %s (%s sample)' % (sub_method_name,
                                                                                            type_string))
                         plot_obj.Draw('colz')
  
                         line = ROOT.TLine(plot_obj.GetXaxis().GetXmin(), 0, plot_obj.GetXaxis().GetXmax(), 0)
                         line.SetLineStyle(2)
                         line.Draw()
  
                         canvas.Update()
  
                         import os
                         output_name = os.path.join(self.base_output_folder, '%s_dev_%s_%s_%s_%i.pdf' %
                                                    (self.mlp_base_file_name, method_name, inortarg, type_string,
                                                     current_canvas))
                         #ROOT.TMVAGlob.imgconv(canvas, output_name)
                         canvas.Print(output_name, 'pdf')
  
                         current_canvas += 1
  
         pass
  

◆ plot_root_hist2d()

def ShowerShapeRegression.ShowerShapeRegressor.plot_root_hist2d	(	histogram,
		file_name = `'root_hist2d.pdf'`,
		x_label = `'X-Axis'`,
		y_label = `'Y-Axis'`,
		z_label = `'Probability'`,
		logscale = `False`,
		normalize = `False`,
		root_file = `None`
	)

static

Definition at line 634 of file ShowerShapeRegression.py.

     def plot_root_hist2d(histogram, file_name='root_hist2d.pdf', x_label='X-Axis', y_label='Y-Axis',
                          z_label='Probability', logscale=False, normalize=False, root_file=None):
         import ROOT
         ROOT.gROOT.SetBatch(True)
         ROOT.gStyle.SetOptStat(1)
  
         from os import path
         ext = path.splitext(path.basename(file_name))[1]
  
         canv = ROOT.TCanvas('canvas', 'test', 0, 0, 600, 600)
         #canv.SetTopMargin(0.05)
         #canv.SetLeftMargin(0.13)
         canv.SetRightMargin(0.15)
         #canv.SetBottomMargin(0.035)
  
         if logscale:
             canv.SetLogz()
  
         if normalize:
             integral = histogram.Integral()
             if integral > 0:
                 histogram.Scale(1.0 / integral)
  
         histogram.GetXaxis().SetTitle(x_label)
         histogram.GetYaxis().SetTitle(y_label)
         histogram.GetZaxis().SetTitle(z_label)
  
         # histogram.GetYaxis().SetLabelSize(0.06)
         # histogram.GetYaxis().SetLabelOffset(0.0015)
         #
         # histogram.GetYaxis().SetTitleSize(0.06)
         # histogram.GetYaxis().SetTitleOffset(0.85)
  
         histogram.Draw('COLZ')
  
         canv.Update()
         canv.Print(file_name, ext)
  
         if root_file:
             histogram.SetDirectory(root_file)
             histogram.Write()
  
         pass
  

◆ process_entry()

def ShowerShapeRegression.ShowerShapeRegressor.process_entry	(	self,
		chain,
		event_number
	)

This processes a single event in the chain. It first checks whether this event should be processed by
calling the take_event_into_account function. If this function returns true processing of this event continues
else the event will be skipped.
Only a single layer (self.selected_layer) will be processed. A histogram will be filled with the energy
distribution vs eta and phi. Corrections by the track to calo extrapolator will be applied here. Once all
events have been finished the histogram will be divided by its integral.
Once the desired number of events (self.cumulated_events_threshold) has been reached the self.create_output()
function will be called.

Definition at line 190 of file ShowerShapeRegression.py.

     def process_entry(self, chain, event_number):
         """ This processes a single event in the chain. It first checks whether this event should be processed by
         calling the take_event_into_account function. If this function returns true processing of this event continues
         else the event will be skipped.
         Only a single layer (self.selected_layer) will be processed. A histogram will be filled with the energy
         distribution vs eta and phi. Corrections by the track to calo extrapolator will be applied here. Once all
         events have been finished the histogram will be divided by its integral.
         Once the desired number of events (self.cumulated_events_threshold) has been reached the self.create_output()
         function will be called.
         """
         import ROOT
  
         if not self.take_event_into_account(chain, event_number):
             return True
  
         num_hits = len(chain.HitX)
  
         eta_phi_efrac_dist = ROOT.TH2D('eta_phi_efrac_dist_%i' % event_number, 'eta vs phi vs efrac',
                                        self.cumulative_etaphi_nbins[0], self.cumulative_etaphi_xrange[0],
                                        self.cumulative_etaphi_xrange[1], self.cumulative_etaphi_nbins[1],
                                        self.cumulative_etaphi_yrange[0], self.cumulative_etaphi_yrange[1])
         for i in range(num_hits):
             layer_id = chain.HitSampling[i]
  
             if not layer_id == self.selected_layer:
                 continue
  
             pos = ROOT.TVector3(chain.HitX[i], chain.HitY[i], chain.HitZ[i])
             eta = pos.PseudoRapidity()
             phi = pos.Phi()
  
             eta_correction = chain.TTC_entrance_eta[0][layer_id]
             phi_correction = chain.TTC_entrance_phi[0][layer_id]
  
             if eta_correction < -900:
                 eta_correction = 0
                 print('no valid eta_correction found')
  
             if phi_correction < -900:
                 phi_correction = 0
                 print('no valid phi_correction found')
  
             d_eta = eta - eta_correction
             d_phi = self.phi_mpi_pi(phi - phi_correction)
  
             eta_phi_efrac_dist.Fill(d_eta, d_phi, chain.HitE[i])
  
         self.cumulated_events += 1
  
         if eta_phi_efrac_dist.Integral() > 0:
                 eta_phi_efrac_dist.Scale(1.0/eta_phi_efrac_dist.Integral())
  
         self.eta_phi_efrac_hists.append(eta_phi_efrac_dist)
         if self.cumulated_events >= self.cumulated_events_threshold:
             self.create_output()
             return False
  
         return True
  

◆ run()

def ShowerShapeRegression.ShowerShapeRegressor.run ( self )

This function iterates over all events and calls for each event the process_entry function. The event loop
breaks either once process_entry returns false or once all events in the input file have been processed.

Definition at line 153 of file ShowerShapeRegression.py.

     def run(self):
         """ This function iterates over all events and calls for each event the process_entry function. The event loop
         breaks either once process_entry returns false or once all events in the input file have been processed.
         """
         import ROOT
  
         print('--------------------------------------------')
         print('Welcome to ShowerShapeRegressor.py')
         print('Using %i neurons of type %s using log of efrac: %s' % (self.num_neurons, self.neuron_type,
                                                                       str(self.log_efrac)))
         print('--------------------------------------------')
  
         self.obtain_output_names()
  
         f = ROOT.TFile(self.root_file_name)  # noqa: F841
         chain = ROOT.gDirectory.Get('ISF_HitAnalysis/CaloHitAna')
         entries = chain.GetEntriesFast()
  
         for current_entry in range(entries):
             print(' Loading entry: %i' % current_entry)
             j = chain.LoadTree(current_entry)
             if j < 0:
                 break
  
             # copy next entry into memory and verify
             bites_read = chain.GetEntry(current_entry)
             if bites_read <= 0:
                 continue
  
             if not self.process_entry(chain, current_entry):
                 break
  
         if len(self.eta_phi_efrac_hists) > 0:
             self.create_output()
  
         pass
  

◆ run_regression()

def ShowerShapeRegression.ShowerShapeRegressor.run_regression ( self )

Definition at line 121 of file ShowerShapeRegression.py.

     def run_regression(self):
         import ROOT
  
         output_file = ROOT.TFile("%s.root" % self.mlp_base_file_name, "RECREATE")
  
         factory = ROOT.TMVA.Factory(self.mlp_base_file_name, output_file, "!V:!Silent:!Color:DrawProgressBar")
  
         factory.AddVariable("d_eta")
         factory.AddVariable("d_phi")
         factory.AddVariable("r")
  
         factory.AddTarget("energy_fraction")
  
         factory.AddRegressionTree(self.cum_distribution_trees, 1.0)
  
         cut = ROOT.TCut('')
  
         factory.PrepareTrainingAndTestTree(cut, 'nTrain_Regression=0:nTest_Regression=0:SplitMode=Random:NormMode=NumEvents:!V')
  
         # Default: HiddenLayers=N+20
         factory.BookMethod(ROOT.TMVA.Types.kMLP, "MLP",
                            "!H:!V:VarTransform=Norm:NeuronType=%s:NCycles=20000:HiddenLayers=N+%i:TestRate=6:TrainingMethod=BFGS:Sampling=0.3:SamplingEpoch=0.8:ConvergenceImprove=1e-6:ConvergenceTests=15:!UseRegulator"
                            % (self.neuron_type, self.num_neurons))
  
         factory.TrainAllMethods()
         factory.TestAllMethods()
         factory.EvaluateAllMethods()
  
         output_file.Close()
  
         pass
  

◆ shower_test_plot_getrandom2()

def ShowerShapeRegression.ShowerShapeRegressor.shower_test_plot_getrandom2	(	self,
		histogram
	)

This function uses the input histogram's GetRandom2 function to draw random numbers according to the
distribution found in the input histogram and fills 50000 entries into a new histogram and in the end prints
it to a file.

Definition at line 413 of file ShowerShapeRegression.py.

     def shower_test_plot_getrandom2(self, histogram):
         """ This function uses the input histogram's GetRandom2 function to draw random numbers according to the
         distribution found in the input histogram and fills 50000 entries into a new histogram and in the end prints
         it to a file.
         """
         import ROOT
         from os import path
         ROOT.gROOT.SetBatch(True)
  
         ROOT.gStyle.SetNumberContours(999)
  
         d_eta = ROOT.Double(0)
         d_phi = ROOT.Double(0)
  
         eta_phi_efrac_dist = self.create_eta_phi_histogram('shower_test_plot',
                                                            'eta vs phi vs efrac - using TH2F::GetRandom2')
  
         accepted_points = 0
         accepted_points_target = 50000
         while accepted_points < accepted_points_target:
             histogram.GetRandom2(d_eta, d_phi)
  
             eta_phi_efrac_dist.Fill(d_eta, d_phi, 1)
             accepted_points += 1
  
         if eta_phi_efrac_dist.Integral() > 0:
             eta_phi_efrac_dist.Scale(1.0/eta_phi_efrac_dist.Integral())
         self.plot_root_hist2d(eta_phi_efrac_dist,
                               path.join(self.base_output_folder, '%s_th2f.pdf' % self.mlp_base_file_name), '#eta', '#phi')
         self.difference_to_reference(histogram, eta_phi_efrac_dist,
                                      path.join(self.base_output_folder, '%s_th2f_diff.pdf' % self.mlp_base_file_name),
                                      '#eta', '#phi')
  
         self.difference_to_reference(histogram, eta_phi_efrac_dist,
                                      path.join(self.base_output_folder, '%s_th2f_diff_inf.pdf' % self.mlp_base_file_name),
                                      '#eta', '#phi', inverted_ratio=True)
  
         pass
  

◆ shower_test_plot_tmva()

def ShowerShapeRegression.ShowerShapeRegressor.shower_test_plot_tmva	(	self,
		reference,
		randomize_pos = `True`
	)

This function uses the NN obtained by TMVA to plot the fitted shower shape and the ratio to the reference.
In case randomize_pos is set to true (which is the default), random positions (flat distributed) in the eta-phi
plane will drawn. The NN will be evaluated there and the value will be filled into the output histogram. This
method is faster than drawing a third random number and accepting the point based on whether this rnd number is
greater than the NN at this point or not.
If randomize_pos is set to false, a histogram will be filled by evaluating the NN at each point and filling the
value into the histogram.

Definition at line 452 of file ShowerShapeRegression.py.

     def shower_test_plot_tmva(self, reference, randomize_pos=True):
         """ This function uses the NN obtained by TMVA to plot the fitted shower shape and the ratio to the reference.
         In case randomize_pos is set to true (which is the default), random positions (flat distributed) in the eta-phi
         plane will drawn. The NN will be evaluated there and the value will be filled into the output histogram. This
         method is faster than drawing a third random number and accepting the point based on whether this rnd number is
         greater than the NN at this point or not.
         If randomize_pos is set to false, a histogram will be filled by evaluating the NN at each point and filling the
         value into the histogram.
         """
         import ROOT
         import math
         ROOT.gROOT.SetBatch(True)
  
         ROOT.gStyle.SetNumberContours(999)
  
         reader = ROOT.TMVA.Reader('!Color')
  
         from array import array
         d_eta = array('f', [0.])
         d_phi = array('f', [0.])
         r = array('f', [0])
         reader.AddVariable('d_eta', d_eta)
         reader.AddVariable('d_phi', d_phi)
         reader.AddVariable('r', r)
  
         from os import path
         base_weight_name = self.mlp_base_file_name
         method_name = self.regression_method
         weight_file = path.join('weights', '%s_%s.weights.xml' % (base_weight_name, method_name))
         reader.BookMVA(self.regression_method, weight_file)
  
         ext = ''
         if randomize_pos:
             ext = ' rnd pos'
  
         eta_phi_efrac_dist = self.create_eta_phi_histogram('shower_test_plot', 'eta vs phi vs efrac - using %s%s' %
                                                            (self.regression_method, ext))
         if randomize_pos:
             accepted_points = 0
             accepted_points_target = 50000
             while accepted_points < accepted_points_target:
                 d_eta[0] = ROOT.gRandom.Rndm() - 0.5
                 d_phi[0] = ROOT.gRandom.Rndm() - 0.5
                 r[0] = math.sqrt(d_eta[0]*d_eta[0] + d_phi[0]*d_phi[0])
  
                 value = reader.EvaluateRegression(self.regression_method)[0]
  
                 if self.log_efrac:
                     value = math.exp(value)
  
                 eta_phi_efrac_dist.Fill(d_eta[0], d_phi[0], value)
                 accepted_points += 1
         else:
             x_axis = reference.GetXaxis()
             y_axis = reference.GetYaxis()
             n_bins_x = x_axis.GetNbins()
             n_bins_y = y_axis.GetNbins()
             for x in range(n_bins_x + 1):
                 d_eta[0] = x_axis.GetBinCenter(x)
                 for y in range(n_bins_y + 1):
                     d_phi[0] = y_axis.GetBinCenter(y)
                     r[0] = math.sqrt(d_eta[0]*d_eta[0] + d_phi[0]*d_phi[0])
  
                     value = reader.EvaluateRegression(self.regression_method)[0]
  
                     if self.log_efrac:
                         value = math.exp(value)
  
                     eta_phi_efrac_dist.Fill(d_eta[0], d_phi[0], value)
  
         if eta_phi_efrac_dist.Integral() > 0:
             eta_phi_efrac_dist.Scale(1.0/eta_phi_efrac_dist.Integral())
  
         if randomize_pos:
             ext = '_rnd'
  
         self.plot_root_hist2d(eta_phi_efrac_dist,
                               path.join(self.base_output_folder, '%s%s.pdf' % (self.mlp_base_file_name, ext)),
                               '#eta', '#phi')
         self.difference_to_reference(reference, eta_phi_efrac_dist,
                                      path.join(self.base_output_folder, '%s%s_diff.pdf' %
                                                (self.mlp_base_file_name, ext)), '#eta', '#phi')
  
         self.difference_to_reference(reference, eta_phi_efrac_dist,
                                      path.join(self.base_output_folder, '%s%s_diff_inv.pdf' %
                                                (self.mlp_base_file_name, ext)), '#eta', '#phi', inverted_ratio=True)
  
         pass
  

◆ take_event_into_account()

def ShowerShapeRegression.ShowerShapeRegressor.take_event_into_account	(	self,
		chain,
		event_number
	)

This function is used to filter events.

Definition at line 249 of file ShowerShapeRegression.py.

     def take_event_into_account(self, chain, event_number):
         """ This function is used to filter events.
         """
         if chain.TruthE[0] < self.truth_energy_threshold:
             print('  Pion does not pass energy threshold of %f skipping process_entry!' %
                   self.truth_energy_threshold)
             return False
  
         return True
  

Member Data Documentation

◆ base_file_name

ShowerShapeRegression.ShowerShapeRegressor.base_file_name

Definition at line 71 of file ShowerShapeRegression.py.

◆ base_output_folder

ShowerShapeRegression.ShowerShapeRegressor.base_output_folder

Definition at line 20 of file ShowerShapeRegression.py.

◆ cum_distribution_trees

ShowerShapeRegression.ShowerShapeRegressor.cum_distribution_trees

Definition at line 60 of file ShowerShapeRegression.py.

◆ cum_efrac_array

ShowerShapeRegression.ShowerShapeRegressor.cum_efrac_array

Definition at line 66 of file ShowerShapeRegression.py.

◆ cum_eta_array

ShowerShapeRegression.ShowerShapeRegressor.cum_eta_array

Definition at line 63 of file ShowerShapeRegression.py.

◆ cum_phi_array

ShowerShapeRegression.ShowerShapeRegressor.cum_phi_array

Definition at line 64 of file ShowerShapeRegression.py.

◆ cum_r_array

ShowerShapeRegression.ShowerShapeRegressor.cum_r_array

Definition at line 65 of file ShowerShapeRegression.py.

◆ cumulated_events

ShowerShapeRegression.ShowerShapeRegressor.cumulated_events

Definition at line 46 of file ShowerShapeRegression.py.

◆ cumulated_events_threshold

ShowerShapeRegression.ShowerShapeRegressor.cumulated_events_threshold

Definition at line 47 of file ShowerShapeRegression.py.

◆ cumulative_etaphi_nbins

ShowerShapeRegression.ShowerShapeRegressor.cumulative_etaphi_nbins

Definition at line 30 of file ShowerShapeRegression.py.

◆ cumulative_etaphi_xrange

ShowerShapeRegression.ShowerShapeRegressor.cumulative_etaphi_xrange

Definition at line 31 of file ShowerShapeRegression.py.

◆ cumulative_etaphi_yrange

ShowerShapeRegression.ShowerShapeRegressor.cumulative_etaphi_yrange

Definition at line 32 of file ShowerShapeRegression.py.

◆ do_regression

ShowerShapeRegression.ShowerShapeRegressor.do_regression

Definition at line 75 of file ShowerShapeRegression.py.

◆ equidistant_binning

ShowerShapeRegression.ShowerShapeRegressor.equidistant_binning

Definition at line 34 of file ShowerShapeRegression.py.

◆ eta_phi_efrac_hists

ShowerShapeRegression.ShowerShapeRegressor.eta_phi_efrac_hists

Definition at line 44 of file ShowerShapeRegression.py.

◆ etaphi_nbins

ShowerShapeRegression.ShowerShapeRegressor.etaphi_nbins

Definition at line 26 of file ShowerShapeRegression.py.

◆ etaphi_xrange

ShowerShapeRegression.ShowerShapeRegressor.etaphi_xrange

Definition at line 27 of file ShowerShapeRegression.py.

◆ etaphi_yrange

ShowerShapeRegression.ShowerShapeRegressor.etaphi_yrange

Definition at line 28 of file ShowerShapeRegression.py.

◆ layer_names

ShowerShapeRegression.ShowerShapeRegressor.layer_names

Definition at line 49 of file ShowerShapeRegression.py.

◆ log_efrac

ShowerShapeRegression.ShowerShapeRegressor.log_efrac

Definition at line 16 of file ShowerShapeRegression.py.

◆ mlp_base_file_name

ShowerShapeRegression.ShowerShapeRegressor.mlp_base_file_name

Definition at line 72 of file ShowerShapeRegression.py.

◆ neuron_type

ShowerShapeRegression.ShowerShapeRegressor.neuron_type

Definition at line 14 of file ShowerShapeRegression.py.

◆ num_neurons

ShowerShapeRegression.ShowerShapeRegressor.num_neurons

Definition at line 18 of file ShowerShapeRegression.py.

◆ regression_method

ShowerShapeRegression.ShowerShapeRegressor.regression_method

Definition at line 68 of file ShowerShapeRegression.py.

◆ root_file_name

ShowerShapeRegression.ShowerShapeRegressor.root_file_name

Definition at line 19 of file ShowerShapeRegression.py.

◆ selected_layer

ShowerShapeRegression.ShowerShapeRegressor.selected_layer

Definition at line 57 of file ShowerShapeRegression.py.

◆ truth_energy_threshold

ShowerShapeRegression.ShowerShapeRegressor.truth_energy_threshold

Definition at line 22 of file ShowerShapeRegression.py.

The documentation for this class was generated from the following file:

ShowerShapeRegression.py

Public Attributes
	neuron_type

	log_efrac

	num_neurons

	root_file_name

	base_output_folder

	truth_energy_threshold

	etaphi_nbins

	etaphi_xrange

	etaphi_yrange

	cumulative_etaphi_nbins

	cumulative_etaphi_xrange

	cumulative_etaphi_yrange

	equidistant_binning

	eta_phi_efrac_hists

	cumulated_events

	cumulated_events_threshold

	layer_names

	selected_layer

	cum_distribution_trees

	cum_eta_array

	cum_phi_array

	cum_r_array

	cum_efrac_array

	regression_method

	base_file_name

	mlp_base_file_name

	do_regression

Public Member Functions

Static Public Member Functions

Public Attributes

Detailed Description

Constructor & Destructor Documentation

◆ __init__()

Member Function Documentation

◆ check_folder()

◆ check_output_folders()

◆ create_eta_phi_histogram()

◆ create_output()

◆ difference_to_reference()

◆ fill_regression_tree()

◆ get_layer_name()

◆ initialize_cumulative_trees()

◆ obtain_output_names()

◆ phi_mpi_pi()

◆ plot_regression_control_plots()

◆ plot_root_hist2d()

◆ process_entry()

◆ run()

◆ run_regression()

◆ shower_test_plot_getrandom2()

◆ shower_test_plot_tmva()

◆ take_event_into_account()

Member Data Documentation

◆ base_file_name

◆ base_output_folder

◆ cum_distribution_trees

◆ cum_efrac_array

◆ cum_eta_array

◆ cum_phi_array

◆ cum_r_array

◆ cumulated_events

◆ cumulated_events_threshold

◆ cumulative_etaphi_nbins

◆ cumulative_etaphi_xrange

◆ cumulative_etaphi_yrange

◆ do_regression

◆ equidistant_binning

◆ eta_phi_efrac_hists

◆ etaphi_nbins

◆ etaphi_xrange

◆ etaphi_yrange

◆ layer_names

◆ log_efrac

◆ mlp_base_file_name

◆ neuron_type

◆ num_neurons

◆ regression_method

◆ root_file_name

◆ selected_layer

◆ truth_energy_threshold

◆ init()