def frame(step):
""" Creates a frame of a given frame (step) number. """
# Show some information about how far we are with rendering
curr_time = step / eval(SETTINGS.NumberFrames) * eval(SETTINGS.FrameTime)
logger.info(" @Time: %.3fs, Step: %d", curr_time, step)
# Getting the total number of frames, see the configuration file
nframes = eval(SETTINGS.NumberFrames)
# Calculates rotation positions
angle_per_frame = math.pi * 2 / nframes # Calculates how much the angle needs to move per frame
angle = angle_per_frame * step # Calculates the angle of the frame
# Gets locations
x = math.cos(angle) * 20
z = math.sin(angle) * 20
# Create objects
sphere = Sphere([6, 2, -2], 3, models.default_sphere_model)
cylinder = Cylinder([-6, -1, 4], [-6, 7, 4], 3, models.default_sphere_model)
legend_1 = legend([-15, 0, 0], 5)
camera = Camera('location', [x, 8, z], 'look_at', [0, 0, 0])
# Return the Scene object containing all objects for rendering
return Scene(camera,
objects=[sphere, models.default_light, models.checkered_ground, cylinder] + legend_1)
def __init__(self,
curve=None,
title='',
subtitle='',
xlabel='',
ylabel='',
scale='1',
xlim=None,
ylim=None):
if curve is None:
self.curves = []
else:
self.curves = [curve]
self.title = title
self.subtitle = subtitle
self.xlabel = xlabel
self.ylabel = ylabel
self.scale = str(scale)
# TODO Extras
self.xlim = xlim # TODO Add error handling
self.ylim = xlim
self.legend = legend(0, 0, anchor='top right')
def plot_hists_dict(hist_dict, setNames=True, **kwargs):
"""
Draws the contents of a dictionary of hists. Styling will be done
automatically. Any additional kwargs are passed to the plot_hists and legend
methods.
Args:
hist_dict: a str, Hist dictionary of the histograms to be drawn.
Keywords:
setNames: a bool to specify whether the keys of the dict will be used as
titles in the legend.
Returns: a handle to the new canvas
"""
items = hist_dict.items()
for hn, h in items:
h.SetName(hn)
if setNames:
h.SetTitle(hn)
hists = [x[1] for x in items]
names = [x[0] for x in items]
ColorStyleGen.style_hists(hists)
canv = plot_hists(hists, **kwargs)
leg = legend(hists, styles=["f", "f"], **kwargs)
canv.LEGEND = leg
return canv
def plot_fit_shapes(var, fitConf, hData, fit_result, lumi, shapes):
tdrstyle()
canv = TCanvas("c", "c")
canv.SetWindowSize(1000, 1000)
canv.SetCanvasSize(1000, 1000)
outfile_name = "fit_plots/" + var.shortName + "_shapes_" + fitConf.name
shapes[-1].SetTitle("QCD")
canv = plot_hists(
canv,
{"shapes": shapes},
x_label=var.displayName,
# title=process,
max_bin_mult=1.4,
)
# Draws the lumi box
# lumibox = lumi_textbox(lumi_iso[channel])
# Draw the legend
leg = legend(shapes, styles=["l"], width=0.2, text_size=0.015) # <<< need to reverse MC order here, mc3 is top-most
lumibox = lumi_textbox(lumi)
leg.Draw()
# print hNonQCD.Integral(), hData.Integral(), hQCD.Integral(), hTotal.Integral(), hQCDp.Integral(), hQCDm.Integral()
canv.Update()
canv.SaveAs(outfile_name + ".png")
canv.SaveAs(outfile_name + ".pdf")
canv.Draw()
return canv
def plot_all_fit_shapes(var, fitConf, hData, fit_result, lumi):
tdrstyle()
canvases = []
infile = "fits/" + var.shortName + "_fit_" + fitConf.name + ".root"
f = TFile(infile)
print fitConf.name
outfile_name = "fit_plots/" + var.shortName + "_all_shapes_" + fitConf.name
cst = TCanvas("Histogram_" + fitConf.name, fitConf.name, 10, 10, 1000, 1000)
hNonQCD = TH1D(f.Get(var.shortName + "__nonqcd"))
hNonQCD.SetTitle("Non-QCD")
hNonQCD.SetLineColor(kRed)
hNonQCD.Scale(1 / hNonQCD.Integral())
hWJets = TH1D(f.Get(var.shortName + "__wjets"))
hWJets.SetTitle("W+Jets")
hWJets.SetLineColor(kGreen + 4)
hWJets.Scale(1 / hWJets.Integral())
hData.SetNameTitle(var.shortName + "__DATA", "Data")
hData.SetMarkerStyle(20)
hData.Scale(1 / hData.Integral())
# print "data integral: ",hData.Integral()
hQCD = f.Get(var.shortName + "__qcd")
hQCD.SetNameTitle(var.shortName + "__qcd", "QCD")
hQCD.SetLineColor(kGray)
hQCD.Scale(1 / hQCD.Integral())
max_bin = hQCD.GetMaximum() * 1.3
hData.SetAxisRange(0, max_bin, "Y")
hData.GetXaxis().SetTitle(var.displayName)
# hTotal.Draw("")
title = fit_result.getTitle()
hData.SetMarkerStyle(20)
hData.Draw("E1")
hQCD.Draw("same")
hNonQCD.Draw("same")
hWJets.Draw("same")
# hData.SetTitle("QCD fit, "+title)
hData.Draw("E1 same")
lumibox = lumi_textbox(lumi)
leg = legend([hData, hQCD, hNonQCD, hWJets], styles=["p", "l"], width=0.2)
leg.Draw()
# print hNonQCD.Integral(), hData.Integral(), hQCD.Integral(), hTotal.Integral(), hQCDp.Integral(), hQCDm.Integral()
cst.Update()
cst.SaveAs(outfile_name + ".png")
cst.SaveAs(outfile_name + ".pdf")
cst.Draw()
return cst
def frame(step):
""" Creates a frame of 4 cones, 4 boxes, 1 sphere and a legend """
# Define textures for different models
sphere_model = Texture(Pigment(
'color',
[1, 0, 1],
), Finish('reflection', 0.5))
box_model = Texture(Pigment(
'color',
[0, 1, 1],
), Finish('reflection', 0))
cone_model = Texture(Pigment(
'color',
[1, 0, 1],
), Finish('reflection', 0))
# Create objects
sphere = Sphere([0, 0, 0], 3, sphere_model)
box_1 = Box([-5, -5, -4], [-3, 5, 4], box_model)
box_2 = Box([3, -5, -4], [5, 5, 4], box_model)
box_3 = Box([-5, 4, -4], [5, 6, 4], box_model)
box_4 = Box([-5, -5, -4], [5, -3, 4], box_model)
cone_1 = Cone([0, 6, 0], 3, [0, 10, 0], 0, cone_model)
cone_2 = Cone([0, -6, 0], 3, [0, -10, 0], 0, cone_model)
cone_3 = Cone([-5, 0, 0], 3, [-9, 0, 0], 0, cone_model)
cone_4 = Cone([5, 0, 0], 3, [9, 0, 0], 0, cone_model)
light_1 = LightSource([0, 10, -25], 'color', [1, 1, 1])
light_2 = LightSource([0, 8, -7], 'color', [1, 1, 1])
xyz_legend = legend([-15, 0, 0], 5)
camera = Camera('location', [0, 7, -30], 'look_at', [0, 2, 1])
# Return the Scene object for rendering
return Scene(camera,
objects=[
sphere, box_1, box_2, box_3, box_4, cone_1, cone_2,
cone_3, cone_4, light_1, light_2
] + xyz_legend)
Styling.mc_style(h_mc3, "W1Jets_exclusive")
Styling.data_style(h_d1)
#Create the canvas
canv = ROOT.TCanvas("c", "c")
canv.SetWindowSize(500, 500)
canv.SetCanvasSize(600, 600)
#!!!!LOOK HERE!!!!!
#----
#Draw the stacked histograms
#----
stacks_d = OrderedDict() #<<< need to use OrderedDict to have data drawn last (dict does not preserve order)
stacks_d["mc"] = [h_mc1, h_mc2, h_mc3] # <<< order is important here, mc1 is bottom-most
stacks_d["data"] = [h_d1]
stacks = plot_hists_stacked(
canv,
stacks_d,
x_label="variable x [GeV]",
y_label="",
do_log_y=False
)
#Draw the legend
import legend
leg = legend.legend(
[h_d1, h_mc3, h_mc2, h_mc1], # <<< need to reverse MC order here
styles=["p", "f"],
width=0.2
)
def plot_fit2(fit_templates_file, qcd_result, other_results, priors, extra={}):
templates_file = fit_templates_file.replace("reversecut", "nocut").replace("qcdcut", "nocut")
var = templates_file.split("/")[1].split("__")[0]
jt = templates_file.split("/")[1].split("__")[1]
channel = templates_file.split("/")[1].split("__")[2]
reg = fit_templates_file.split("/")[1].split("__")[3]
added = fit_templates_file.split("/")[1].split("__")[4]
f = TFile(templates_file)
print "using templates", templates_file
extra_string = ""
extra_id = ""
for k, v in extra.items():
extra_id += k + "_" + v
vark = (
k.replace("varMC_QCDMC", ", QCD from MC")
.replace("varMC_", ", MC antiiso")
.replace("isovar", "Anti-iso range")
)
extra_string += "%s %s" % (vark, v)
display_var = var.replace("qcd_mva", "QCD BDT")
print channel, jt
tdrstyle()
gStyle.SetOptTitle(1)
canvases = []
# infile = "fits/"+var.shortName+"_fit_"+jt+"_"+channel+"_"+cutid+".root"
# infile2 = "templates/"+var.shortName+"_templates_"+jt+"_"+channel+"_"+cutid+".root"
# cutinfo = "_".join(cutid.split("_")[1:])
# infile3 = "templates/"+var.shortName+"_templates_"+jt+"_"+channel+"_nocut_"+cutinfo+".root"
# outfile_wd = "fits/"+var+"_fit_"+jt+"_"+channel+".root"
# outf = TFile(outfile_wd, "recreate")
# outf.cd()
try:
os.mkdir("fit_plots")
except Exception as e:
logging.warning(str(e))
outfile_name = (
"fit_plots/" + var + "_Fit_" + jt + "_" + channel + "_" + reg + "_" + added + "_" + extra_id
) # +"_"+cutid
outfile_name = outfile_name.replace(".root_", "")
hData = TH1D(f.Get(var + "__DATA"))
hQCD = TH1D(f.Get(var + "__QCD"))
# print qcd_result
print "initial"
print "QCD:", hQCD.Integral()
print "Data:", hData.Integral()
hQCD.Scale(qcd_result[0])
hTotal = TH1D(hQCD)
"""cut = get_cut(var, channel)
for bin in range(hData.GetNbinsX()+2):
if abs(hData.GetBinLowEdge(bin) - cut) < 1e-8:
low = bin
high = hData.GetNbinsX()+1
"""
other_templates = {}
# print other_results
first = True
for temp in other_results.keys():
other_templates[temp] = TH1D(f.Get(var + "__" + temp))
# print other_templates[temp].Integral(), other_results[temp][0], priors[temp]
sf = math.e ** (other_results[temp][0] * priors[temp])
print temp, other_templates[temp].Integral()
other_templates[temp].Scale(sf)
if first:
hNonQCD = TH1D(f.Get(var + "__" + temp))
first = False
else:
hNonQCD.Add(other_templates[temp])
hTotal.Add(other_templates[temp])
# print "SF", temp, other_templates[temp].Integral()
chi2 = hData.Chi2Test(hTotal, "UW CHI2/NDF")
print "chi2", chi2
cst = TCanvas("Histogram_", "histo", 10, 10, 1000, 1000)
cst.SetTopMargin(0.9)
# print infile
# print f
# print f.Get(var.shortName+"__nonqcd")
hNonQCD.SetTitle("Non-QCD")
hNonQCD.SetLineColor(kRed)
hData.SetMarkerStyle(20)
hQCD.SetNameTitle("QCD", "QCD")
hQCD.SetLineColor(kYellow)
"""hQCDp=TH1D(hQCD)
hQCDp.Scale(QCDRATE_UP/QCDRATE)
hQCDm=TH1D(hQCD)
hQCDm.Scale(QCDRATE_DOWN/QCDRATE)
hQCDp.SetName(var.shortName+"__qcd_plus")
hQCDp.Write()
hQCDm.SetName(var.shortName+"__qcd_minus")
hQCDm.Write()
hQCDp.SetLineColor(kGreen)
hQCDp.SetTitle("QCD #pm 1 #sigma")
hQCDm.SetLineColor(kGreen)
hQCDm.SetTitle("QCD #pm 1 #sigma")
hTotal=TH1D(hNonQCD)
hTotal.Add(hQCD)
"""
hTotal.SetLineColor(kBlue)
hTotal.SetTitle("Fitted total")
max_bin = hData.GetBinContent(hData.GetMaximumBin()) * 1.5
hData.SetAxisRange(0, max_bin, "Y")
hData.GetXaxis().SetTitle(display_var)
# hTotal.Draw("")
title = jt + ", " + channel
hData.SetTitle(title)
hData.SetMarkerStyle(20)
# print "data", hData.Integral()
hData.Draw("E1")
# print "data", hData.Integral()
# hNonQCDp.Draw("same")
hQCD.SetLineColor(kGreen + 2)
hQCD.SetLineWidth(3)
hQCD.Draw("same")
hNonQCD.Draw("same hist")
# hNonQCDm.Draw("same")
# hQCDp.Draw("same")
# hQCDm.Draw("same")
hTotal.Draw("same hist")
# hData.SetTitle("QCD fit, "+title)
hData.Draw("E1 same")
lumi = 15311
if channel == "ele":
lumi = 16934
lumibox = lumi_textbox(lumi)
hDataCopy = TH1D(hData)
hDataCopy.SetTitle("Data")
leg = legend([hDataCopy, hQCD, hNonQCD, hTotal], styles=["p", "l"], width=0.2) # hQCDp, # hNonQCDp,
# outf.Close()
cst.Update()
cst.SaveAs(outfile_name + ".png")
cst.SaveAs(outfile_name + ".pdf")
# cst.Draw()
print "final"
print "QCD:", hQCD.Integral()
print "Data:", hData.Integral()
print "other", hNonQCD.Integral()
print "total", hTotal.Integral()
"""
def plot_fit(var, fitConf, hData, fit_result, lumi):
tdrstyle()
canvases = []
infile = "fits/" + var.shortName + "_fit_" + fitConf.name + ".root"
outfile_wd = "fits/" + var.shortName + "_fit_" + fitConf.name + "_withData.root"
f = TFile(infile)
f2 = TFile(outfile_wd, "recreate")
f2.cd()
try:
os.mkdir("fit_plots")
except Exception as e:
logging.warning(str(e))
outfile_name = "fit_plots/" + var.shortName + "_Fit_" + fitConf.name
# print fit_result
QCDRATE = fit_result.qcd
# QCDRATE_UP = fit_result.qcd + fit_result.qcd_uncert
# QCDRATE_DOWN = fit_result.qcd - fit_result.qcd_uncert
QCDRATE_UP = (
fit_result.qcd
* math.e ** ((fit_result.result["qcd"]["beta_signal"][0][0] + fit_result.result["qcd"]["beta_signal"][0][1]))
/ math.e ** ((fit_result.result["qcd"]["beta_signal"][0][0]))
)
QCDRATE_DOWN = (
fit_result.qcd
* math.e ** ((fit_result.result["qcd"]["beta_signal"][0][0] - fit_result.result["qcd"]["beta_signal"][0][1]))
/ math.e ** ((fit_result.result["qcd"]["beta_signal"][0][0]))
)
# NONQCDRATE = fit_result.nonqcd
# NONQCDRATE_UP = fit_result.nonqcd + fit_result.nonqcd_uncert
# NONQCDRATE_DOWN = fit_result.nonqcd - fit_result.nonqcd_uncert
NONQCDRATE = math.e ** ((fit_result.result["qcd"]["nonqcd_rate"][0][0]) * fit_result.coeff["nonqcd"])
NONQCDRATE_UP = math.e ** (
(fit_result.result["qcd"]["nonqcd_rate"][0][0] + fit_result.result["qcd"]["nonqcd_rate"][0][1])
* fit_result.coeff["nonqcd"]
)
NONQCDRATE_DOWN = math.e ** (
(fit_result.result["qcd"]["nonqcd_rate"][0][0] - fit_result.result["qcd"]["nonqcd_rate"][0][1])
* fit_result.coeff["nonqcd"]
)
WJETS = math.e ** ((fit_result.result["qcd"]["wjets_rate"][0][0]) * fit_result.coeff["wjets"])
WJETS_UP = math.e ** (
(fit_result.result["qcd"]["wjets_rate"][0][0] + fit_result.result["qcd"]["wjets_rate"][0][1])
* fit_result.coeff["wjets"]
)
WJETS_DOWN = math.e ** (
(fit_result.result["qcd"]["wjets_rate"][0][0] - fit_result.result["qcd"]["wjets_rate"][0][1])
* fit_result.coeff["wjets"]
)
print "Results with fit uncertainty included"
print "QCD: ", QCDRATE, "+", (QCDRATE_UP - QCDRATE), "-", (QCDRATE - QCDRATE_DOWN)
print "Non-QCD: ", NONQCDRATE, "+", (NONQCDRATE_UP - NONQCDRATE), "-", (NONQCDRATE - NONQCDRATE_DOWN)
print "W+jets: ", WJETS, "+", (WJETS_UP - WJETS), "-", (WJETS - WJETS_DOWN)
# WJETS = fit_result.wjets
# WJETS_UP = fit_result.wjets + fit_result.wjets_uncert
# WJETS_DOWN = fit_result.wjets - fit_result.wjets_uncert
cst = TCanvas("Histogram_" + fitConf.name, fitConf.name, 10, 10, 1000, 1000)
hNonQCD = TH1D(f.Get(var.shortName + "__nonqcd"))
hNonQCD.Write()
hNonQCD.SetTitle("Non-QCD")
hNonQCD.SetLineColor(kRed)
hNonQCDp = TH1D(hNonQCD)
hNonQCDp.SetName(var.shortName + "__nonqcd__plus")
hNonQCDp.Scale(NONQCDRATE_UP / NONQCDRATE)
hNonQCDp.Write()
hNonQCDm = TH1D(hNonQCD)
hNonQCDm.Scale(NONQCDRATE_DOWN / NONQCDRATE)
hNonQCDm.SetName(var.shortName + "__nonqcd__minus")
hNonQCDm.Write()
hNonQCDp.SetLineColor(kOrange)
hNonQCDp.SetTitle("Non-QCD #pm 1 #sigma")
hNonQCDm.SetLineColor(kOrange)
hNonQCDm.SetTitle("non-QCD - 1 sigma")
hWJets = TH1D(f.Get(var.shortName + "__wjets"))
hWJets.Write()
hWJets.SetTitle("W+Jets")
hWJets.SetLineColor(kGreen + 4)
hWJetsp = TH1D(hWJets)
hWJetsp.SetName(var.shortName + "__wjets_plus")
hWJetsm = TH1D(hWJets)
hWJetsm.SetName(var.shortName + "__wjets_minus")
if WJETS > 0:
hWJetsp.Scale(WJETS_UP / WJETS)
hWJetsm.Scale(WJETS_DOWN / WJETS)
hWJetsp.Write()
hWJetsm.Write()
hWJetsp.SetLineColor(kGreen + 8)
hWJetsp.SetTitle("W+Jets #pm 1 #sigma")
hWJetsm.SetLineColor(kGreen + 8)
hWJetsm.SetTitle("W+Jets - 1 sigma")
hData.SetNameTitle(var.shortName + "__DATA", "Data")
hData.Write()
hData.SetMarkerStyle(20)
# print "data integral: ",hData.Integral()
hQCD = f.Get(var.shortName + "__qcd")
hQCD.Write()
hQCD.SetNameTitle(var.shortName + "__qcd", "QCD")
hQCD.SetLineColor(kYellow)
hQCDp = TH1D(hQCD)
hQCDp.Scale(QCDRATE_UP / QCDRATE)
hQCDm = TH1D(hQCD)
hQCDm.Scale(QCDRATE_DOWN / QCDRATE)
hQCDp.SetName(var.shortName + "__qcd_plus")
hQCDp.Write()
hQCDm.SetName(var.shortName + "__qcd_minus")
hQCDm.Write()
hQCDp.SetLineColor(kGreen)
hQCDp.SetTitle("QCD #pm 1 #sigma")
hQCDm.SetLineColor(kGreen)
hQCDm.SetTitle("QCD #pm 1 #sigma")
hTotal = TH1D(hNonQCD)
hTotal.Add(hQCD)
hTotal.Add(hWJets)
hTotal.SetLineColor(kBlue)
hTotal.SetTitle("Fitted total")
max_bin = hData.GetMaximum() * 1.6
hData.SetAxisRange(0, max_bin, "Y")
hData.GetXaxis().SetTitle(var.displayName)
# hTotal.Draw("")
title = fit_result.getTitle()
hData.SetMarkerStyle(20)
hData.Draw("E1")
hNonQCDp.Draw("same")
hQCD.Draw("same")
hNonQCD.Draw("same")
hNonQCDm.Draw("same")
hQCDp.Draw("same")
hQCDm.Draw("same")
hWJets.Draw("same")
hWJetsp.Draw("same")
hWJetsm.Draw("same")
hTotal.Draw("same")
# hData.SetTitle("QCD fit, "+title)
hData.Draw("E1 same")
lumibox = lumi_textbox(lumi)
leg = legend([hData, hQCD, hQCDp, hNonQCD, hNonQCDp, hWJets, hWJetsp, hTotal], styles=["p", "l"], width=0.2)
leg.Draw()
# print hNonQCD.Integral(), hData.Integral(), hQCD.Integral(), hTotal.Integral(), hQCDp.Integral(), hQCDm.Integral()
cst.Update()
cst.SaveAs(outfile_name + ".png")
cst.SaveAs(outfile_name + ".pdf")
cst.Draw()
return cst
Styling.mc_style(h_mc2, "T_t")
Styling.mc_style(h_mc3, "W1Jets_exclusive")
Styling.data_style(h_d1)
#Create the canvas
canv = ROOT.TCanvas("c", "c")
canv.SetWindowSize(500, 500)
canv.SetCanvasSize(600, 600)
#!!!!LOOK HERE!!!!!
#----
#Draw the stacked histograms
#----
stacks_d = OrderedDict(
) #<<< need to use OrderedDict to have data drawn last (dict does not preserve order)
stacks_d["mc"] = [h_mc1, h_mc2,
h_mc3] # <<< order is important here, mc1 is bottom-most
stacks_d["data"] = [h_d1]
stacks = plot_hists_stacked(canv,
stacks_d,
x_label="variable x [GeV]",
y_label="",
do_log_y=False)
#Draw the legend
import legend
leg = legend.legend(
[h_d1, h_mc3, h_mc2, h_mc1], # <<< need to reverse MC order here
styles=["p", "f"],
width=0.2)
