def plot2d(histogram, plot_fn_base, width=1200, height=900):
canvas = ROOT.TCanvas('c1', 'c1', width, height)
ROOT.gStyle.SetOptStat(0)
histogram.Draw('col text')
canvas.SetLogy()
canvas.SaveAs('%s.png' % plot_fn_base)
canvas.SaveAs('%s.pdf' % plot_fn_base)
del canvas
def plot(input_files, output_files, title, legend):
handle_lhe = Handle('LHEEventProduct')
label_lhe = ('externalLHEProducer')
bins = array.array('f', [1. * i for i in range(0, 201)])
h = ROOT.TH1F(legend, legend, len(bins) - 1, bins)
for input_file in input_files:
events = Events(input_file)
for event in events:
event.getByLabel(label_lhe, handle_lhe)
lhe = handle_lhe.product()
invmass = []
for status, pdg, mom in zip(lhe.hepeup().ISTUP,
lhe.hepeup().IDUP,
lhe.hepeup().PUP):
if status == 1 and abs(pdg) in [11, 13, 15]:
l = ROOT.TLorentzVector(mom.x[0], mom.x[1], mom.x[2],
mom.x[3])
invmass.append(l)
if len(invmass) == 2:
h.Fill((invmass[0] + invmass[1]).M())
else:
raise RuntimeError(
'Did not find exactly 2 but %d LHE leptons in %s' %
(len(invmass), input_file))
c = ROOT.TCanvas('c', 'c')
c.SetLogy()
c.SetGrid()
h.SetTitle(title)
h.GetXaxis().SetTitle('m_ll [GeV]')
h.GetYaxis().SetTitle('# events')
h.Draw()
for output_file in output_files:
c.SaveAs(output_file)
del h
del c
def plot(input_files, output_files, title, expected_neff, mode):
histogram_dict = {}
for sample_name, sample_entry in input_files.items():
if not hdfs.isfile(sample_entry['input']):
logging.error('Could not find file {}'.format(sample_entry['input']))
continue
root_file = ROOT.TFile.Open(sample_entry['input'], 'read')
logging.debug('Opened file {}'.format(sample_entry['input']))
root_directories = list(filter(
lambda root_dir: root_dir != None, [
root_file.Get(os.path.join(key.GetName(), mode, 'genEvt')) \
for key in root_file.GetListOfKeys() if key.GetClassName() == 'TDirectoryFile'
]
))
if len(root_directories) != 1:
raise RuntimeError('Expected single directory in %s' % sample_entry['input'])
root_dir = root_directories[0]
histogram_dirs = [
root_dir.Get(key.GetName()) \
for key in root_dir.GetListOfKeys() if key.GetClassName() == 'TDirectoryFile'
]
if len(histogram_dirs) != 1:
raise RuntimeError(
'Expected single directory containing lumiScale histograms in %s' % sample_entry['input']
)
histogram_dir = histogram_dirs[0]
histograms = [
key.GetName() for key in histogram_dir.GetListOfKeys() \
if key.GetClassName().startswith('TH1') and 'lumiScale' in key.GetName()
]
for histogram_name_actual in histograms:
histogram_name = histogram_name_actual.replace('_lumiScale', '').replace('CMS_ttHl_', '') \
if histogram_name_actual != 'lumiScale' else 'central'
histogram = histogram_dir.Get(histogram_name_actual).Clone()
histogram.SetDirectory(0)
if histogram.GetEntries() != sample_entry['nentries'] and mode == 'unbiased':
raise RuntimeError('Expected {} entries from {} in file {}, but got {} entries'.format(
sample_entry['nentries'], histogram_name, sample_entry['input'], histogram.GetEntries(),
))
if histogram_name not in histogram_dict:
histogram_dict[histogram_name] = {
'histogram' : histogram,
'nentries' : histogram.GetEntries(),
'nfiles' : 1,
}
else:
histogram_dict[histogram_name]['histogram'].Add(histogram)
histogram_dict[histogram_name]['nentries'] += histogram.GetEntries()
histogram_dict[histogram_name]['nfiles'] += 1
root_file.Close()
if not histogram_dict:
logging.error('Could not find histograms for samples {}'.format(', '.join(list(input_files.keys()))))
return
if len(set(histogram_dict[histogram_name]['nfiles'] for histogram_name in histogram_dict)) != 1:
raise RuntimeError(
'Inconsistent number of files found for samples %s' % ', '.join(list(input_files.keys()))
)
if len(set(histogram_dict[histogram_name]['nentries'] for histogram_name in histogram_dict)) != 1:
raise RuntimeError(
'Inconsistent number of entries found in samples %s' % ', '.join(list(input_files.keys()))
)
min_y = -1
max_y = -1
nentries = -1
for histograms in histogram_dict.values():
histogram = histograms['histogram']
y_content = histogram.GetBinContent(1)
y_error = histogram.GetBinError(1)
y_down = y_content - y_error
y_up = y_content + y_error
if min_y < 0:
min_y = y_down
if max_y < 0:
max_y = y_up
if y_down < min_y:
min_y = y_down
if y_up > max_y:
max_y = y_up
if nentries < 0:
nentries = histograms['nentries']
else:
assert(nentries == histograms['nentries'])
if not (y_down < expected_neff < y_up) and mode == 'unbiased':
logging.warning(
"Effective event count {} not within {} +- {}".format(expected_neff, y_content, y_error)
)
if mode == 'unbiased':
min_y = min(min_y, expected_neff)
max_y = max(max_y, expected_neff)
diff = 0.2 * (max_y - min_y)
min_y -= diff
max_y += diff
canvas = ROOT.TCanvas('c', 'c', 1200, 900)
canvas.SetGrid()
ROOT.gStyle.SetOptStat(0)
legend = ROOT.TLegend(0.1, 0.7, 0.48, 0.9)
legend.SetHeader('N_{eff} (%d entries)' % nentries)
expected_histogram = None
line_width = 3
marker_style = 20
fill_style = 4000
lines = []
for idx, histogram_name in enumerate(sorted(histogram_dict.keys())):
histogram = histogram_dict[histogram_name]['histogram']
color = 2 + idx
histogram.SetTitle(title)
histogram.SetAxisRange(min_y, max_y, "Y")
histogram.SetLineColor(color)
histogram.SetMarkerColor(color)
histogram.SetLineWidth(line_width)
histogram.SetMarkerStyle(marker_style)
histogram.SetFillStyle(fill_style)
histogram.Draw("l e1%s" % (" same" if idx > 0 else ""))
y_content = histogram.GetBinContent(1)
y_error = histogram.GetBinError(1)
y_up = y_content + y_error
y_down = y_content - y_error
bin_width = histogram.GetBinWidth(1)
bin_center = histogram.GetBinCenter(1)
line_min_x = bin_center - bin_width / 4
line_max_x = bin_center + bin_width / 4
line_down = ROOT.TLine(line_min_x, y_down, line_max_x, y_down)
line_down.SetLineColor(color)
line_down.SetLineWidth(line_width)
line_down.Draw()
lines.append(line_down)
line_up = ROOT.TLine(line_min_x, y_up, line_max_x, y_up)
line_up.SetLineColor(color)
line_up.SetLineWidth(line_width)
line_up.Draw()
lines.append(line_up)
sig_digits = max(8 - int(math.ceil(math.log10(y_content))), 1) if y_content > 0. else 1
leg_pattern = '%s (%.{}f #pm %.{}f)'.format(sig_digits, sig_digits)
leg_name = leg_pattern % (histogram_name, y_content, y_error)
legend.AddEntry(histogram, leg_name)
logging.debug(
'Effective event count for the sys unc option {} is {} +- {}'.format(
histogram_name, y_content, y_error
)
)
if not expected_histogram and mode == 'unbiased':
expected_histogram = histogram.Clone()
expected_histogram.Reset()
expected_histogram.SetBinContent(1, expected_neff)
expected_histogram.SetBinError(1, 0)
expected_histogram.SetLineColor(ROOT.kBlack)
expected_histogram.SetMarkerColor(ROOT.kBlack)
expected_histogram.SetLineWidth(line_width)
expected_histogram.SetMarkerStyle(marker_style)
expected_histogram.SetLineStyle(9)
expected_histogram.SetFillStyle(fill_style)
if expected_histogram:
logging.debug('Expecting {} events'.format(expected_neff))
expected_histogram.Draw("e2 same")
legend.AddEntry(expected_histogram, 'expected (%.1f)' % expected_neff)
legend.Draw()
for output_file in output_files:
canvas.SaveAs(output_file)
canvas.Close()
legend.Delete()
if expected_histogram:
expected_histogram.Delete()
for histogram_name in histogram_dict:
histogram_dict[histogram_name]['histogram'].Delete()
for line in lines:
line.Delete()
for idx in range(1, nof_files + 1):
if idx in blacklist:
continue
input_file = os.path.join(base_path, '%04d' % (idx // 1000),
'tree_%d.root' % idx)
output_file = os.path.join(current_subfolder,
'histogram_%d.root' % idx)
file_exists = project(input_file, output_file, binning)
if file_exists:
output_files.append(output_file)
hadd(output_files, final_output_file)
root_file = ROOT.TFile.Open(final_output_file, 'read')
for binning_key in binning:
canvas = ROOT.TCanvas('c', 'c', 1200, 900)
histogram = root_file.Get(binning_key)
canvas.SetLogy()
if binning_key == 'LHE_HT':
canvas.SetLogx()
histogram.SetTitle(sample_name)
histogram.SetXTitle(binning_key)
histogram.Draw()
for extension in ['pdf', 'png']:
plot_file = os.path.join(
output_plot_dir,
'%s_%s.%s' % (sample_name, binning_key, extension))
canvas.SaveAs(plot_file)
del histogram
del canvas
output_file_ptr.write('{:.6e}'.format(most_frequent_weight))
logging.info("Wrote output file: {}".format(output_file))
if plot_files:
has_neg_weights = any(weight_freq[0] < 0
for weight_freq in weights_by_frequency)
binning_max = MAX_CUTOFF * most_frequent_weight
binning_min = -MAX_CUTOFF * most_frequent_weight
if not has_neg_weights:
binning_min = max(0., binning_min)
binning = array.array(
'f', list(np.linspace(binning_min, binning_max, PLOT_BINS + 1)))
histogram = ROOT.TH1D(GENWEIGHT_NAME, GENWEIGHT_NAME, PLOT_BINS, binning)
for weight_freq in weights_by_frequency:
histogram.Fill(weight_freq[0], weight_freq[1])
histogram.GetXaxis().SetRange(0, histogram.GetNbinsX() + 1)
for plot_file in plot_files:
title = os.path.splitext(os.path.basename(plot_file))[0]
histogram.SetTitle(title)
canvas = ROOT.TCanvas()
canvas.SetCanvasSize(1000, 800)
canvas.SetLogy(True)
histogram.Draw('hist')
canvas.SaveAs(plot_file)
canvas.Close()
del canvas
del histogram
