def DoSignalHeightScan(self, heights=None, hits_per_height=300000):
gc.disable()
starttime = datetime.today()
# ROOT Logfile:
# path = "MC/Performance_Results/"+str(starttime)
path = "MC/Performance_Results/_" + str(
self.minimum_statistics) + "_" + str(
self.binning) + "_" + str(hits_per_height) + "_"
os.makedirs(path)
rootfile = TFile(path + '/MCPerformanceLog.root', 'RECREATE')
LogTree = TTree('LogTree', 'MC Log Tree')
RealSignalAmplitude = array('f', [0])
Repetition = array('i', [0])
TrueNPeaks = array('i', [0])
Ninjas = array('i', [0])
Ghosts = array('i', [0])
Minimas = array('i', [0])
RecSA_Quantiles = array('f', [0])
RecSA_MinMax = array('f', [0])
LogTree.Branch('RealSignalAmplitude', RealSignalAmplitude,
'RealSignalAmplitude/F')
LogTree.Branch('Repetition', Repetition, 'Repetition/I')
LogTree.Branch('TrueNPeaks', TrueNPeaks, 'TrueNPeaks/I')
LogTree.Branch('Ninjas', Ninjas, 'Ninjas/I')
LogTree.Branch('Ghosts', Ghosts, 'Ghosts/I')
LogTree.Branch('Minimas', Minimas, 'Minimas/I')
LogTree.Branch('RecSA_Quantiles', RecSA_Quantiles, 'RecSA_Quantiles/F')
LogTree.Branch('RecSA_MinMax', RecSA_MinMax, 'RecSA_MinMax/F')
# copy Config files:
shutil.copy("Configuration/MonteCarloConfig.cfg",
path + "/MonteCarloConfig.cfg")
shutil.copy("Configuration/AnalysisConfig.cfg",
path + "/AnalysisConfig.cfg")
if heights == None:
heights = [
0.0, 0.05, 0.08, 0.1, 0.125, 0.15, 0.175, 0.2, 0.3, 0.5, 0.8,
1.0
]
# infofile:
infofile = open(path + "/info.txt", "w")
infofile.write("DoSignalHeightScan\n\n")
infofile.write("Timestamp: " + str(starttime) + "\n\n")
infofile.write("Number of Repetitions for each Amplitude: " +
str(self.tries) + "\n")
infofile.write("Number of different Amplitudes: " +
str(len(heights)) + "\n")
infofile.write("Hits per Amplitude: " +
str(hits_per_height) + "\n")
infofile.write("Quantiles: " +
str(self.min_percent) + "/" + str(self.max_percent) +
"\n")
infofile.write("Binning: " +
str(self.binning) + "\n")
infofile.write("Minimum Statistics: " +
str(self.minimum_statistics) + "\n")
infofile.write("Extrema Configuration: " +
self.extremaconfiguration)
success_prob = []
ghost_prob = []
cycle_nr = 0
cycles = self.tries * len(heights)
for height in heights: # add more statistics for each height, not just one try..
fails = 0
tot_ghosts = 0
peaks_generated = 0
for repetition in range(self.tries):
cycle_nr += 1
print("\n{0}th repetition with Signal height set to: {1}\n".
format(repetition, height))
run_object = MCRun(validate=False,
verbose=self.verbose,
run_number=364)
run_object.MCAttributes['PeakHeight'] = height
run_object.SetNumberOfHits(hits_per_height)
print("newAnalysis = Analysis(run_object)")
newAnalysis = Analysis(run_object, verbose=self.verbose)
print("newAnalysis.FindMaxima()")
newAnalysis.FindMaxima(
binning=self.binning,
minimum_bincontent=self.minimum_statistics)
print("newAnalysis.FindMinima()")
newAnalysis.FindMinima(
binning=self.binning,
minimum_bincontent=self.minimum_statistics)
npeaks = newAnalysis.ExtremeAnalysis.ExtremaResults[
'TrueNPeaks']
ninjas = newAnalysis.ExtremeAnalysis.ExtremaResults['Ninjas']
ghosts = newAnalysis.ExtremeAnalysis.ExtremaResults['Ghosts']
maxima = newAnalysis.ExtremeAnalysis.ExtremaResults[
'FoundMaxima']
minima = newAnalysis.ExtremeAnalysis.ExtremaResults[
'FoundMinima']
# Reconstruct Signal Amplitude:
if len(maxima) * len(minima) > 0:
maxbin = newAnalysis.ExtremeAnalysis.Pad.GetBinByCoordinates(
*(maxima[0]))
maxbin.FitLandau()
minbin = newAnalysis.ExtremeAnalysis.Pad.GetBinByCoordinates(
*(minima[0]))
minbin.FitLandau()
rec_sa_minmax = maxbin.Fit['MPV'] / minbin.Fit['MPV'] - 1.
else:
rec_sa_minmax = -99
q = array('d', [
1. * self.min_percent / 100., 1. * self.max_percent / 100.
])
y = array('d', [0, 0])
newAnalysis.ExtremeAnalysis.CreateMeanSignalHistogram()
newAnalysis.ExtremeAnalysis.MeanSignalHisto.GetQuantiles(
2, y, q)
rec_sa_quantiles = y[1] / y[0] - 1.
# Fill ROOT file:
RealSignalAmplitude[0] = height
Repetition[0] = repetition
TrueNPeaks[0] = npeaks
Ninjas[0] = ninjas
Ghosts[0] = ghosts
Minimas[0] = len(minima)
RecSA_Quantiles[0] = rec_sa_quantiles
RecSA_MinMax[0] = rec_sa_minmax
LogTree.Fill()
assert (npeaks > 0), 'no peak in MC created'
peaks_generated += npeaks
fails += ninjas
tot_ghosts += ghosts
# self.AddAnalysis(newAnalysis)
del newAnalysis
del run_object
elapsed_time = datetime.today() - starttime
estimated_time = elapsed_time / cycle_nr * cycles
remaining_time = estimated_time - elapsed_time
print("\n\nAPPROXIMATED TIME LEFT: " + str(remaining_time) +
"\n")
success = 1. * (peaks_generated - fails) / peaks_generated
ghost = 4. * ghosts / self.tries
success_prob.append(success)
ghost_prob.append(ghost)
print("Write ROOT-file")
rootfile.Write()
rootfile.Close()
print("ROOT File written. Write infofile")
infofile.write("\nTotal Time elapsed: " +
str(datetime.today() - starttime))
infofile.close()
print("infofile written")
# canvas = ROOT.TCanvas('canvas', 'canvas') # HERE IT CRASHES DUE TO MEMORY PROBLEMS
# canvas.cd()
# graph1 = ROOT.TGraph()
# graph1.SetNameTitle('graph1', 'success')
# graph1.SaveAs(path+"/SuccessGraph.root")
# graph2 = ROOT.TGraph()
# graph2.SetNameTitle('graph2', 'ghosts')
# graph2.SaveAs(path+"/GhostsGraph.root")
# for i in range(len(heights)):
# graph1.SetPoint(i, heights[i], success_prob[i])
# graph2.SetPoint(i, heights[i], ghost_prob[i])
# graph1.Draw('ALP*')
# graph2.Draw('SAME LP*')
# self.SavePlots("PerformanceResult", "png", path+"/")
answer = input('Wanna crash?')
ROOT.gDirectory.GetList().ls()
ROOT.gROOT.GetListOfFiles().ls()
if answer == 'yes':
gc.collect()
