def get_unfold_histogram_tuple(
inputfile,
variable,
channel,
met_type = 'patType1CorrectedPFMet',
centre_of_mass = 8,
ttbar_xsection = 245.8,
luminosity = 19712,
load_fakes = False,
scale_to_lumi = False,
):
folder = None
h_truth = None
h_measured = None
h_response = None
h_fakes = None
if not channel == 'combined':
if not 'HT' in variable:
folder = inputfile.Get( 'unfolding_%s_analyser_%s_channel_%s' % ( variable, channel, met_type ) )
else:
folder = inputfile.Get( 'unfolding_%s_analyser_%s_channel' % ( variable, channel ) )
h_truth = asrootpy( folder.truth.Clone() )
h_measured = asrootpy( folder.measured.Clone() )
# response matrix is always without fakes
# fake subtraction from measured is performed automatically in RooUnfoldSvd (h_measured - h_response->ProjectionX())
# or manually for TSVDUnfold
# fix for a bug/typo in NTupleTools
h_response = asrootpy( folder.response_without_fakes.Clone() )
if load_fakes:
h_fakes = asrootpy( folder.fake.Clone() )
else:
return get_combined_unfold_histogram_tuple( inputfile = inputfile,
variable = variable,
met_type = met_type,
centre_of_mass = centre_of_mass,
ttbar_xsection = ttbar_xsection,
luminosity = luminosity,
load_fakes = load_fakes,
scale_to_lumi = scale_to_lumi,
)
if scale_to_lumi:
nEvents = inputfile.EventFilter.EventCounter.GetBinContent( 1 ) # number of processed events
lumiweight = ttbar_xsection * luminosity / nEvents
if load_fakes:
h_fakes.Scale( lumiweight )
h_truth.Scale( lumiweight )
h_measured.Scale( lumiweight )
h_response.Scale( lumiweight )
h_truth, h_measured, h_response = [ fix_overflow( hist ) for hist in [h_truth, h_measured, h_response] ]
if load_fakes:
h_fakes = fix_overflow( h_fakes )
return h_truth, h_measured, h_response, h_fakes
def test_overflow_2D( self ):
before_fix = check_overflow_in_2DHist(self.h2)
has_overflow_in_x = before_fix['has_overflow_in_x']
has_overflow_in_y = before_fix['has_overflow_in_y']
self.assertGreater(has_overflow_in_x, 0, '2D hist: No overflow in x present, wrong setup.')
self.assertGreater(has_overflow_in_y, 0, '2D hist: No overflow in y present, wrong setup.')
h2 = fix_overflow( self.h2 )
after_fix = check_overflow_in_2DHist(h2)
has_overflow_in_x = after_fix['has_overflow_in_x']
has_overflow_in_y = after_fix['has_overflow_in_y']
# check if overflow has been reset
self.assertEqual( has_overflow_in_x, 0, '2D hist: Overflow in x is not 0.' )
self.assertEqual( has_overflow_in_y, 0, '2D hist: Overflow in y is not 0.' )
# now check if new last bin content is equal to the old one plus overflow
overflow_x_before = before_fix['overflow_x']
overflow_y_before = before_fix['overflow_y']
last_bin_content_x_before = before_fix['last_bin_content_x']
last_bin_content_y_before = before_fix['last_bin_content_y']
last_bin_content_x_after = after_fix['last_bin_content_x']
last_bin_content_y_after = after_fix['last_bin_content_y']
check_last_bin_content_x = [overflow + last_bin_content for overflow,last_bin_content in zip(overflow_x_before, last_bin_content_x_before)]
check_last_bin_content_y = [overflow + last_bin_content for overflow,last_bin_content in zip(overflow_y_before, last_bin_content_y_before)]
# remember, the last item in each list is actually the overflow, which should be 0 and the above calculation is not correct.
self.assertTrue(check_equal_lists(check_last_bin_content_x[:-2], last_bin_content_x_after[:-2]), '2D hist: last bins in x are not correct.')
self.assertTrue(check_equal_lists(check_last_bin_content_y[:-2], last_bin_content_y_after[:-2]), '2D hist: last bins in y are not correct.')
def calculate_normalisation(self):
'''
1. get file names
2. get histograms from files
3. ???
4. calculate normalisation based on self.method
'''
if self.have_normalisation:
return
histograms = self.measurement.histograms
for sample, hist in histograms.items():
# TODO: this should be a list of bin-contents
hist = fix_overflow(hist)
histograms[sample] = hist
self.initial_normalisation[
sample] = hist_to_value_error_tuplelist(hist)
if self.method == self.BACKGROUND_SUBTRACTION and sample != 'TTJet':
self.normalisation[sample] = self.initial_normalisation[sample]
if self.method == self.BACKGROUND_SUBTRACTION:
self.background_subtraction(histograms)
if self.method == self.SIMULTANEOUS_FIT:
self.simultaneous_fit(histograms)
# next, let's round all numbers (they are event numbers after all
for sample, values in self.normalisation.items():
new_values = [(round(v, 1), round(e, 1)) for v, e in values]
self.normalisation[sample] = new_values
self.have_normalisation = True
def setUp( self ):
# create histograms
# self.h1 = Hist2D( 15, 0, 15, 15, 0, 15 )
# x_1 = [1, 3, 7, 7, 8, 1, 12, 7, 8, 6]
# y_1 = [1, 7, 3, 7, 8, 12, 7, 12, 13, 11]
self.h1 = Hist2D( 60, 40, 100, 60, 40, 100 )
n_1 = 100000
x_1 = 60 + 10 * np.random.randn( n_1 )
y_1 = x_1 + np.random.randn( n_1 )
z_1 = np.vstack( ( x_1, y_1 ) ).T
self.h1.fill_array( z_1 )
self.h1 = fix_overflow( self.h1 )
self.histogram_information = [
{'hist': self.h1,
'CoM': 7,
'channel':'test_1'},
]
self.histograms = [info['hist'] for info in self.histogram_information]
# requirements for new binning
self.p_min, self.s_min, self.n_min = 0.5, 0.5, 1000
self.bin_edges = []
self.purities_GetBinContent = []
self.stabilities_GetBinContent = []
self.n_events_GetBinContent = []
self.purities_Integral = []
self.stabilities_Integral = []
self.n_events_Integral = []
first_hist = self.histograms[0]
n_bins = first_hist.GetNbinsX()
current_bin_start = 0
current_bin_end = 0
while current_bin_end < n_bins:
current_bin_end, p, s, n_gen_and_reco = pick_bins.get_next_end( self.histograms, current_bin_start, current_bin_end, self.p_min, self.s_min, self.n_min )
if not self.bin_edges:
# if empty
self.bin_edges.append( first_hist.GetXaxis().GetBinLowEdge( current_bin_start + 1 ) )
self.bin_edges.append( first_hist.GetXaxis().GetBinLowEdge( current_bin_end ) + first_hist.GetXaxis().GetBinWidth( current_bin_end ) )
self.purities_Integral.append(p)
self.stabilities_Integral.append(s)
self.n_events_Integral.append(n_gen_and_reco)
current_bin_start = current_bin_end
self.h1_rebinned = rebin_2d(self.h1, self.bin_edges, self.bin_edges)
self.purities_GetBinContent = calculate_purities( self.h1_rebinned )
self.stabilities_GetBinContent = calculate_stabilities( self.h1_rebinned )
self.n_events_GetBinContent = [int( self.h1_rebinned.GetBinContent( i, i ) ) for i in range( 1, len( self.bin_edges ) )]
def test_overflow_1D( self ):
last_bin = self.h1.nbins()
overflow_bin = last_bin + 1
overflow = self.h1.GetBinContent( overflow_bin )
last_bin_content = self.h1.GetBinContent( last_bin )
self.assertGreater( overflow, 0, '1D hist: No overflow present, wrong setup.' )
h1 = fix_overflow( self.h1 )
self.assertEqual( h1.GetBinContent( overflow_bin ), 0., '1D hist: Overflow bin is not 0.' )
self.assertEqual( h1.GetBinContent( last_bin ), last_bin_content + overflow, '1D hist: last bin is not correct.' )
def test_overflow_1D(self):
last_bin = self.h1.nbins()
overflow_bin = last_bin + 1
overflow = self.h1.GetBinContent(overflow_bin)
last_bin_content = self.h1.GetBinContent(last_bin)
self.assertGreater(overflow, 0,
'1D hist: No overflow present, wrong setup.')
h1 = fix_overflow(self.h1)
self.assertEqual(h1.GetBinContent(overflow_bin), 0.,
'1D hist: Overflow bin is not 0.')
self.assertEqual(h1.GetBinContent(last_bin),
last_bin_content + overflow,
'1D hist: last bin is not correct.')
def test_overflow_2D(self):
before_fix = check_overflow_in_2DHist(self.h2)
has_overflow_in_x = before_fix['has_overflow_in_x']
has_overflow_in_y = before_fix['has_overflow_in_y']
self.assertGreater(has_overflow_in_x, 0,
'2D hist: No overflow in x present, wrong setup.')
self.assertGreater(has_overflow_in_y, 0,
'2D hist: No overflow in y present, wrong setup.')
h2 = fix_overflow(self.h2)
after_fix = check_overflow_in_2DHist(h2)
has_overflow_in_x = after_fix['has_overflow_in_x']
has_overflow_in_y = after_fix['has_overflow_in_y']
# check if overflow has been reset
self.assertEqual(has_overflow_in_x, 0,
'2D hist: Overflow in x is not 0.')
self.assertEqual(has_overflow_in_y, 0,
'2D hist: Overflow in y is not 0.')
# now check if new last bin content is equal to the old one plus overflow
overflow_x_before = before_fix['overflow_x']
overflow_y_before = before_fix['overflow_y']
last_bin_content_x_before = before_fix['last_bin_content_x']
last_bin_content_y_before = before_fix['last_bin_content_y']
last_bin_content_x_after = after_fix['last_bin_content_x']
last_bin_content_y_after = after_fix['last_bin_content_y']
check_last_bin_content_x = [
overflow + last_bin_content for overflow, last_bin_content in zip(
overflow_x_before, last_bin_content_x_before)
]
check_last_bin_content_y = [
overflow + last_bin_content for overflow, last_bin_content in zip(
overflow_y_before, last_bin_content_y_before)
]
# remember, the last item in each list is actually the overflow, which should be 0 and the above calculation is not correct.
self.assertTrue(
check_equal_lists(check_last_bin_content_x[:-2],
last_bin_content_x_after[:-2]),
'2D hist: last bins in x are not correct.')
self.assertTrue(
check_equal_lists(check_last_bin_content_y[:-2],
last_bin_content_y_after[:-2]),
'2D hist: last bins in y are not correct.')
