def cutflowplot(config_list):
##Import math for s/sqrt(s+b)
from math import sqrt
##Name of output root file
root_file = "/cutflow_" + channel + ".root"
##Construct list of weights = [no cut, cut1, cut1+cut2, ..., cut1+...+cutN]
cut_strings = [parameter_info[param][4] for param in cut_info[0 if weight == "(njetspt30==0)" else 1][channel].keys()]
cut_values, cut_side = [[entry[index] for entry in cut_info[0 if weight == "(njetspt30==0)" else 1][channel].values()] for index in [0,1]]
weight_list = ["1"] + ["*".join(cut_strings[index].format(cut = cut, side = side) for index, (cut, side) in enumerate(zip(cut_values[:1 + index2], cut_side[:1 + index2]))) for index2 in range(len(cut_strings))]
##Using harry.py the magician to get root files, read out for each process the total number of events and save them in hists as list for each weight
hists = []
configs = []
for index, weights in enumerate(weight_list):
config = configmaster.ConfigMaster(*base_config(x[0], weight_input = weights))
config.change_config_info("plot_modules", "ExportRoot")
config.pop(["www", "www_nodate", "legend_markers"])
config.change_config_info("filename", str(index))
configs.append(config.return_config())
higgsplot.HiggsPlotter(list_of_config_dicts=configs, n_plots = len(configs), batch = "rwthcondor")
for index in range(len(weight_list)):
root = ROOT.TFile.Open(flavio_path + "/VisibleMass" + weight_name + str(index) + ".root")
hists.append([root.Get(process).Integral() for process in sample_list])
hists[index].append(hists[index][0]/sqrt(hists[index][0] + sum(hists[index][1:])))
for filetype in [".root", ".json"]:
for index in range(len(weight_list)):
os.remove(flavio_path + "/VisibleMass" + weight_name + str(index) + filetype)
##Use ROOT to fill hists in a new root file in which each process is a hist with the Number of events for each weight applied
if not os.path.exists(flavio_path):
os.mkdir(flavio_path)
if os.path.isfile(flavio_path + root_file):
os.remove(flavio_path + root_file)
for index1, process in enumerate(sample_list + ["soverb"]):
output = ROOT.TFile(flavio_path + root_file, "UPDATE")
root_hist = ROOT.TH1F(process, process, len(hists), 0, len(hists))
for index2, hist in enumerate(hists):
root_hist.SetBinContent(index2+1, hist[index1])
output.Write()
output.Close()
##Write config for cutflow plots
config = configmaster.ConfigMaster(base_config(x[0], no_plot = True)[0])
config.add_config_info(cutflow_config(x[0])[0], 0)
config.add_config_info(cutflow_config(x[0])[1], 5)
config.pop(["directories", "x_expressions", "x_bins"])
config.change_config_info(["directories", "x_expressions", "x_tick_labels"], [flavio_path, sample_list + ["soverb"], ["no cuts"] + cut_strings])
config.print_config()
config_list.append(config.return_config())
return config_list
def main():
###Call and get parser arguments
args = parser()
###Define global constants
global_constants(args)
##Define libary with all analysis modules of flavio
Analysismodule_libary = {
Analysismodule.control_plot: controlplot,
Analysismodule.effiency_plot: efficiencyplot,
Analysismodule.shape_plot: shapeplot,
Analysismodule.cut_optimization: cutoptimization,
Analysismodule.nminus1_plot: nminus1plot,
Analysismodule.cutflow_plot: cutflowplot,
Analysismodule.limit_plot: limitplot
}
###Write config using your desired analysis function
config_list = []
config_list = Analysismodule_libary[args.analysis](config_list)
###Call MVP harry.py to get your job done
higgsplot.HiggsPlotter(list_of_config_dicts=config_list, n_plots = len(config_list), batch = "rwthcondor" if len(config_list) != 1 else None)
if len(config_list) != 1: subprocess.Popen("cd $CMSSW_BASE/src/websync/; webplot.py {channel} -r -c".format(channel = channel), shell = True)
def plots(channel, method):
##Plot postfit shapes
config_list = []
for category in categories:
for shape in ["_prefit", "_postfit"]:
config = {
"files":
output_dir + channel + "/" + "postfit_" + method + ".root",
"folders":
category[1] + shape,
"x_expressions":
sorted(backgrounds.keys(), reverse=True) + ["TotalBkg"] +
data.keys() + ["Z" + channel.upper()],
"stacks": ["bkg" for bkg in backgrounds.keys()] +
["unc", "data", "sig", "ratio1", "ratio2"],
"markers": ["HIST" for hist in backgrounds.keys()] +
["E2", "E", "LINE", "E", "E2"],
"formats": ["png", "pdf"],
"nicks":
sorted(backgrounds.values(), reverse=True) + ["Uncertainty"] +
data.values() + ["z" + channel],
"colors":
sorted(backgrounds.values(), reverse=True) + ["kBlack"] +
+data.values() + ["z" + channel, "kBlack", "kBlack"],
"legend": [0.65, 0.5, 0.95, 0.85],
"y_subplot_lims": [0.5, 1.5],
"analysis_modules": ["Ratio", "MaskHistograms"],
"mask_histogram_nicks": ["data", "ratio1"],
"mask_above_reference_nick":
"z" + channel,
"mask_above_reference_value":
50 if category[0] == 0 else 10,
"ratio_denominator_nicks": ["Uncertainty", "Uncertainty"],
"ratio_numerator_nicks": ["data", "Uncertainty"],
"ratio_result_nicks": ["ratio1", "ratio2"],
"x_label":
"BDT score" if method == "BDT" else "m_{vis}",
"filename":
method + "_" + category[1] + shape,
"lumis": [35.87],
"energies": [13],
"year":
"2016",
"scale_factors":
["1"
for proc in backgrounds.keys() + data.keys()] + ["1", "10"],
"title": {
"em": "e#mu",
"et": "e#tau",
"mt": "#mu#tau"
}[channel],
"www":
channel + "/Limits",
"www_nodate":
True,
}
config_list.append(config)
higgsplot.HiggsPlotter(list_of_config_dicts=config_list)
def prefit_postfit_plots(self, datacards_cbs, datacards_postfit_shapes, plotting_args=None, n_processes=1, *args):
if plotting_args is None:
plotting_args = {}
plot_configs = []
bkg_plotting_order = ["ZTT", "ZLL", "ZL", "ZJ", "TTJ", "TT", "VV", "WJ", "W", "QCD"]
for level in ["prefit", "postfit"]:
for index, (fit_type, datacards_postfit_shapes_dict) in enumerate(datacards_postfit_shapes.iteritems()):
if (index == 0) or (level == "postfit"):
for datacard, postfit_shapes in datacards_postfit_shapes_dict.iteritems():
for category in datacards_cbs[datacard].cp().bin_set():
bkg_processes = datacards_cbs[datacard].cp().bin([category]).backgrounds().process_set()
bkg_processes.sort(key=lambda process: bkg_plotting_order.index(process) if process in bkg_plotting_order else len(bkg_plotting_order))
config = {}
config["files"] = [postfit_shapes]
config["folders"] = [category+"_"+level]
config["x_expressions"] = ["TotalSig"] + bkg_processes + ["TotalBkg", "data_obs"]
config["nicks"] = ["TotalSig"] + bkg_processes + ["TotalBkg", "data_obs"]
config["stacks"] = ["sig_bkg"] + (["sig_bkg"]*len(bkg_processes)) + ["bkg_unc", "data"]
config["labels"] = [label.lower() for label in ["TotalSig"] + bkg_processes + ["TotalBkg", "data_obs"]]
config["colors"] = [color.lower() for color in ["TotalSig"] + bkg_processes + ["TotalBkg", "data_obs"]]
config["markers"] = ["LINE"] + (["HIST"]*len(bkg_processes)) + ["E2", "E"]
config["legend_markers"] = ["L"] + (["F"]*len(bkg_processes)) + ["F", "ELP"]
config["y_lims"] = [0.0]
config["y_rel_lims"] = [0.0, 1.4]
config["legend"] = [0.23, 0.73, 0.9, 0.89]
config["legend_cols"] = 3
config["title"] = "channel_"+(category.split("_")[0])
config["output_dir"] = os.path.join(os.path.dirname(datacard), "plots")
config["filename"] = level+("_"+fit_type if level == "postfit" else "")+"_"+category
if not "NormalizeByBinWidth" in config.get("analysis_modules", []):
config.setdefault("analysis_modules", []).append("NormalizeByBinWidth")
config["y_label"] = "Entries / bin"
if plotting_args.get("ratio", False):
if not "Ratio" in config.get("analysis_modules", []):
config.setdefault("analysis_modules", []).append("Ratio")
config.setdefault("ratio_numerator_nicks", []).extend(["noplot_TotalBkg", "noplot_TotalBkg TotalSig", "data_obs"])
config.setdefault("ratio_denominator_nicks", []).extend(["noplot_TotalBkg"] * 3)
config.setdefault("ratio_result_nicks", []).extend(["ratio_unc", "ratio_sig", "ratio"])
config.setdefault("colors", []).extend(["totalbkg", "totalsig", "#000000"])
config.setdefault("markers", []).extend(["E2", "LINE", "E"])
config.setdefault("legend_markers", []).extend(["F", "L", "ELP"])
config.setdefault("labels", []).extend([""] * 3)
config["legend"] = [0.7, 0.5, 0.95, 0.92]
plot_configs.append(config)
# create result plots HarryPlotter
return higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs, list_of_args_strings=[plotting_args.get("args", "")], n_processes=n_processes)
def pull_plots(self,
datacards_postfit_shapes,
s_fit_only=False,
plotting_args=None,
n_processes=1):
if plotting_args is None:
plotting_args = {}
plot_configs = []
for index, (fit_type, datacards_postfit_shapes_dict) in enumerate(
datacards_postfit_shapes.iteritems()):
if (index == 0):
for datacard, postfit_shapes in datacards_postfit_shapes_dict.iteritems(
):
config = {}
config["files"] = [
os.path.join(os.path.dirname(datacard), "mlfit.root")
]
config["input_modules"] = ["InputRootSimple"]
config["root_names"] = [
"fit_s", "fit_b", "nuisances_prefit"
]
if s_fit_only:
config["root_names"] = ["fit_s", "nuisances_prefit"]
config["fit_s_only"] = [True]
config["analysis_modules"] = ["ComputePullValues"]
config["nicks_blacklist"] = ["graph_b"]
config["fit_poi"] = plotting_args.get("fit_poi", "r")
config["left_pad_margin"] = 0.40
config["labels"] = ["prefit", "S+B model"]
config["markers"] = ["L2", "P"]
config["fill_styles"] = [3001, 0]
config["legend"] = [0.75, 0.8]
config["legend_markers"] = ["LF", "LP"]
config["x_lims"] = [-5.0, 5.0]
config["x_label"] = "Pull values"
config["output_dir"] = os.path.join(
os.path.dirname(datacard), "plots")
config["filename"] = "pulls"
plot_configs.append(config)
# create result plots HarryPlotter
return higgsplot.HiggsPlotter(
list_of_config_dicts=plot_configs,
list_of_args_strings=[plotting_args.get("args", "")],
n_processes=n_processes)
def get_hists(parameter, base_config, binning, channel):
##Get shapes for variables
config = base_config + {
"directories": "/net/scratch_cms3b/{}/artus/AllSamples/merged/".format(os.environ["USER"]),
"x_expressions": parameter,
"plot_modules": ["ExportRoot"],
"output_dir": plot_dir + channel,
"x_bins": binning,
"file_mode": "UPDATE",
"filename": parameter,
}
config.pop("legend_markers")
higgsplot.HiggsPlotter(list_of_config_dicts=[config])
if args.y_quantity is not None: y_bins = [x for x in range(int(float(args.y_range[0])*100),int(float(args.y_range[1])*100)+1, int((float(args.y_range[1])-float(args.y_range[0]))*4))] config["y_expressions"] = [args.y_quantity] config["y_bins"] = [" ".join([str(x/100.0) for x in y_bins])] variables.append(args.y_quantity) if args.z_quantity is not None: z_bins = [x for x in range(int(float(args.z_range[0])*100),int(int(float(args.z_range[1])*100)+1), int((float(args.z_range[1])-float(args.z_range[0]))*4))] config["z_expressions"] = [args.z_quantity] config["z_bins"] = [" ".join([str(x/100.0) for x in z_bins])] variables.append(args.z_quantity) current_cuts = [] current_max = 0 real_values = [] for i in range(3): if i == 0: higgsplot.HiggsPlotter(list_of_config_dicts=[config]) else: b_max_index = current_cuts[0]-1 b_min = x_bins[b_max_index-3] if b_max_index >=3 else x_bins[0] b_max = x_bins[b_max_index+3] if (len(x_bins) - b_max_index) >3 else x_bins[-1] b_min = int(b_min*100) b_max = int(b_max*100) step = (b_max-b_min)/10 if (b_max-b_min)/10 >= 1 else 1 x_bins = [x_bins[0]]+[x/100.0 for x in range(b_min, b_max, int(step))]+[x_bins[-1]] config["x_bins"] = [" ".join([str(x/100.0) for x in x_bins])] print config["x_bins"] if args.y_quantity is not None: b_max_index = current_cuts[1]-1 b_min = y_bins[b_max_index-3] if b_max_index >=3 else y_bins[0] b_max = y_bins[b_max_index+3] if (len(y_bins) - b_max_index) >3 else y_bins[-1] b_min = int(b_min*100)
)
configs.extend(
selection_check_lMu_PPV.return_json_with_changed_x_and_weight(
x_expressions=["leadingMuon_PhotonsFromFirstPVPtFlow_full"]))
selection_check_lMu_PPV_peak = selection_check_lMu_PPV.clone(
name="selection_check_lMu_PPV_peak",
legend=data_embedded_mirrored_random_legend_upper_right,
title="#DeltaR in [0.21,0.22]; peak region (#pm 10 GeV)",
#title = "DeltaR 0.21-0.22 & peak region"
)
configs.extend(
selection_check_lMu_PPV_peak.return_json_with_changed_x_and_weight(
x_expressions=["leadingMuon_PhotonsFromFirstPVPtFlow_peak"]))
selection_check_lMu_PPV_sideband = selection_check_lMu_PPV.clone(
name="selection_check_lMu_PPV_sideband",
legend=data_embedded_mirrored_random_legend_upper_right,
title="#DeltaR in [0.21,0.22]; sideband region (#pm 10 GeV)",
#title = "DeltaR 0.21-0.22 & sideband region"
)
configs.extend(
selection_check_lMu_PPV_sideband.return_json_with_changed_x_and_weight(
x_expressions=["leadingMuon_PhotonsFromFirstPVPtFlow_sideband"]))
higgs_plotter = higgsplot.HiggsPlotter(list_of_config_dicts=configs,
list_of_args_strings=[""])
def main(args):
signal_names = ["ggh", "qqh"]
signal = ["ggH125", "qqH125"]
if args.emb and args.ff:
bkg_processes_names = ["emb", "zll", "ttl", "vvl", "fakes"]
bkg_processes = ["EMB", "ZL", "TTL", "VVL", "jetFakes"]
elif args.emb:
bkg_processes_names = [
"emb", "zll", "zj", "ttl", "ttj", "vvl", "vvj", "w", "qcd"
]
bkg_processes = [
"EMB", "ZL", "ZJ", "TTL", "TTJ", "VVL", "VVJ", "W", "QCD"
]
elif args.ff:
bkg_processes_names = [
"ztt", "zll", "ttt", "ttl", "vvt", "vvl", "fakes"
] # enforced by HarryPlotter
bkg_processes = ["ZTT", "ZL", "TTT", "TTL", "VVT", "VVL", "jetFakes"]
else:
bkg_processes_names = [
"ztt", "zll", "zj", "ttl", "ttt", "ttj", "vvl", "vvt", "vvj", "w",
"qcd"
]
bkg_processes = [
"ZTT", "ZL", "ZJ", "TTL", "TTT", "TTJ", "VVL", "VVT", "VVJ", "W",
"QCD"
]
channels = args.channels
analysis = args.analysis
era = args.era
if args.lumi is None:
if "2016" in era:
lumi = 35.9
elif "2017" in era:
lumi = 41.5
elif "2018" in era:
lumi = 59.7
else:
lumi = args.lumi
output_dir = args.output_dir
y_log = args.y_log
mass = args.mass
variables = args.variables
categories = [c for c in args.categories
] if args.categories is not None else None
if "all" in variables:
variables = [
"mt_1", "mt_2", "pt_1", "pt_2", "eta_1", "eta_2", "m_vis", "ptvis",
"npv", "njets", "nbtag", "jpt_1", "jpt_2", "jeta_1", "jeta_2",
"met", "mjj", "dijetpt", "pZetaMissVis", "m_1", "m_2",
"decayMode_1", "decayMode_2", "iso_1", "iso_2", "rho", "mt_tot",
"d0_1", "d0_2", "dZ_1", "dZ_2"
]
if args.www:
config_template['www'] = ''
if args.www_dir:
config_template['www_dir'] = args.www_dir
if args.www_no_overwrite:
config_template['www_no_overwrite'] = True
config_template['no_overwrite'] = True
if args.no_overwrite:
config_template['no_overwrite'] = True
if args.redo_cache:
config_template['redo_cache'] = True
if args.filename_prefix != '':
args.filename_prefix = args.filename_prefix if args.filename_prefix[
0] == '_' else '_' + args.filename_prefix
# if args.log_level == 'debug':
# config['log_level'] =
configs = []
for channel in channels:
if args.categories is None:
categories = [channel + "_" + v for v in variables]
shapes = args.shapes if args.shapes is not None else "shapes_{}.root".format(
channel)
if "em" in channel:
bkg_processes = [
p for p in bkg_processes if p not in ["ZJ", "TTJ", "VVJ"]
]
bkg_processes_names = [
p for p in bkg_processes_names
if p not in ["zj", "ttj", "vvj"]
]
if len(variables) == 1:
variables = [variables[0]] * len(categories)
for variable, category in zip(variables, categories):
config = deepcopy(config_template)
if args.chi2:
config["analysis_modules"].append("AddHistograms")
config["analysis_modules"].append("Chi2Test")
config["add_nicks"] = [" ".join(bkg_processes_names)]
config["add_result_nicks"] = ["tot_background_noplot"]
config["chi2test_nicks"] = ["tot_background_noplot data"]
config["chi2test_compare"] = ["UU CHI2/NDF"]
config["files"] = [shapes]
config["lumis"] = [lumi]
config["year"] = era.strip("Run")
config["output_dir"] = output_dir + "/" + channel
config["y_log"] = True if ((variable in logvars)
or y_log) else False
config["y_rel_lims"] = [5, 500] if (variable
in logvars) else [0.9, 1.5]
config["markers"] = ["HIST"] * len(
bkg_processes_names) + ["LINE"] * len(signal_names) + ["P"] + [
"E2"
] + ["LINE"] * len(signal_names) + ["P"]
config["legend_markers"] = ["F"] * (
len(bkg_processes_names)) + ["LINE"] * len(signal_names) + [
"ELP"
] + ["E2"] + ["L"] * len(signal_names) + ["P"]
config["labels"] = bkg_processes_names + signal_names + [
"data"
] + [""] * len(signal_names) + config["labels"]
config["colors"] = bkg_processes_names + signal_names + [
"data"
] + ["#B7B7B7"] + signal_names + ["#000000"]
config["nicks"] = bkg_processes_names + signal_names + ["data"]
config["x_expressions"] = [
"#" + "#".join([
channel, category, process, analysis, era, variable, mass
]) + "#" for process in bkg_processes + signal
] + [
"#" + "#".join([
channel, category, "data_obs", analysis, era, variable,
mass
]) + "#"
]
if args.emb == False:
config["filename"] = "_".join([
channel, category, analysis, era, variable, mass
]) + args.filename_prefix
else:
config["filename"] = "_".join([
channel, category, analysis, era, variable, mass, "emb"
]) + args.filename_prefix
if args.comparison:
config["filename"] = "_".join([
channel, category, analysis, era, variable, mass,
"comparison"
]) + args.filename_prefix
if args.ff:
config["filename"] = config["filename"] + "_ff"
if not args.x_label == None:
config["x_label"] = args.x_label
else:
config["x_label"] = "_".join([channel, variable])
config["title"] = "_".join(["channel", channel])
config["stacks"] = ["mc"
] * len(bkg_processes_names) + signal_names + [
"data"
] + [x + "Ratio"
for x in signal_names] + config["stacks"]
config["ratio_denominator_nicks"] = [
" ".join(bkg_processes_names)
] * (2 + len(signal_names))
config["ratio_numerator_nicks"] = [
" ".join(bkg_processes_names)
] + [" ".join(bkg_processes_names + [signal_names[0]])] + [
" ".join(bkg_processes_names + [signal_names[1]])
] + ["data"]
config["ratio_result_nicks"] = ["bkg_ratio"] + [
x + "_ratio" for x in signal_names
] + ["data_ratio"]
if args.comparison:
config["markers"] = ["HIST"] + ["LINE"] + ["HIST"] * (
len(bkg_processes_names) - 1) + ["EX0"] + ["E2", "EX0"
] + ["EX0"]
config["legend_markers"] = ["F"] + ["L"] + ["F"] * (
len(bkg_processes_names) - 1) + ["PEX0"] + ["F", "PEX0"
] + ["PEX0"]
config["labels"] = [bkg_processes_names[0]] + [
"Z #rightarrow #tau#tau (simulation)"
] + bkg_processes_names[1:] + ["data"] + [
"Bkg. (embedded) unc.", "Obs./Bkg. (simulation)",
"Obs./Bkg. (embedded)"
]
config["colors"] = [bkg_processes_names[0]] + [
"#C70039"
] + bkg_processes_names[1:] + ["data"] + [
"#B7B7B7", " #C70039 ", "#000000"
]
config["subplot_legend"] = [0.195, 0.77, 0.92, 0.93],
config["subplot_legend_cols"] = 3
config["subplot_legend_fontsize"] = 0.06
config["y_subplot_label"] = ""
config["legend_fontsize"] = 0.044
bkg_processes_names_noplot = [
"ztt_noplot", "ttt_noplot", "vvt_noplot"
] + [
x + "_noplot"
for x in bkg_processes_names if not "emb" in x
]
config[
"nicks"] = bkg_processes_names + bkg_processes_names_noplot + [
"data"
]
config["analysis_modules"].append("AddHistograms")
config["add_nicks"] = [" ".join(bkg_processes_names_noplot)]
config["add_result_nicks"] = ["ztt"]
config["add_result_position"] = 1
config["x_expressions"] = [
"#" + "#".join([
channel, category, process, analysis, era, variable,
mass
]) + "#" for process in bkg_processes
] + [
"#" + "#".join([
channel, category, process, analysis, era, variable,
mass
]) + "#" for process in ["ZTT", "TTT", "VVT"] +
[x for x in bkg_processes if x != "EMB"]
] + [
"#" + "#".join([
channel, category, "data_obs", analysis, era, variable,
mass
]) + "#"
]
config["stacks"] = [
"mc"
] + ["ztt"] + ["mc"] * (len(bkg_processes_names) - 1) + [
"data"
] + config_template["stacks"] + ["ratio_data_zttmc"]
config["ratio_denominator_nicks"] = [
" ".join(bkg_processes_names)
] + [" ".join(bkg_processes_names_noplot)
] + [" ".join(bkg_processes_names)]
config["ratio_numerator_nicks"] = [
" ".join(bkg_processes_names), "data", "data"
]
config["ratio_result_nicks"] = [
"ratio_Bkg", "ratio_Data_to_MC", "ratio_Data"
]
configs.append(config)
import pprint as pp
pp.pprint(configs)
higgsplot.HiggsPlotter(list_of_config_dicts=configs,
list_of_args_strings=[args.additional_arguments],
n_processes=args.num_processes)
def main(args):
bkg_processes_names = ["ztt", "zl", "zj", "wj", "ttt", "ttj",
"qcd", "vv", "ewk"] # enforced by HarryPlotter
bkg_processes = ["ZTT", "ZL", "ZJ", "W", "TTT", "TTJ",
"QCD", "VV", "EWKZ"] # names in ROOT file
signal_processes_names = ["htt125", "ggh125",
"qqh125"] # enforced by HarryPlotter
signal_processes = ["HTT", "qqH", "ggH"] # name in ROOT file
if args.emb and args.ff:
bkg_processes_names = ["emb", "zll", "ttl", "vvl", "fakes"]
bkg_processes = ["EMB", "ZL", "TTL", "VVL", "jetFakesEMB"]
elif args.emb:
bkg_processes_names = ["emb", "zll", "zj", "ttl", "ttj", "vvl", "vvj", "w", "qcd"]
bkg_processes = ["EMB", "ZL", "ZJ", "TTL", "TTJ", "VVL", "VVJ", "W", "QCD"]
elif args.ff:
bkg_processes_names = ["ztt", "zll", "ttt", "ttl", "vvt", "vvl", "fakes"] # enforced by HarryPlotter
bkg_processes = ["ZTT", "ZL", "TTT", "TTL", "VVT", "VVL", "jetFakes"]
else:
bkg_processes_names = ["ztt", "zll", "zj", "ttl", "ttt", "ttj", "vvl", "vvt", "vvj", "w", "qcd"]
bkg_processes = ["ZTT", "ZL", "ZJ", "TTL", "TTT", "TTJ", "VVL", "VVT", "VVJ", "W", "QCD"]
if args.bg_proc:
bkg_processes = args.bg_proc
bkg_processes_names = [i.lower() for i in bkg_processes]
bkg_processes_names = ['fakes' if i.startswith('jetfakes') else i for i in bkg_processes_names]
bkg_processes_names = ['qcd' if i.startswith('qcd') else i for i in bkg_processes_names]
bkg_processes_names = ['w' if i.startswith('w') else i for i in bkg_processes_names]
bkg_processes_names = ["zll" if i == 'zl' else i for i in bkg_processes_names]
logger.info('bkg_processes:' + ' '.join(bkg_processes))
logger.info('bkg_processes_names:' + ' '.bkg_processes_names)
if args.no_signal:
signal_processes_names = []
signal_processes = []
scale_signal = args.scale_signal
config_template["scales"] = scale_signal
config_template["scale_nicks"] = signal_processes_names
if scale_signal != 1:
postfix = "_" + str(scale_signal)
else:
postfix = ""
signal_processes_labels = [
label + postfix for label in signal_processes_names
]
if args.www:
config_template['www'] = ''
if args.www_dir:
config_template['www_dir'] = args.www_dir
if args.www_no_overwrite:
config_template['www_no_overwrite'] = True
config_template['no_overwrite'] = True
if args.no_overwrite:
config_template['no_overwrite'] = True
if args.redo_cache:
config_template['redo_cache'] = True
if args.filename_prefix != '':
args.filename_prefix = args.filename_prefix if args.filename_prefix[0] == '_' else '_' + args.filename_prefix
categories = args.categories
channel = args.channel
analysis = args.analysis
era = args.era
if "2016" in era:
config_template['lumi'] = [35.9]
config_template['year'] = '2016'
elif "2017" in era:
config_template['lumi'] = [41.5]
config_template['year'] = '2017'
elif "2018" in era:
config_template['lumi'] = [59.7]
config_template['year'] = '2018'
mass = args.mass
variables = args.variables
configs = []
for category in categories:
for variable in variables:
config = deepcopy(config_template)
config["files"] = [args.shapes]
config["markers"] = ["HIST"] * len(bkg_processes_names) + [
"LINE"
] * len(signal_processes_names) + ["E"] + config["markers"]
config["legend_markers"] = ["F"] * len(
bkg_processes_names) + ["L"] * len(signal_processes_names) + [
"ELP"
] + config["legend_markers"]
config[
"labels"] = bkg_processes_names + signal_processes_labels + [
"data"
] + config["labels"]
config["colors"] = bkg_processes_names + signal_processes_names + [
"data"
] + config["colors"]
config["nicks"] = bkg_processes_names + signal_processes_names + [
"data"
]
config["x_expressions"] = [
"#" + "#".join([
channel, category, process, analysis, era, variable, mass
]) + "#" for process in bkg_processes + signal_processes
] + [
"#" + "#".join([
channel, category, "data_obs", analysis, era, variable,
mass
]) + "#"
]
config["filename"] = "_".join(
[channel, category, analysis, era, variable, mass]) + args.filename_prefix
if not args.x_label == None:
config["x_label"] = args.x_label
else:
config["x_label"] = "_".join([channel, variable])
config["title"] = "_".join(["channel", channel])
config["stacks"] = ["mc"] * len(
bkg_processes_names) + signal_processes_names + [
"data"
] + config["stacks"]
config["ratio_denominator_nicks"] = [
" ".join(bkg_processes_names)
] * 2
config["ratio_numerator_nicks"] = [
" ".join(bkg_processes_names), "data"
]
configs.append(config)
import pprint as pp
pp.pprint(configs[0])
higgsplot.HiggsPlotter(
list_of_config_dicts=configs,
list_of_args_strings=[""],
n_processes=args.num_processes)
#!/usr/bin/env python
# -*- coding: utf-8 -*-
import logging
import Artus.Utility.logger as logger
log = logging.getLogger(__name__)
import sys
import HiggsAnalysis.KITHiggsToTauTau.plotting.higgsplot as higgsplot
if __name__ == "__main__":
"""This script executes all steps necessary to create a plot."""
if len(sys.argv) == 1:
sys.argv.append("-h")
higgsplot.HiggsPlotter()
def main(args):
bkg_processes_names = [
"ztt", "zl", "zj", "wj", "ttt", "ttj", "qcd", "vv", "ewk"
] # enforced by HarryPlotter
bkg_processes = [
"ZTT", "ZL", "ZJ", "W", "TTT", "TTJ", "QCD", "VV", "EWKZ"
] # names in ROOT file
signal_processes_names = ["htt125", "ggh125",
"qqh125"] # enforced by HarryPlotter
signal_processes = ["HTT", "qqH", "ggH"] # name in ROOT file
scale_signal = args.scale_signal
config_template["scales"] = scale_signal
config_template["scale_nicks"] = signal_processes_names
if scale_signal != 1:
postfix = "_" + str(scale_signal)
else:
postfix = ""
signal_processes_labels = [
label + postfix for label in signal_processes_names
]
categories = args.categories
channel = args.channel
analysis = args.analysis
era = args.era
mass = args.mass
variables = args.variables
configs = []
for category in categories:
for variable in variables:
config = deepcopy(config_template)
config["files"] = [args.shapes]
config["markers"] = ["HIST"] * len(bkg_processes_names) + [
"LINE"
] * len(signal_processes_names) + ["E"] + config["markers"]
config["legend_markers"] = ["F"] * len(bkg_processes_names) + [
"L"
] * len(signal_processes_names) + ["ELP"
] + config["legend_markers"]
config[
"labels"] = bkg_processes_names + signal_processes_labels + [
"data"
] + config["labels"]
config["colors"] = bkg_processes_names + signal_processes_names + [
"data"
] + config["colors"]
config["nicks"] = bkg_processes_names + signal_processes_names + [
"data"
]
config["x_expressions"] = [
"#" + "#".join([
channel, category, process, analysis, era, variable, mass
]) + "#" for process in bkg_processes + signal_processes
] + [
"#" + "#".join([
channel, category, "data_obs", analysis, era, variable,
mass
]) + "#"
]
config["filename"] = "_".join(
[channel, category, analysis, era, variable, mass])
if not args.x_label == None:
config["x_label"] = args.x_label
else:
config["x_label"] = "_".join([channel, variable])
config["title"] = "_".join(["channel", channel])
config["stacks"] = ["mc"] * len(
bkg_processes_names) + signal_processes_names + [
"data"
] + config["stacks"]
config["ratio_denominator_nicks"] = [
" ".join(bkg_processes_names)
] * 2
config["ratio_numerator_nicks"] = [
" ".join(bkg_processes_names), "data"
]
configs.append(config)
for key in configs[0]:
print(key, configs[0][key])
higgsplot.HiggsPlotter(list_of_config_dicts=configs,
list_of_args_strings=[""],
n_processes=args.num_processes)
# delete existing output files
tmp_output_files = list(
set([
os.path.join(config["output_dir"], config["filename"] + ".root")
for config in plot_configs[:args.n_plots[0]]
]))
for output_file_iterator in tmp_output_files:
if os.path.exists(output_file_iterator):
os.remove(output_file_iterator)
log.debug("Removed file \"" + output_file_iterator +
"\" before it is recreated again.")
output_files = list(set(output_files))
# create input histograms with HarryPlotter
higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs,
list_of_args_strings=[args.args],
n_processes=args.n_processes,
n_plots=args.n_plots[0])
if args.n_plots[0] != 0:
tools.parallelize(_call_command,
hadd_commands,
n_processes=args.n_processes)
if args.debug_plots:
debug_plot_configs = []
for output_file in merged_output_files:
debug_plot_configs.extend(plotconfigs.PlotConfigs().all_histograms(
output_file,
plot_config_template={
"markers": ["E"],
"colors": ["#FF0000"]
}))
higgsplot.HiggsPlotter(list_of_config_dicts=debug_plot_configs,
def create_input_root_files(datacards, args):
''' Configuring Harry plotter according to the samples and creating input root files according to the args.'''
plot_configs = []
output_files = []
merged_output_files = []
hadd_commands = []
sample_settings = samples.Samples()
expression_settings = expressions.ExpressionsDict()
binnings_settings = binnings.BinningsDict()
systematics_factory = systematics.SystematicsFactory()
datacards.configs._mapping_process2sample = {
"data_obs": "data",
"EWKZ": "ewkz",
"QCD": "qcd",
"TT": "ttj",
"TTT": "ttt",
"TTJ": "ttjj",
"VV": "vv",
"VVT": "vvt",
"VVJ": "vvj",
"W": "wj",
"ZJ": "zj",
"ZL": "zl",
"ZLL": "zll",
"ZTTPOSPOL": "zttpospol",
"ZTTNEGPOL": "zttnegpol",
"ZTT_GEN_DM_ZERO": "ztt_gen_dm_zero",
"ZTT_GEN_DM_ONE": "ztt_gen_dm_one",
"ZTT_GEN_DM_TWO": "ztt_gen_dm_two",
"ZTT_GEN_DM_TEN": "ztt_gen_dm_ten",
"ZTT_GEN_DM_ELEVEN": "ztt_gen_dm_eleven",
}
for index, (channel,
categories) in enumerate(zip(args.channel, args.categories)):
for category in categories:
datacards_per_channel_category = datacardsbase.Datacards(
cb=datacards.cb.cp().channel([channel]).bin([category]))
higgs_masses = [
mass for mass in datacards_per_channel_category.cb.mass_set()
if mass != "*"
]
output_file = os.path.join(
args.output_dir,
"input/{ANALYSIS}_{CHANNEL}_{BIN}_{ERA}.root".format(
ANALYSIS="ztt", CHANNEL=channel, BIN=category,
ERA="13TeV"))
output_files.append(output_file)
tmp_output_files = []
for shape_systematic, list_of_samples in datacards_per_channel_category.get_samples_per_shape_systematic(
).iteritems():
nominal = (shape_systematic == "nominal")
list_of_samples = [
datacards.configs.process2sample(process)
for process in list_of_samples
]
if ("wj"
in list_of_samples) and not ("qcd" in list_of_samples):
list_of_samples.append("qcd")
elif ("qcd"
in list_of_samples) and not ("wj" in list_of_samples):
list_of_samples.append("wj")
asimov_nicks = []
if args.use_asimov_dataset:
asimov_nicks = [
nick.replace("zttpospol", "zttpospol_noplot").replace(
"zttnegpol", "zttnegpol_noplot")
for nick in list_of_samples
]
if "data" in asimov_nicks:
asimov_nicks.remove("data")
for shift_up in ([True] if nominal else [True, False]):
systematic = "nominal" if nominal else (
shape_systematic + ("Up" if shift_up else "Down"))
log.debug(
"Create inputs for (samples, systematic) = ([\"{samples}\"], {systematic}), (channel, category) = ({channel}, {category})."
.format(samples="\", \"".join(list_of_samples),
channel=channel,
category=category,
systematic=systematic))
tmp_quantity = args.quantity
tmp_omega_version = args.omega_version
if args.fixed_variables == "best_choice":
if channel in ["tt"]:
if category in [
channel + "_" + cat for cat in [
"combined_rho_oneprong",
"combined_oneprong_oneprong"
]
]:
tmp_quantity = "m_vis"
tmp_omega_version = None
elif category in [
channel + "_" + cat
for cat in ["combined_a1_rho"]
]:
tmp_quantity = None
tmp_omega_version = None
elif category in [
channel + "_" + cat for cat in
["combined_a1_a1", "combined_a1_oneprong"]
]:
tmp_quantity = None
tmp_omega_version = "BarSvfitM91"
elif category in [
channel + "_" + cat
for cat in ["combined_rho_rho", "rho"]
]:
tmp_quantity = None
tmp_omega_version = "VisibleSvfit"
elif channel in ["mt", "et"]:
if category in [
channel + "_" + cat
for cat in ["combined_a1_oneprong"]
]:
tmp_quantity = "m_vis"
tmp_omega_version = None
elif category in [
channel + "_" + cat
for cat in ["combined_rho_oneprong"]
]:
tmp_quantity = None
tmp_omega_version = None
elif category in [
channel + "_" + cat for cat in [
"combined_oneprong_oneprong", "a1",
"oneprong"
]
]:
tmp_quantity = None
tmp_omega_version = "BarSvfitM91"
elif category in [
channel + "_" + cat for cat in ["rho"]
]:
tmp_quantity = None
tmp_omega_version = "VisibleSvfit"
elif channel in ["em"]:
if category in [
channel + "_" + cat
for cat in ["combined_oneprong_oneprong"]
]:
tmp_quantity = "m_vis"
tmp_omega_version = None
elif args.fixed_variables == "best_choice_no_svfit":
tmp_quantity = "m_vis"
tmp_omega_version = None
if channel in ["tt", "mt", "et"]:
if category in [
channel + "_" + cat
for cat in ["combined_rho_rho", "rho"]
]:
tmp_quantity = None
tmp_omega_version = "VisibleSvfit"
x_expression = None
if tmp_quantity:
x_expression = tmp_quantity
else:
x_expression = "testZttPol13TeV_" + category
if tmp_omega_version:
x_expression = expression_settings.expressions_dict[
x_expression].replace("BarSvfit",
tmp_omega_version)
x_expression = expression_settings.expressions_dict.get(
x_expression, x_expression)
# prepare plotting configs for retrieving the input histograms
config = sample_settings.get_config(
samples=[
getattr(samples.Samples, sample)
for sample in list_of_samples
],
channel=channel,
category="catZttPol13TeV_" + category,
weight=args.weight,
lumi=args.lumi * 1000,
higgs_masses=higgs_masses,
estimationMethod="new",
polarisation_bias_correction=True,
cut_type="low_mvis_smhtt2016",
exclude_cuts=(["m_vis"]
if x_expression == "m_vis" else []),
no_ewk_samples=args.no_ewk_samples,
no_ewkz_as_dy=True,
asimov_nicks=asimov_nicks)
systematics_settings = systematics_factory.get(
shape_systematic)(config)
# TODO: evaluate shift from datacards_per_channel_category.cb
config = systematics_settings.get_config(shift=(
0.0 if nominal else (1.0 if shift_up else -1.0)))
#config["qcd_subtract_shape"] =[args.qcd_subtract_shapes]
config["x_expressions"] = [
("0" if (("gen_zttpospol" in nick) or
("gen_zttnegpol" in nick)) else x_expression)
for nick in config["nicks"]
]
binnings_key = "binningZttPol13TeV_" + category + "_" + x_expression
if not (binnings_key in binnings_settings.binnings_dict):
binnings_key = "binningZttPol13TeV_" + category + (
("_" + tmp_quantity) if tmp_quantity else "")
if binnings_key in binnings_settings.binnings_dict:
config["x_bins"] = [
("1,-1,1" if
(("gen_zttpospol" in nick) or
("gen_zttnegpol" in nick)) else binnings_key)
for nick in config["nicks"]
]
if args.fixed_binning:
if args.fixed_variables:
if channel == "tt":
config["x_bins"] = [
args.fixed_binning.split(",")[0] +
",-1.0001,1.0001"
for nick in config["nicks"]
]
else:
config["x_bins"] = [
args.fixed_binning
for nick in config["nicks"]
]
else:
config["x_bins"] = [
args.fixed_binning for nick in config["nicks"]
]
config["directories"] = [args.input_dir]
histogram_name_template = "${BIN}/${PROCESS}" if nominal else "${BIN}/${PROCESS}_${SYSTEMATIC}"
config["labels"] = [
histogram_name_template.replace("$", "").format(
PROCESS=datacards.configs.sample2process(
sample.replace("asimov", "data")),
BIN=category,
SYSTEMATIC=systematic)
for sample in config["labels"]
]
tmp_output_file = os.path.join(
args.output_dir,
"input/{ANALYSIS}_{CHANNEL}_{BIN}_{SYSTEMATIC}_{ERA}.root"
.format(ANALYSIS="ztt",
CHANNEL=channel,
BIN=category,
SYSTEMATIC=systematic,
ERA="13TeV"))
tmp_output_files.append(tmp_output_file)
config["output_dir"] = os.path.dirname(tmp_output_file)
config["filename"] = os.path.splitext(
os.path.basename(tmp_output_file))[0]
config["plot_modules"] = ["ExportRoot"]
config["file_mode"] = "UPDATE"
if "legend_markers" in config:
config.pop("legend_markers")
plot_configs.append(config)
hadd_commands.append("hadd -f {DST} {SRC} && rm {SRC}".format(
DST=output_file, SRC=" ".join(tmp_output_files)))
tmp_output_files = list(
set([
os.path.join(config["output_dir"], config["filename"] + ".root")
for config in plot_configs[:args.n_plots[0]]
]))
for output_file in tmp_output_files:
if os.path.exists(output_file):
os.remove(output_file)
log.debug("Removed file \"" + output_file +
"\" before it is recreated again.")
output_files = list(set(output_files))
higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs,
list_of_args_strings=[args.args],
n_processes=args.n_processes,
n_plots=args.n_plots[0],
batch=args.batch)
if args.n_plots[0] != 0:
tools.parallelize(_call_command,
hadd_commands,
n_processes=args.n_processes)
debug_plot_configs = []
for output_file in (output_files):
debug_plot_configs.extend(plotconfigs.PlotConfigs().all_histograms(
output_file,
plot_config_template={
"markers": ["E"],
"colors": ["#FF0000"]
}))
if args.www:
for debug_plot_config in debug_plot_configs:
debug_plot_config["www"] = debug_plot_config["output_dir"].replace(
args.output_dir, args.www)
#higgsplot.HiggsPlotter(list_of_config_dicts=debug_plot_configs, list_of_args_strings=[args.args], n_processes=args.n_processes, n_plots=args.n_plots[0])
return None
exclude_cuts=args.exclude_cuts
)
firsts = [exp_dict.replace_expressions(s) for s in args.first_category]
seconds = [exp_dict.replace_expressions(s) for s in args.second_category]
config["x_expressions"] = ["1*({first})+2*({second})".format(first=" || ".join(firsts), second=" || ".join(seconds))]
#config["x_expressions"] = ["m_sv"]
#config["x_ticks"] = [1,2,3]
config["x_bins"] = ["3,0.5,3.5"]
#config["x_tick_labels"] = args.Ticklabels
config["filename"] = args.filename
config["output_dir"] = args.output_dir
config["directories"] = [args.input_dir]
args.args += " --plot-modules ExportRoot"
#print config
higgsplot.HiggsPlotter(list_of_config_dicts=[config],
list_of_args_strings=[args.args])
names = []
ratio_names = []
for nick in args.Samples:
for mass in args.higgs_masses:
add = nick
if nick in ("ggh", "qqh", "vh"):
add += "%s"%mass
names.append(add)
if nick in args.ratio:
ratio_names.append(add)
#sys.exit()
tfile = ROOT.TFile(os.path.join(args.output_dir, args.filename+".root"), "READ")
#ztt = tfile.Get("ztt")
#zll = tfile.Get("zll")
# delete existing output files
output_files = list(
set([
os.path.join(config["output_dir"],
config["filename"] + ".root")
for config in input_plot_configs[:args.n_plots[0]]
]))
for output_file in output_files:
if os.path.exists(output_file):
os.remove(output_file)
log.debug("Removed file \"" + output_file +
"\" before it is recreated again.")
# create input histograms with HarryPlotter
higgsplot.HiggsPlotter(list_of_config_dicts=input_plot_configs,
n_processes=args.n_processes,
n_plots=args.n_plots[0])
tools.parallelize(_call_command,
hadd_commands,
n_processes=args.n_processes)
# update CombineHarvester with the yields and shapes
datacards.extract_shapes(os.path.join(
args.output_dir, input_root_filename_template.replace("$", "")),
bkg_histogram_name_template,
sig_histogram_name_template,
bkg_syst_histogram_name_template,
sig_syst_histogram_name_template,
update_systematics=False)
# create morphing
def cutoptimization(config_list):
##Information to create weight string for harry plotter
cut_side = ["<" for parameter in x]
cut_values = range(len(cut_side))
cut_strings = [parameter_info[parameter][4] for parameter in x]
for index1 in range(3):
for index2, parameter in enumerate(x):
##Create N-1 weight for Nth parameter if not zero Iteration
if index1 != 0:
weights = "*".join([
cut_strings[index3].format(side=side, cut=value)
for index3, (side,
value) in enumerate(zip(cut_side, cut_values))
if index2 != index3
])
##Create harry plotter config
config = configmaster.ConfigMaster(
*base_config(parameter,
nick_suffix="noplot",
no_plot=True,
weight_input="1" if index1 == 0 else weights))
config.add_config_info(
sumofhists_config([
"zttnoplot zllnoplot ttjnoplot vvnoplot wjnoplot qcdnoplot"
], ["bkg_sum"]), 1)
config.add_config_info(
efficiency_config(
parameter, ["bkg_sum"],
lower_cut=True if cut_side[index2] == "<" else False)[1],
2)
config.pop(["www", "www_nodate", "cms", "extra_text"])
config.change_config_info(["filename", "plot_modules"],
["_" + channel, "ExportRoot"])
##Create config in zero iteration to check on which side the cut should be applied
if (index1 == 0):
config2 = config.copy()
config2.replace("select_lower_values", False)
config2.change_config_info("filename", "_2")
##Harry.py the MVP
higgsplot.HiggsPlotter(
list_of_config_dicts=[config.return_config()] if index1 != 0
else [config.return_config(),
config2.return_config()],
n_plots=2 if index1 == 0 else 1)
##Readout best cut value
cut_file = ROOT.TFile.Open(flavio_path + "/" +
config.return_config()["filename"] +
".root")
histogram = cut_file.Get("sOverSqrtSB")
cut_values[index2] = (histogram.GetXaxis().GetBinCenter(
histogram.GetMaximumBin()))
for filetype in [".json", ".root"]:
os.remove(flavio_path + "/" +
config.return_config()["filename"] + filetype)
##Choose which side to cut on in zero iteration
if index1 == 0:
cut_file2 = ROOT.TFile.Open(
flavio_path + "/" + config2.return_config()["filename"] +
".root")
histogram2 = cut_file2.Get("sOverSqrtSB")
if histogram.GetMaximum() < histogram2.GetMaximum():
cut_values[index2] = histogram2.GetXaxis().GetBinCenter(
histogram2.GetMaximumBin())
cut_side[index2] = ">"
for filetype in [".json", ".root"]:
os.remove(flavio_path + "/" +
config2.return_config()["filename"] + filetype)
##print progress
if index1 == 0:
print "Cut sides which are choosen: {sides}".format(sides=cut_side)
print "Cut values for iteration {index}: {values}".format(
index=index1, values=cut_values)
##Save data in cut.yaml
cut_info = yaml.load(
open(
os.path.abspath(
os.path.expandvars(
"$CMSSW_BASE/src/HiggsAnalysis/KITHiggsToTauTau/python/lfv/cuts.yaml"
)), "r"))
if cut_info == None:
for category in [0 if weight == "(njetspt30==0)" else 1]:
cut_info = {category: {channel: {}}}
for (cut, side, parameter) in zip(cut_values, cut_side, x):
cut_info[category][channel][parameter] = [cut, side]
else:
for category in [0 if weight == "(njetspt30==0)" else 1]:
try:
cut_info[category][channel] = {}
except KeyError:
cut_info[category] = {channel: {}}
for (cut, side, parameter) in zip(cut_values, cut_side, x):
cut_info[category][channel][parameter] = [cut, side]
yaml.dump(
cut_info,
open(
os.path.abspath(
os.path.expandvars(
"$CMSSW_BASE/src/HiggsAnalysis/KITHiggsToTauTau/python/lfv/cuts.yaml"
)), "w"),
default_flow_style=False)
sys.exit()
def main(args):
'''
### Use this for inclusive background process labels (probably obsolete)
if args.emb:
bkg_processes_names = [
"emb", "zll","ttl", "vv","w", "qcd"
]
bkg_processes = ["EMB", "ZL", "TTL", "VV", "W", "QCD"] # names in ROOT file
else:
bkg_processes_names = [
"ztt", "zll", "tt", "vv", "ewk", "w", "qcd"
] # enforced by HarryPlotter
bkg_processes = ["ZTT", "ZL", "TT", "VV", "EWK", "W", "QCD"] # names in ROOT file
'''
### Use this for fake-factor shapes
if args.emb and args.ff:
bkg_processes_names = ["emb", "zll", "ttl", "vvl", "fakes"]
bkg_processes = ["EMB", "ZL", "TTL", "VVL",
"jetFakes"] # names in ROOT file
elif args.emb:
bkg_processes_names = [
"emb", "zll", "zj", "ttl", "ttj", "ewkl", "ewkj", "vvl", "vvj",
"w", "qcd"
]
bkg_processes = [
"EMB", "ZL", "ZJ", "TTL", "TTJ", "EWKL", "EWKJ", "VVL", "VVJ",
"WEMB", "QCDEMB"
] # names in ROOT file
elif args.ff:
bkg_processes_names = [
"ztt", "zll", "ttt", "ttl", "vvt", "vvl", "ewk", "fakes"
] # enforced by HarryPlotter
bkg_processes = [
"ZTT", "ZL", "TTT", "TTL", "VVT", "VVL", "EWK", "jetFakes"
] # names in ROOT file
else:
bkg_processes_names = [
"ztt", "zll", "zj", "tt", "vv", "ewk", "w", "qcd"
] # enforced by HarryPlotter
bkg_processes = ["ZTT", "ZL", "ZJ", "TT", "VV", "EWK", "W",
"QCD"] # names in ROOT file
qcd_scale_factors = {"mt": 1.0, "et": 1.0, "tt": 1.0, "em": 1.0}
channels = args.channels
analysis = args.analysis
era = args.era
output_dir = args.output_dir
y_log = args.y_log
mass = args.mass
variables = args.variables
categories = [c for c in args.categories
] if args.categories is not None else None
if "all" in variables:
variables = [
"mt_1", "mt_2", "pt_1", "pt_2", "eta_1", "eta_2", "m_vis", "ptvis",
"npv", "njets", "nbtag", "jpt_1", "jpt_2", "jeta_1", "jeta_2",
"met", "mjj", "dijetpt", "pZetaMissVis", "m_1", "m_2",
"decayMode_1", "decayMode_2", "iso_1", "iso_2", "rho", "mt_tot",
"d0_1", "d0_2", "dZ_1", "dZ_2"
]
configs = []
for channel in channels:
if args.categories is None:
categories = [channel + "_" + v for v in variables]
shapes = args.shapes if args.shapes is not None else "shapes_{}.root".format(
channel)
if "em" in channel:
bkg_processes = [
p for p in bkg_processes
if p not in ["ZJ", "TTJ", "VVJ", "EWKJ", "WEMB"]
]
bkg_processes_names = [
p for p in bkg_processes_names
if p not in ["zj", "ttj", "vvj", "ewkj"]
]
if not "W" in bkg_processes:
bkg_processes.insert(bkg_processes_names.index("w"), "W")
elif "tt" in channel and not args.ff:
bkg_processes = [
p for p in bkg_processes if p not in ["WEMB", "QCDEMB"]
]
if not "W" in bkg_processes:
bkg_processes.insert(bkg_processes_names.index("w"), "W")
if not "QCD" in bkg_processes:
bkg_processes.insert(bkg_processes_names.index("qcd"), "QCD")
for variable, category in zip(variables, categories):
config = deepcopy(config_template)
if args.chi2:
config["analysis_modules"].append("AddHistograms")
config["analysis_modules"].append("Chi2Test")
config["add_nicks"] = [" ".join(bkg_processes_names)]
config["add_result_nicks"] = ["tot_background_noplot"]
config["chi2test_nicks"] = ["tot_background_noplot data"]
config["chi2test_compare"] = ["UW CHI2/NDF"]
config["files"] = [shapes]
config["lumis"] = [args.lumi]
config["output_dir"] = output_dir + "/" + channel
config["y_log"] = True if (variable in logvars) else False
config["y_rel_lims"] = [0.9, 50] if y_log else [0.9, 1.5]
config["markers"] = ["HIST"] * len(bkg_processes_names) + [
"P"
] + config["markers"]
config["legend_markers"] = ["F"] * (len(bkg_processes_names)) + [
"ELP"
] + config["legend_markers"]
config["labels"] = bkg_processes_names + ["data"
] + config["labels"]
config["colors"] = bkg_processes_names + ["data"
] + config["colors"]
config["nicks"] = bkg_processes_names + ["data"]
config["x_expressions"] = [
"#" + "#".join([
channel, category, process, analysis, "Run2017ReReco31Mar",
variable, mass
]) + "#" for process in bkg_processes
] + [
"#" + "#".join([
channel, category, "data_obs", analysis,
"Run2017ReReco31Mar", variable, mass
]) + "#"
]
if args.emb == False:
config["filename"] = "_".join([
channel, category, analysis, "Run2017ReReco31Mar",
variable, mass
])
else:
config["filename"] = "_".join([
channel, category, analysis, "Run2017ReReco31Mar",
variable, mass, "emb"
])
if args.ff:
config["filename"] = config["filename"] + "_ff"
if not args.x_label == None:
config["x_label"] = args.x_label
else:
config["x_label"] = "_".join([channel, variable])
config["title"] = "_".join(["channel", channel])
config["stacks"] = ["mc"] * len(bkg_processes_names) + [
"data"
] + config["stacks"]
config["ratio_denominator_nicks"] = [
" ".join(bkg_processes_names)
] * 2
config["ratio_numerator_nicks"] = [
" ".join(bkg_processes_names), "data"
]
#config["scale_factors"] = [1.0 if x != "qcd" else qcd_scale_factors[channel] for x in bkg_processes_names] + [1.0]
configs.append(config)
higgsplot.HiggsPlotter(list_of_config_dicts=configs,
list_of_args_strings=[args.additional_arguments],
n_processes=args.num_processes)
def main(args, heavy_mass, light_mass):
signal = []
signal_names = []
prefit_categories = {1: "Inclusive"}
# folders = ["htt_{}_{}_2016_prefit".format(args.channels[0],x,era.strip("Run") for x in categories.keys())
# signal_names=["ggh","qqh"]
# signal=["ggH125","qqH125"]
if "em" in args.channels:
args.ff = False
prefit_categories.update({4: "ss"})
signal.append("NMSSM_{}_125_{}".format(heavy_mass, light_mass))
signal_names.append("nmssm_{}_125_{}".format(heavy_mass, light_mass))
if args.emb and args.ff:
bkg_processes_names = ["emb", "zll", "ttl", "vvl", "fakes", "htt"]
bkg_processes = ["EMB", "ZL", "TTL", "VVL", "jetFakes", "HTT"]
elif args.emb:
bkg_processes_names = [
"emb", "zll", "zj", "ttl", "ttj", "vvl", "vvj", "w", "qcd"
]
bkg_processes = [
"EMB", "ZL", "ZJ", "TTL", "TTJ", "VVL", "VVJ", "W", "QCD"
]
elif args.ff:
bkg_processes_names = [
"ztt", "zll", "ttt", "ttl", "vvt", "vvl", "fakes"
] # enforced by HarryPlotter
bkg_processes = ["ZTT", "ZL", "TTT", "TTL", "VVT", "VVL", "jetFakes"]
else:
bkg_processes_names = [
"ztt", "zll", "zj", "ttl", "ttt", "ttj", "vvl", "vvt", "vvj", "w",
"qcd"
]
bkg_processes = [
"ZTT", "ZL", "ZJ", "TTL", "TTT", "TTJ", "VVL", "VVT", "VVJ", "W",
"QCD"
]
if "em" in args.channels:
bkg_processes = [
p for p in bkg_processes if p not in ["ZJ", "TTJ", "VVJ"]
]
bkg_processes_names = [
p for p in bkg_processes_names if p not in ["zj", "ttj", "vvj"]
]
bkg_processes_save = [x for x in bkg_processes]
bkg_processes_names_save = [x for x in bkg_processes_names]
rfile = ROOT.TFile(args.shapes)
channels = args.channels
analysis = args.analysis
era = args.era
if args.lumi is None:
if "2016" in era:
lumi = 35.9
elif "2017" in era:
lumi = 41.5
elif "2018" in era:
lumi = 59.7
else:
lumi = args.lumi
output_dir = args.output_dir + "/" + era
y_log = args.y_log
x_log = args.x_log
mass = args.mass
variables = args.variables
if args.www:
config_template['www'] = ''
if args.www_dir:
config_template['www_dir'] = args.www_dir
if args.www_no_overwrite:
config_template['www_no_overwrite'] = True
config_template['no_overwrite'] = True
if args.no_overwrite:
config_template['no_overwrite'] = True
if args.redo_cache:
config_template['redo_cache'] = True
if args.filename_prefix != '':
args.filename_prefix = args.filename_prefix if args.filename_prefix[
0] == '_' else '_' + args.filename_prefix
# if args.log_level == 'debug':
# config['log_level'] =
configs = []
for channel in channels:
if args.categories is None:
categories = [channel + "_" + v for v in variables]
shapes = args.shapes if args.shapes is not None else "shapes_{}.root".format(
channel)
if len(variables) == 1:
variables = [variables[0]] * len(prefit_categories.keys())
for variable, category in zip(variables, prefit_categories.keys()):
config = deepcopy(config_template)
bkg_processes = [p for p in bkg_processes_save]
bkg_processes_names = [p for p in bkg_processes_names_save]
empty_processes = []
empty_names = []
fit_mode = "postfit" if "postfit" in shapes else "prefit"
for process, name in zip(bkg_processes, bkg_processes_names):
if not rfile.Get("htt_{}_{}_{}_{}/{}".format(
args.channels[0], category, era.strip("Run"), fit_mode,
process)):
print "Process {} not found for category htt_{}_{}_{}_{}. Continuing.".format(
process, args.channels[0], category, era.strip("Run"),
fit_mode)
empty_processes.append(process)
empty_names.append(name)
for process, name in zip(signal, signal_names):
if not rfile.Get("htt_{}_{}_{}_{}/{}".format(
args.channels[0], category, era.strip("Run"), fit_mode,
process)):
print "Process {} not found for category htt_{}_{}_{}_{}. Continuing.".format(
process, args.channels[0], category, era.strip("Run"),
fit_mode)
empty_processes.append(process)
empty_names.append(name)
signal = ["TotalSig"]
bkg_processes = [
p for p in bkg_processes if not p in empty_processes
]
bkg_processes_names = [
p for p in bkg_processes_names if not p in empty_names
]
config["files"] = [shapes]
config["folders"] = "htt_{}_{}_{}_{}".format(
args.channels[0], category, era.strip("Run"), fit_mode)
config["lumis"] = [lumi]
config["year"] = era.strip("Run")
config[
"output_dir"] = output_dir + "/" + channel #+"/"+prefit_categories[category]
config["y_log"] = True if ((variable in logvars)
or y_log) else False
config["x_log"] = True if x_log else False
gof_file = "CMSSW_10_2_16_UL/src/output/gof/{era}_{channel}_{mass}_{batch}_{cat}/gof.json".format(
era=era.strip("Run"),
channel=channel,
mass=args.mass,
batch=args.batch,
cat=category)
if category == 5:
gof_file = "gof_categories/output/gof/{era}_{channel}_{mass}_{batch}_{cat}/gof.json".format(
era=era.strip("Run"),
channel=channel,
mass=args.mass,
batch=args.batch,
cat=category)
p_value = -1
if os.path.exists(gof_file):
with open(gof_file, "r") as f:
gof_result = json.load(f)
temp_mass = args.heavy_mass
print args.heavy_mass
print gof_file
if int(args.batch) in [5, 6] and int(
args.heavy_mass) == 2000:
temp_mass = 3000
try:
p_value = gof_result["{}.0".format(temp_mass)]["p"]
except:
p_value = -1
# config["subplot_legend"] = [0.15, 0.7, 0.4, 0.8]
# config["subplot_legend_fontsize"] = 0.14
config["y_rel_lims"] = [10, 1000000] if ((variable in logvars)
or y_log) else [0., 1.5]
config["markers"] = ["HIST"] * len(bkg_processes_names) + [
"E2"
] + ["LINE"] * len(signal_names) + ["P"] + [
"E2"
] + ["LINE"] * len(signal_names) + ["P"]
config["legend_markers"] = ["F"] * (len(bkg_processes_names)) + [
"F"
] + ["LX0"] * len(signal_names) + ["ELP"] + [
"E2"
] + ["L"] * len(signal_names) + [""]
p_label = [""] if (fit_mode == "postfit"
or category > 10) else ["p={}".format(p_value)]
signal_label = [
'#scale[0.85]{H(%s)#rightarrowh_{SM}(125)h_{S}(%s) (5 pb)}' %
(heavy_mass, light_mass)
]
if channel == "tt":
signal_label = [
'#scale[0.85]{H(%s)#rightarrowh_{SM}(125)h_{S}(%s) (0.05 pb)}'
% (heavy_mass, light_mass)
]
config["labels"] = bkg_processes_names + [
"Bkg.unc."
] + signal_label + ["data"
] + [""] * len(signal_names) + [""] + p_label
config["nicks"] = bkg_processes_names + signal_names + ["data"]
config["colors"] = bkg_processes_names + [
"transgrey"
] + ["#CF5E61"] * len(signal_names) + ["data"] + [
"#B7B7B7"
] + ["#CF5E61"] * len(signal_names) + ["#000000"]
config["x_expressions"] = bkg_processes + signal + ["data_obs"]
config["scale_factors"] = [1] * len(bkg_processes_names) + [
0.5
] * len(signal_label) + [1] if channel == "tt" else [1] * len(
bkg_processes_names) + [50] * len(signal_label) + [1]
if args.emb == False:
config["filename"] = "_".join([
channel,
str(category), analysis, era, variable, mass,
str(heavy_mass),
str(light_mass)
])
else:
config["filename"] = "_".join([
channel,
str(category), analysis, era, variable, mass,
str(heavy_mass),
str(light_mass), "emb"
])
if args.comparison:
config["filename"] = "_".join([
channel,
str(category), analysis, era, variable, mass, "comparison"
]) + args.filename_prefix
if args.ff:
config["filename"] = config["filename"] + "_ff"
if not args.x_label == None:
config["x_label"] = args.x_label
else:
config["x_label"] = "_".join([channel, variable])
config["filename"] = config["filename"] + "_" + fit_mode
title_dict = {}
title_dict["mt"] = "#mu#tau_{h}"
title_dict["tt"] = "#tau_{h}#tau_{h}"
title_dict["et"] = "e#tau_{h}"
title_dict["em"] = "e#mu"
config["title"] = title_dict[channel]
config["stacks"] = ["mc"] * len(
bkg_processes_names) + signal_names + ["bkg_unc"] + [
"data"
] + [x + "Ratio" for x in signal_names] + config["stacks"]
config["ratio_denominator_nicks"] = [
" ".join(bkg_processes_names)
] * (2 + len(signal_names))
config["ratio_numerator_nicks"] = [" ".join(bkg_processes_names)]
for i in range(len(signal_names)):
config["ratio_numerator_nicks"].append(
" ".join(bkg_processes_names + [signal_names[i]]))
config["ratio_numerator_nicks"].append("data")
config["ratio_result_nicks"] = ["bkg_ratio"] + [
x + "_ratio" for x in signal_names
] + ["data_ratio"]
config["analysis_modules"].append("AddHistograms")
config["add_nicks"] = [" ".join(bkg_processes_names)]
config["add_result_nicks"] = ["background_uncertainty"]
if args.blind or category > 99:
for key in config.keys():
if isinstance(config[key], list):
if isinstance(config[key][0], str):
config[key] = [
x for x in config[key] if (not "data" in x)
]
if "marker" in key:
config[key] = [
x for x in config[key] if (x != "P")
]
if key in [
"ratio_denominator_nicks", "scale_factors",
"colors"
]:
config[key] = config[key][:-1]
#config["legend_cols"] = 1
# import pprint
# pp = pprint.PrettyPrinter(indent=2)
# pp.pprint(config)
configs.append(config)
higgsplot.HiggsPlotter(list_of_config_dicts=configs,
list_of_args_strings=[args.additional_arguments],
n_processes=args.num_processes)
rfile.Close()
config["labels"][index] = os.path.join(
"%s_%s" %
(channel_renamings.get(channel, channel),
category_renamings.get(category, category)),
label_renamings.get(label, label))
config = uncertainty(config, name).get_config(shift)
if "PrintInfos" in config.get("analysis_modules", []):
config.get("analysis_modules", []).remove("PrintInfos")
harry_configs.append(config)
harry_args.append("-d %s --formats png pdf %s" %
(args["input_dir"], args["args"]))
higgs_plotter = higgsplot.HiggsPlotter(list_of_config_dicts=harry_configs,
list_of_args_strings=harry_args,
n_processes=args["n_processes"],
n_plots=args["n_plots"])
root_outputs = list(
set([
output
for output in tools.flattenList(higgs_plotter.output_filenames)
if output.endswith(".root")
]))
command = "hadd -f %s %s" % (args["root_output"], " ".join(root_outputs))
log.info(command)
logger.subprocessCall(shlex.split(command))
log.info("Merged ROOT output is saved to \"%s\"." % args["root_output"])
plot_configs[-1]["files"] = [args["input"]]
plot_configs[-1]["folders"] = [directory]
plot_configs[-1]["x_expressions"] = background + signal + data
plot_configs[-1]["nicks"] = background + (
["sig"] * len(signal)) + (["data"] * len(data))
plot_configs[-1]["stacks"] = (["bkg"] * len(background)) + (
["sig"] * (1 if len(signal) > 0 else 0)) + ["data"]
plot_configs[-1]["markers"] = (["hist"] * len(background)) + (
["line"] * (1 if len(signal) > 0 else 0)) + ["E"]
plot_configs[-1]["ratio"] = args["ratio"]
plot_configs[-1]["colors"] = [
colors.get(x, "#000000")
for index, x in enumerate(plot_configs[-1]["x_expressions"])
]
plot_configs[-1]["labels"] = background + (
["Signal (%d)" % mass] *
(1 if len(signal) > 0 else 0)) + ["Data"]
plot_configs[-1]["legend"] = [0.75, 0.6]
plot_configs[-1]["x_label"] = ""
plot_configs[-1]["y_label"] = "Events"
plot_configs[-1]["output_dir"] = os.path.join(
"plots", "prefit",
os.path.splitext(os.path.basename(args["input"]))[0])
plot_configs[-1]["filename"] = directory + "_mH" + str(mass)
higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs,
list_of_args_strings=[args["args"]],
n_processes=args["n_processes"],
n_plots=args["n_plots"])
def prefit_postfit_plots(self, datacards_cbs, datacards_postfit_shapes, plotting_args=None, n_processes=1, signal_stacked_on_bkg=False, *args, **kwargs):
if plotting_args is None:
plotting_args = {}
base_path = reduce(lambda datacard1, datacard2: tools.longest_common_substring(datacard1, datacard2), datacards_cbs.keys())
plot_configs = []
bkg_plotting_order = ["ZTTPOSPOL", "ZTTNEGPOL", "ZTT", "ZLL", "ZL", "ZJ", "EWKZ", "TTTAUTAU", "TTT", "TTJJ", "TTJ", "TT", "VVT", "VVJ", "VV", "WJ", "W", "hww_gg125", "hww_qq125", "EWK", "QCD"]
for level in ["prefit", "postfit"]:
for index, (fit_type, datacards_postfit_shapes_dict) in enumerate(datacards_postfit_shapes.iteritems()):
if (index == 0) or (level == "postfit"):
for datacard, postfit_shapes in datacards_postfit_shapes_dict.iteritems():
for category in datacards_cbs[datacard].cp().bin_set():
stacked_processes = []
if signal_stacked_on_bkg:
stacked_processes.extend(datacards_cbs[datacard].cp().bin([category]).signals().process_set())
stacked_processes.extend(datacards_cbs[datacard].cp().bin([category]).backgrounds().process_set())
stacked_processes.sort(key=lambda process: bkg_plotting_order.index(process) if process in bkg_plotting_order else len(bkg_plotting_order))
#stacked_processes = [process for process in bkg_plotting_order if datacards_cbs[datacard].cp().bin([category]).backgrounds().process([process]).GetRate() > 0.0]
config = {}
processes_to_plot = list(stacked_processes)
# merge backgrounds from dictionary if provided
if plotting_args.get("merge_backgrounds", False):
if not "SumOfHistograms" in config.get("analysis_modules", []):
config.setdefault("analysis_modules", []).append("SumOfHistograms")
merge_backgrounds = plotting_args.get("merge_backgrounds", {})
for new_background, backgrounds_to_merge in merge_backgrounds.iteritems():
if new_background not in stacked_processes:
backgrounds_to_remove = ""
for background in backgrounds_to_merge:
if background in stacked_processes:
stacked_processes = [p + ("_noplot" if p == background else "") for p in stacked_processes]
backgrounds_to_remove += background + "_noplot "
config.setdefault("sum_nicks", []).append(backgrounds_to_remove)
config.setdefault("sum_result_nicks", []).append(new_background)
processes_to_plot = [p for p in stacked_processes if not "noplot" in p]
for new_background in merge_backgrounds:
if new_background not in stacked_processes:
processes_to_plot.append(new_background)
processes_to_plot.sort(key=lambda process: bkg_plotting_order.index(process) if process in bkg_plotting_order else len(bkg_plotting_order))
config["files"] = [postfit_shapes]
config["folders"] = [category+"_"+level]
config["x_expressions"] = [p.strip("_noplot") for p in stacked_processes] + ["TotalSig"] + ["data_obs", "TotalBkg"]
config["nicks"] = stacked_processes + ["TotalSig" + ("_noplot" if signal_stacked_on_bkg else "")] + ["data_obs", "TotalBkg" + ("_noplot" if signal_stacked_on_bkg else "")]
config["stacks"] = (["stack"]*len(processes_to_plot)) + (["data"] if signal_stacked_on_bkg else ["sig", "data", "bkg_unc"])
config["labels"] = [label.lower() for label in processes_to_plot + (["data_obs"] if signal_stacked_on_bkg else ["TotalSig", "data_obs", "TotalBkg"])]
config["colors"] = [color.lower() for color in processes_to_plot + (["data_obs"] if signal_stacked_on_bkg else ["TotalSig", "data_obs", "TotalBkg"])]
config["markers"] = (["HIST"]*len(processes_to_plot)) + (["E"] if signal_stacked_on_bkg else ["LINE", "E", "E2"])
config["legend_markers"] = (["F"]*len(processes_to_plot)) + (["ELP"] if signal_stacked_on_bkg else ["L", "ELP", "F"])
config["x_label"] = category.split("_")[0]+"_"+plotting_args.get("x_expressions", None)
config["title"] = "channel_"+category.split("_")[0]
config["energies"] = [13.0]
config["lumis"] = [float("%.1f" % plotting_args.get("lumi", 1.0))]
if plotting_args.get("era", False):
config["year"] = plotting_args.get("era")
config["legend"] = [0.7, 0.6, 0.9, 0.88]
config["y_lims"] = [0.0]
config["output_dir"] = os.path.join(os.path.dirname(datacard), "plots")
config["filename"] = level+("_"+fit_type if level == "postfit" else "")+"_"+category
if plotting_args.get("www", False):
config["www"] = os.path.join(config["output_dir"].replace(base_path, plotting_args["www"]+"/"))
if plotting_args.get("normalize", False):
config.setdefault("analysis_modules", []).append("NormalizeByBinWidth")
if plotting_args.get("ratio", False):
if signal_stacked_on_bkg:
if not "SumOfHistograms" in config.get("analysis_modules", []):
config.setdefault("analysis_modules", []).append("SumOfHistograms")
config.setdefault("sum_nicks", []).append("TotalBkg_noplot TotalSig_noplot")
config.setdefault("sum_result_nicks", []).append("TotalBkg")
if not "Ratio" in config.get("analysis_modules", []):
config.setdefault("analysis_modules", []).append("Ratio")
# add signal/bkg ratio first
if plotting_args.get("add_soverb_ratio", False):
if not "SumOfHistograms" in config.get("analysis_modules", []):
config.setdefault("analysis_modules", []).append("SumOfHistograms")
config.setdefault("sum_nicks", []).append("TotalBkg TotalSig")
config.setdefault("sum_result_nicks", []).append("TotalSignalPlusBackground_noplot")
config.setdefault("ratio_numerator_nicks", []).append("TotalSignalPlusBackground_noplot")
config.setdefault("ratio_denominator_nicks", []).append("TotalBkg")
config.setdefault("ratio_result_nicks", []).append("ratio_soverb")
# now add data/bkg ratio
config.setdefault("ratio_numerator_nicks", []).extend(["TotalBkg", "data_obs"])
config.setdefault("ratio_denominator_nicks", []).extend(["TotalBkg"] * 2)
config.setdefault("ratio_result_nicks", []).extend(["ratio_unc", "ratio"])
config["ratio_denominator_no_errors"] = True
if plotting_args.get("add_soverb_ratio", False):
config.setdefault("colors", []).append("kRed")
config.setdefault("markers", []).append("LINE")
config.setdefault("legend_markers", []).append("L")
config.setdefault("labels", []).append("")
config.setdefault("stacks", []).append("ratio_soverb")
config.setdefault("colors", []).extend(["totalbkg", "#000000"])
config.setdefault("markers", []).extend(["E2", "E"])
config.setdefault("legend_markers", []).extend(["F", "ELP"])
config.setdefault("labels", []).extend([""] * 2)
config.setdefault("stacks", []).extend(["unc", "ratio"])
config["legend"] = [0.7, 0.4, 0.95, 0.83]
config["subplot_grid"] = "True"
config["y_subplot_lims"] = [0.5, 1.5]
config["y_subplot_label"] = "Obs./Exp."
# update ordering if backgrounds were merged
if plotting_args.get("merge_backgrounds", False):
config["nicks_whitelist"] = processes_to_plot + ["TotalSig" + ("_noplot" if signal_stacked_on_bkg else "")] + ["data_obs", "TotalBkg" + ("_noplot" if signal_stacked_on_bkg else "")]
if plotting_args.get("ratio", False):
if plotting_args.get("add_soverb_ratio", False):
config["nicks_whitelist"].append("ratio_soverb")
config["nicks_whitelist"].extend(["ratio_unc", "ratio"])
plot_configs.append(config)
if plotting_args.get("return_configs", False):
return plot_configs
else:
# create result plots HarryPlotter
return higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs, list_of_args_strings=[plotting_args.get("args", "")], n_processes=n_processes)
plot_configs.append(config)
hadd_commands.append("hadd -f {DST} {SRC}".format(
DST=output_file, SRC=" ".join(tmp_output_files)))
# # Delete existing output files
# output_files = list(set([os.path.join(config["output_dir"], config["filename"]+".root") for config in plot_configs[:args.n_plots[0]]]))
# for output_file in output_files:
# if os.path.exists(output_file):
# os.remove(output_file)
# log.debug("Removed file \""+output_file+"\" before it is recreated again.")
# Create input histograms with HarryPlotter
if "inputs" in args.steps:
higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs,
n_processes=args.n_processes,
n_plots=args.n_plots[0])
if args.n_plots[0] != 0:
tools.parallelize(_call_command,
hadd_commands,
n_processes=args.n_processes)
# Update CombineHarvester with the yields and shapes
datacards.extract_shapes(os.path.join(
args.output_dir, input_root_filename_template.replace("$", "")),
bkg_histogram_name_template,
sig_histogram_name_template,
bkg_syst_histogram_name_template,
sig_syst_histogram_name_template,
update_systematics=True)
"_alphas_only_": plot_configs_alphas_only,
"_scale_only_": plot_configs_scale_only
}
for category in args.categories:
log.debug(category)
for (key, value) in configs_dict.items():
merge(value[category])
if category != 'inclusive' and 'inclusive' in args.categories:
value[category][0] = sample_settings.merge_configs(
value[category][0], value['inclusive'][0])
log.debug(args.quantities[0] + key + str(category))
value[category][0]["filename"] = args.samples[0] + args.quantities[
0] + key + str(category)
fout = open("merged" + key + str(category) + ".json", "w")
fout.write(
pprint.pformat(value[category][0]).replace("u'D",
"'D").replace(
"'", '"'))
if log.isEnabledFor(logging.DEBUG):
print "Addititonal args for the configuration files:", [args.args]
# This is not working BECAUSE of the single quotes.
#pprint.pprint([configs_dict.values()[0][0]])
higgsplot.HiggsPlotter(
list_of_config_dicts=[configs_dict["_all_"][args.categories[0]][0]],
list_of_args_strings=[args.args],
n_processes=args.n_processes,
n_plots=args.n_plots)
#higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs[args.categories[0]][0], list_of_args_strings=[args.args], n_processes=args.n_processes, n_plots=args.n_plots)
def harry_do_your_job(config):
higgsplot.HiggsPlotter(list_of_config_dicts=[config])
output_files = list(set(output_files))
# create input histograms with HarryPlotter
if not args.batch:
pool = Pool(cpu_count())
for config in plot_configs:
pool.apply_async(harry_do_your_job, args=(config, ))
pool.close()
pool.join()
else:
# create input histograms with HarryPlotter
higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs,
n_processes=1,
n_plots=len(plot_configs),
batch="rwthcondor")
tools.parallelize(_call_command, hadd_commands, n_processes=1)
datacards.extract_shapes(os.path.join(
args.output_dir, input_root_filename_template.replace("$", "")),
bkg_histogram_name_template,
sig_histogram_name_template,
bkg_syst_histogram_name_template,
sig_syst_histogram_name_template,
update_systematics=True)
"""
datacards.add_bin_by_bin_uncertainties(
processes=datacards.cb.cp().backgrounds().process_set(),
add_threshold=0.05, merge_threshold=0.8, fix_norm=False
w_gen = [wll_gen, whl_gen, whh_gen]
w_reco = [wll_reco, whl_reco, whh_reco]
plot_configs = []
#comparing gen and reco
for i in xrange(3):
config = jsonTools.JsonDict(
os.path.expandvars(
"$CMSSW_BASE/src/HiggsAnalysis/KITHiggsToTauTau/data/plots/configs/reco_CP/recogenPhiStarCP.json"
))
config["files"] = files[i]
for j in xrange(3):
config["filename"] = "recogenPhiStarCP_" + samples[
i] + "_" + channels[j]
config["directories"] = args.input
config["folders"] = [folders[j]]
config["texts"] = texts[j]
config["weights"] = [w_gen[j] + "&&" + w_reco[j]]
plot_configs.append(config.copy())
if log.isEnabledFor(logging.DEBUG):
import pprint
pprint.pprint(plot_configs)
higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs,
list_of_args_strings=[args.args],
n_processes=args.n_processes,
n_plots=args.n_plots)
config["fill_styles"] = [3001, 0]
config["legend"] = [0.75, 0.8]
config["legend_markers"] = ["LF", "LP"]
config["x_lims"] = [-5.0, 5.0]
config["x_label"] = "Pull values"
config["output_dir"] = os.path.join(os.path.dirname(datacard), "plots")
config["filename"] = "pulls"
if plotting_args.get("www", False):
config["www"] = os.path.join(config["output_dir"].replace(base_path, plotting_args["www"]+"/"))
plot_configs.append(config)
# create result plots HarryPlotter
<<<<<<< HEAD
return higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs, list_of_args_strings=[plotting_args.get("args", "")], n_processes=n_processes)
def nuisance_impacts(self, datacards_cbs, datacards_workspaces, n_processes=1, *args, **kwargs):
=======
return higgsplot.HiggsPlotter(list_of_config_dicts=plot_configs, list_of_args_strings=[plotting_args.get("args", "")], n_processes=n_processes)
def nuisance_impacts(self, datacards_cbs, datacards_workspaces, n_processes=1, *args):
>>>>>>> origin/master
tmp_args = " ".join(args)
for key, value in kwargs.items():
higgs_mass = value if "higgs_mass" in key else "0"
commandsInitialFit = []
commandsInitialFit.extend([[
"combineTool.py -M Impacts -d {WORKSPACE} -m {MASS} --robustFit 1 --minimizerTolerance 0.1 --minimizerStrategy 0 --minimizerAlgoForMinos Minuit2,migrad --doInitialFit --allPars {ARGS}".format(
