def test_packed_selection():
from coffea.analysis_tools import PackedSelection
sel = PackedSelection()
shape = (10, )
all_true = np.full(shape=shape, fill_value=True, dtype=np.bool)
all_false = np.full(shape=shape, fill_value=False, dtype=np.bool)
fizz = np.arange(shape[0]) % 3 == 0
buzz = np.arange(shape[0]) % 5 == 0
ones = np.ones(shape=shape, dtype=np.uint64)
wrong_shape = ones = np.ones(shape=(shape[0] - 5, ), dtype=np.bool)
sel.add("all_true", all_true)
sel.add("all_false", all_false)
sel.add("fizz", fizz)
sel.add("buzz", buzz)
assert np.all(sel.require(all_true=True, all_false=False) == all_true)
# allow truthy values
assert np.all(sel.require(all_true=1, all_false=0) == all_true)
assert np.all(sel.all("all_true", "all_false") == all_false)
assert np.all(sel.any("all_true", "all_false") == all_true)
assert np.all(
sel.all("fizz", "buzz") == np.array([
True, False, False, False, False, False, False, False, False, False
]))
assert np.all(
sel.any("fizz", "buzz") == np.array(
[True, False, False, True, False, True, True, False, False, True]))
with pytest.raises(ValueError):
sel.add("wrong_shape", wrong_shape)
with pytest.raises(ValueError):
sel.add("ones", ones)
with pytest.raises(RuntimeError):
overpack = PackedSelection()
for i in range(65):
overpack.add("sel_%d", all_true)
def test_packed_selection():
from coffea.analysis_tools import PackedSelection
sel = PackedSelection()
counts, test_eta, test_pt = dummy_jagged_eta_pt()
all_true = np.full(shape=counts.shape, fill_value=True, dtype=np.bool)
all_false = np.full(shape=counts.shape, fill_value=False, dtype=np.bool)
ones = np.ones(shape=counts.shape, dtype=np.uint64)
wrong_shape = ones = np.ones(shape=(counts.shape[0] - 5, ), dtype=np.bool)
sel.add("all_true", all_true)
sel.add("all_false", all_false)
assert np.all(sel.require(all_true=True, all_false=False) == all_true)
assert np.all(sel.all("all_true", "all_false") == all_false)
try:
sel.require(all_true=1, all_false=0)
except ValueError:
pass
try:
sel.add("wrong_shape", wrong_shape)
except ValueError:
pass
try:
sel.add("ones", ones)
except ValueError:
pass
try:
overpack = PackedSelection()
for i in range(65):
overpack.add("sel_%d", all_true)
except RuntimeError:
pass
def SS_selection(lep1, lep2):
selection = PackedSelection()
is_dilep = ((ak.num(lep1) + ak.num(lep2))==2)
pos_charge = ((ak.sum(lep1.pdgId, axis=1) + ak.sum(lep2.pdgId, axis=1))<0)
neg_charge = ((ak.sum(lep1.pdgId, axis=1) + ak.sum(lep2.pdgId, axis=1))>0)
dilep2 = choose(lep2, 2)
dilep1 = choose(lep1, 2)
dilep = cross(lep2, lep1)
is_SS = ( ak.any((dilep2['0'].charge * dilep2['1'].charge)>0, axis=1) | \
ak.any((dilep1['0'].charge * dilep1['1'].charge)>0, axis=1) | \
ak.any((dilep['0'].charge * dilep['1'].charge)>0, axis=1) )
selection.add('SS', is_SS)
ss_reqs = ['SS']
ss_reqs_d = {sel: True for sel in ss_reqs}
ss_selection = selection.require(**ss_reqs_d)
return ss_selection
def process(self, events):
output = self.accumulator.identity()
# use a very loose preselection to filter the events
presel = ak.num(events.Jet) > 2
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Muons
muon = Collections(ev, "Muon", "tight").get()
vetomuon = Collections(ev, "Muon", "veto").get()
dimuon = choose(muon, 2)
SSmuon = ak.any((dimuon['0'].charge * dimuon['1'].charge) > 0, axis=1)
OSmuon = ak.any((dimuon['0'].charge * dimuon['1'].charge) < 0, axis=1)
leading_muon_idx = ak.singletons(ak.argmax(muon.pt, axis=1))
leading_muon = muon[leading_muon_idx]
## Electrons
electron = Collections(ev, "Electron", "tight").get()
vetoelectron = Collections(ev, "Electron", "veto").get()
dielectron = choose(electron, 2)
SSelectron = ak.any(
(dielectron['0'].charge * dielectron['1'].charge) > 0, axis=1)
OSelectron = ak.any(
(dielectron['0'].charge * dielectron['1'].charge) < 0, axis=1)
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[leading_electron_idx]
## Merge electrons and muons - this should work better now in ak1
lepton = ak.concatenate([muon, electron], axis=1)
dilepton = cross(muon, electron)
SSlepton = ak.any((dilepton['0'].charge * dilepton['1'].charge) > 0,
axis=1)
OSlepton = ak.any((dilepton['0'].charge * dilepton['1'].charge) < 0,
axis=1)
leading_lepton_idx = ak.singletons(ak.argmax(lepton.pt, axis=1))
leading_lepton = lepton[leading_lepton_idx]
trailing_lepton_idx = ak.singletons(ak.argmin(lepton.pt, axis=1))
trailing_lepton = lepton[trailing_lepton_idx]
## Jets
jet = getJets(ev, minPt=25, maxEta=4.7)
jet = jet[(jet.pt > 25) & (jet.jetId > 1)]
jet = jet[~match(jet, muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(
jet, electron,
deltaRCut=0.4)] # remove jets that overlap with electrons
## event selectors
filters = getFilters(ev, year=self.year, dataset=dataset)
triggers = getTriggers(ev, year=self.year, dataset=dataset)
dilep = ((ak.num(electron) == 1) & (ak.num(muon) == 1))
lep0pt = ((ak.num(electron[(electron.pt > 25)]) +
ak.num(muon[(muon.pt > 25)])) > 0)
lep1pt = ((ak.num(electron[(electron.pt > 20)]) +
ak.num(muon[(muon.pt > 20)])) > 1)
lepveto = ((ak.num(vetoelectron) + ak.num(vetomuon)) == 2)
# define the weight
weight = Weights(len(ev))
if not dataset == 'MuonEG':
# lumi weight
weight.add("weight", ev.weight * cfg['lumi'][self.year])
## PU weight - not in the babies...
#weight.add("PU", ev.puWeight, weightUp=ev.puWeightUp, weightDown=ev.puWeightDown, shift=False)
# b-tag SFs
#weight.add("btag", self.btagSF.Method1a(btag, light))
# lepton SFs
weight.add("lepton", self.leptonSF.get(electron, muon))
selection = PackedSelection()
selection.add('lepveto', lepveto)
selection.add('dilep', dilep)
selection.add('trigger', (triggers))
selection.add('filter', (filters))
selection.add('p_T(lep0)>25', lep0pt)
selection.add('p_T(lep1)>20', lep1pt)
selection.add('OS', OSlepton)
selection.add('N_jet>2', (ak.num(jet) >= 3))
selection.add('MET>30', (ev.MET.pt > 30))
os_reqs = [
'lepveto', 'dilep', 'trigger', 'filter', 'p_T(lep0)>25',
'p_T(lep1)>20', 'OS'
]
bl_reqs = os_reqs + ['N_jet>2', 'MET>30']
os_reqs_d = {sel: True for sel in os_reqs}
os_selection = selection.require(**os_reqs_d)
bl_reqs_d = {sel: True for sel in bl_reqs}
BL = selection.require(**bl_reqs_d)
cutflow = Cutflow(output, ev, weight=weight)
cutflow_reqs_d = {}
for req in bl_reqs:
cutflow_reqs_d.update({req: True})
cutflow.addRow(req, selection.require(**cutflow_reqs_d))
# first, make a few super inclusive plots
output['PV_npvs'].fill(dataset=dataset,
multiplicity=ev.PV[os_selection].npvs,
weight=weight.weight()[os_selection])
output['PV_npvsGood'].fill(dataset=dataset,
multiplicity=ev.PV[os_selection].npvsGood,
weight=weight.weight()[os_selection])
output['N_jet'].fill(dataset=dataset,
multiplicity=ak.num(jet)[os_selection],
weight=weight.weight()[os_selection])
output['MET'].fill(dataset=dataset,
pt=ev.MET[os_selection].pt,
phi=ev.MET[os_selection].phi,
weight=weight.weight()[os_selection])
output['j1'].fill(dataset=dataset,
pt=ak.flatten(jet.pt[:, 0:1][BL]),
eta=ak.flatten(jet.eta[:, 0:1][BL]),
phi=ak.flatten(jet.phi[:, 0:1][BL]),
weight=weight.weight()[BL])
# Now, take care of systematic unceratinties
if not dataset == 'MuonEG':
alljets = getJets(ev, minPt=0, maxEta=4.7)
alljets = alljets[(alljets.jetId > 1)]
for var in self.variations:
# get the collections that change with the variations
jet_var = getPtEtaPhi(alljets, pt_var=var)
jet_var = jet_var[(jet_var.pt > 25)]
jet_var = jet_var[~match(
jet_var, muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet_var = jet_var[~match(
jet_var, electron,
deltaRCut=0.4)] # remove jets that overlap with electrons
# get the modified selection -> more difficult
selection.add(
'N_jet>2_' + var, (ak.num(jet_var.pt) > 3)
) # something needs to be improved with getPtEtaPhi function
selection.add('MET>30_' + var, (getattr(ev.MET, var) > 30))
bl_reqs = os_reqs + ['N_jet>2_' + var, 'MET>30_' + var]
bl_reqs_d = {sel: True for sel in bl_reqs}
BL = selection.require(**bl_reqs_d)
# the OS selection remains unchanged
output['N_jet_' + var].fill(
dataset=dataset,
multiplicity=ak.num(jet_var)[os_selection],
weight=weight.weight()[os_selection])
# We don't need to redo all plots with variations. E.g., just add uncertainties to the jet plots.
output['j1_' + var].fill(dataset=dataset,
pt=ak.flatten(jet_var.pt[:, 0:1][BL]),
eta=ak.flatten(jet_var.eta[:,
0:1][BL]),
phi=ak.flatten(jet_var.phi[:,
0:1][BL]),
weight=weight.weight()[BL])
return output
def process(self, events):
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet) > 0
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Electrons
electron = Collections(ev, "Electron", "tight").get()
electron = electron[(electron.miniPFRelIso_all < 0.12)
& (electron.pt > 20) & (abs(electron.eta) < 2.4)]
gen_matched_electron = electron[((electron.genPartIdx >= 0) & (abs(
electron.matched_gen.pdgId) == 11))]
n_gen = ak.num(gen_matched_electron)
is_flipped = ((gen_matched_electron.matched_gen.pdgId *
(-1) == gen_matched_electron.pdgId) &
(abs(gen_matched_electron.pdgId) == 11))
#is_flipped = (abs(ev.GenPart[gen_matched_electron.genPartIdx].pdgId) == abs(gen_matched_electron.pdgId))&(ev.GenPart[gen_matched_electron.genPartIdx].pdgId/abs(ev.GenPart[gen_matched_electron.genPartIdx].pdgId) != gen_matched_electron.pdgId/abs(gen_matched_electron.pdgId))
flipped_electron = gen_matched_electron[is_flipped]
n_flips = ak.num(flipped_electron)
sielectron = choose(electron, 1)
dielectron = choose(electron, 2)
SSelectron = ak.any(
(dielectron['0'].charge * dielectron['1'].charge) > 0, axis=1)
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[leading_electron_idx]
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
# setting up the various weights
weight = Weights(len(ev))
weight2 = Weights(len(ev))
if not dataset == 'MuonEG':
# generator weight
weight.add("weight", ev.genWeight)
weight2.add("weight", ev.genWeight)
weight2.add("charge flip",
self.charge_flip_ratio.flip_ratio(sielectron['0']))
#selections
filters = getFilters(ev, year=self.year, dataset=dataset)
electr = ((ak.num(electron) == 2))
ss = (SSelectron)
gen = (n_gen >= 1)
flip = (n_flips >= 1)
selection = PackedSelection()
selection.add('filter', (filters))
selection.add('electr', electr)
selection.add('ss', ss)
selection.add('flip', flip)
selection.add('gen', gen)
bl_reqs = ['filter', 'electr']
bl_reqs_d = {sel: True for sel in bl_reqs}
baseline = selection.require(**bl_reqs_d)
s_reqs = bl_reqs + ['ss']
s_reqs_d = {sel: True for sel in s_reqs}
ss_sel = selection.require(**s_reqs_d)
f_reqs = bl_reqs + ['gen', 'flip']
f_reqs_d = {sel: True for sel in f_reqs}
flip_sel = selection.require(**f_reqs_d)
#outputs
output['N_ele'].fill(dataset=dataset,
multiplicity=ak.num(electron)[flip_sel],
weight=weight.weight()[flip_sel])
output['electron_flips'].fill(dataset=dataset,
multiplicity=n_flips[flip_sel],
weight=weight.weight()[flip_sel])
output['N_ele2'].fill(dataset=dataset,
multiplicity=ak.num(electron)[baseline],
weight=weight2.weight()[baseline])
output['electron_flips2'].fill(dataset=dataset,
multiplicity=n_flips[baseline],
weight=weight2.weight()[baseline])
output["electron"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(flipped_electron[flip_sel].pt)),
eta=ak.to_numpy(ak.flatten(abs(flipped_electron[flip_sel].eta))),
#phi = ak.to_numpy(ak.flatten(leading_electron[baseline].phi)),
weight=weight.weight()[flip_sel])
output["electron2"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_electron[baseline].pt)),
eta=ak.to_numpy(ak.flatten(abs(leading_electron[baseline].eta))),
#phi = ak.to_numpy(ak.flatten(leading_electron[baseline].phi)),
weight=weight2.weight()[baseline])
return output
def process(self, events):
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet) > 0
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Electrons
electron = Collections(ev, "Electron", "tightFCNC", 0, self.year).get()
electron = electron[(electron.pt > 15) & (np.abs(electron.eta) < 2.4)]
electron = electron[(electron.genPartIdx >= 0)]
electron = electron[(np.abs(electron.matched_gen.pdgId) == 11
)] #from here on all leptons are gen-matched
electron = electron[((electron.genPartFlav == 1) |
(electron.genPartFlav
== 15))] #and now they are all prompt
is_flipped = (((electron.matched_gen.pdgId * (-1) == electron.pdgId) |
(find_first_parent(electron.matched_gen) *
(-1) == electron.pdgId)) &
(np.abs(electron.pdgId) == 11))
flipped_electron = electron[is_flipped]
n_flips = ak.num(flipped_electron)
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[leading_electron_idx]
leading_flipped_electron_idx = ak.singletons(
ak.argmax(flipped_electron.pt, axis=1))
leading_flipped_electron = electron[leading_flipped_electron_idx]
def getMVAscore(electron):
if self.year == 2016:
MVA = electron.mvaSpring16GP
return MVA
elif self.year == 2017:
MVA = electron.mvaFall17V2noIso
return MVA
elif self.year == 2018:
MVA = np.minimum(
np.maximum(electron.mvaFall17V2noIso, -1.0 + 1.e-6),
1.0 - 1.e-6)
return -0.5 * np.log(2 / (MVA + 1) - 1)
# setting up the various weights
#weight = Weights( len(ev) )
#if not dataset=='MuonEG':
# generator weight
# weight.add("weight", ev.genWeight)
#selections
filters = getFilters(ev, year=self.year, dataset=dataset)
electr = ((ak.num(electron) >= 1))
flip = (n_flips >= 1)
selection = PackedSelection()
selection.add('filter', (filters))
selection.add('electr', electr)
selection.add('flip', flip)
bl_reqs = ['filter', 'electr']
bl_reqs_d = {sel: True for sel in bl_reqs}
baseline = selection.require(**bl_reqs_d)
f_reqs = bl_reqs + ['flip']
f_reqs_d = {sel: True for sel in f_reqs}
flip_sel = selection.require(**f_reqs_d)
#adjust weights to prevent length mismatch
#ak_weight_gen = ak.ones_like(electron[baseline].pt) * weight.weight()[baseline]
#ak_weight_flip = ak.ones_like(flipped_electron[flip_sel].pt) * weight.weight()[flip_sel]
#output['N_ele'].fill(dataset=dataset, multiplicity=ak.num(electron)[baseline], weight=weight.weight()[baseline])
#output['electron_flips'].fill(dataset=dataset, multiplicity=n_flips[baseline], weight=weight.weight()[baseline])
output["electron"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(electron[baseline].pt)),
eta=abs(ak.to_numpy(ak.flatten(electron[baseline].eta))),
)
output["electron2"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(electron[baseline].pt)),
eta=ak.to_numpy(ak.flatten(electron[baseline].eta)),
)
output["flipped_electron"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(flipped_electron[flip_sel].pt)),
eta=abs(ak.to_numpy(ak.flatten(flipped_electron[flip_sel].eta))),
)
output["flipped_electron2"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(flipped_electron[flip_sel].pt)),
eta=ak.to_numpy(ak.flatten(flipped_electron[flip_sel].eta)),
)
output["mva_id"].fill(
dataset=dataset,
mva_id=ak.to_numpy(ak.flatten(getMVAscore(electron)[baseline])),
eta=np.abs(ak.to_numpy(ak.flatten(electron.etaSC[baseline]))),
)
output["mva_id2"].fill(
dataset=dataset,
mva_id=ak.to_numpy(ak.flatten(getMVAscore(electron)[baseline])),
pt=ak.to_numpy(ak.flatten(electron.pt[baseline])),
)
output["isolation"].fill(
dataset=dataset,
isolation1=ak.to_numpy(ak.flatten(electron.jetRelIso[baseline])),
isolation2=ak.to_numpy(ak.flatten(electron.jetPtRelv2[baseline])),
)
return output
def process(self, events):
output = self.accumulator.identity()
# use a very loose preselection to filter the events
presel = ak.num(events.Jet)>2
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Generated leptons
gen_lep = ev.GenL
leading_gen_lep = gen_lep[ak.singletons(ak.argmax(gen_lep.pt, axis=1))]
trailing_gen_lep = gen_lep[ak.singletons(ak.argmin(gen_lep.pt, axis=1))]
## Muons
muon = Collections(ev, "Muon", "tightSSTTH").get()
vetomuon = Collections(ev, "Muon", "vetoTTH").get()
dimuon = choose(muon, 2)
SSmuon = ak.any((dimuon['0'].charge * dimuon['1'].charge)>0, axis=1)
leading_muon_idx = ak.singletons(ak.argmax(muon.pt, axis=1))
leading_muon = muon[leading_muon_idx]
## Electrons
electron = Collections(ev, "Electron", "tightSSTTH").get()
vetoelectron = Collections(ev, "Electron", "vetoTTH").get()
dielectron = choose(electron, 2)
SSelectron = ak.any((dielectron['0'].charge * dielectron['1'].charge)>0, axis=1)
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[leading_electron_idx]
## Merge electrons and muons - this should work better now in ak1
dilepton = cross(muon, electron)
SSlepton = ak.any((dilepton['0'].charge * dilepton['1'].charge)>0, axis=1)
lepton = ak.concatenate([muon, electron], axis=1)
leading_lepton_idx = ak.singletons(ak.argmax(lepton.pt, axis=1))
leading_lepton = lepton[leading_lepton_idx]
trailing_lepton_idx = ak.singletons(ak.argmin(lepton.pt, axis=1))
trailing_lepton = lepton[trailing_lepton_idx]
n_nonprompt = getNonPromptFromFlavour(electron) + getNonPromptFromFlavour(muon)
n_chargeflip = getChargeFlips(electron, ev.GenPart) + getChargeFlips(muon, ev.GenPart)
## Jets
jet = getJets(ev, minPt=25, maxEta=4.7, pt_var='pt_nom')
jet = jet[ak.argsort(jet.pt_nom, ascending=False)] # need to sort wrt smeared and recorrected jet pt
jet = jet[~match(jet, muon, deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(jet, electron, deltaRCut=0.4)] # remove jets that overlap with electrons
central = jet[(abs(jet.eta)<2.4)]
btag = getBTagsDeepFlavB(jet, year=self.year) # should study working point for DeepJet
light = getBTagsDeepFlavB(jet, year=self.year, invert=True)
fwd = getFwdJet(light)
fwd_noPU = getFwdJet(light, puId=False)
## forward jets
j_fwd = fwd[ak.singletons(ak.argmax(fwd.p, axis=1))] # highest momentum spectator
jf = cross(j_fwd, jet)
mjf = (jf['0']+jf['1']).mass
j_fwd2 = jf[ak.singletons(ak.argmax(mjf, axis=1))]['1'] # this is the jet that forms the largest invariant mass with j_fwd
delta_eta = abs(j_fwd2.eta - j_fwd.eta)
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
## other variables
ht = ak.sum(jet.pt, axis=1)
st = met_pt + ht + ak.sum(muon.pt, axis=1) + ak.sum(electron.pt, axis=1)
## event selectors
filters = getFilters(ev, year=self.year, dataset=dataset)
dilep = ((ak.num(electron) + ak.num(muon))==2)
pos_charge = ((ak.sum(electron.pdgId, axis=1) + ak.sum(muon.pdgId, axis=1))<0)
neg_charge = ((ak.sum(electron.pdgId, axis=1) + ak.sum(muon.pdgId, axis=1))>0)
lep0pt = ((ak.num(electron[(electron.pt>25)]) + ak.num(muon[(muon.pt>25)]))>0)
lep0pt_40 = ((ak.num(electron[(electron.pt>40)]) + ak.num(muon[(muon.pt>40)]))>0)
lep0pt_100 = ((ak.num(electron[(electron.pt>100)]) + ak.num(muon[(muon.pt>100)]))>0)
lep1pt = ((ak.num(electron[(electron.pt>20)]) + ak.num(muon[(muon.pt>20)]))>1)
lep1pt_30 = ((ak.num(electron[(electron.pt>30)]) + ak.num(muon[(muon.pt>30)]))>1)
lepveto = ((ak.num(vetoelectron) + ak.num(vetomuon))==2)
# define the weight
weight = Weights( len(ev) )
#mult = 1
#if dataset=='inclusive': mult = 0.0478/47.448
#if dataset=='plus': mult = 0.0036/7.205
if not dataset=='MuonEG':
# lumi weight
weight.add("weight", ev.weight*cfg['lumi'][self.year])
#weight.add("weight", ev.genWeight*cfg['lumi'][self.year]*mult)
# PU weight - not in the babies...
weight.add("PU", ev.puWeight, weightUp=ev.puWeightUp, weightDown=ev.puWeightDown, shift=False)
# b-tag SFs
weight.add("btag", self.btagSF.Method1a(btag, light))
# lepton SFs
weight.add("lepton", self.leptonSF.get(electron, muon))
selection = PackedSelection()
selection.add('lepveto', lepveto)
selection.add('dilep', dilep )
selection.add('filter', (filters) )
selection.add('p_T(lep0)>25', lep0pt )
selection.add('p_T(lep0)>40', lep0pt_40 )
selection.add('p_T(lep1)>20', lep1pt )
selection.add('p_T(lep1)>30', lep1pt_30 )
selection.add('SS', ( SSlepton | SSelectron | SSmuon) )
selection.add('pos', ( pos_charge ) )
selection.add('neg', ( neg_charge ) )
selection.add('N_jet>3', (ak.num(jet)>=4) )
selection.add('N_jet>4', (ak.num(jet)>=5) )
selection.add('N_central>2', (ak.num(central)>=3) )
selection.add('N_central>3', (ak.num(central)>=4) )
selection.add('N_btag>0', (ak.num(btag)>=1) )
selection.add('MET>50', (ev.MET.pt>50) )
selection.add('ST', (st>600) )
selection.add('N_fwd>0', (ak.num(fwd)>=1 ))
selection.add('delta_eta', (ak.any(delta_eta>2, axis=1) ) )
selection.add('fwd_p>500', (ak.any(j_fwd.p>500, axis=1) ) )
ss_reqs = ['lepveto', 'dilep', 'SS', 'filter', 'p_T(lep0)>25', 'p_T(lep1)>20', 'N_jet>3', 'N_central>2', 'N_btag>0']
bl_reqs = ss_reqs + ['N_fwd>0', 'N_jet>4', 'N_central>3', 'ST', 'MET>50', 'delta_eta']
sr_reqs = bl_reqs + ['fwd_p>500', 'p_T(lep0)>40', 'p_T(lep1)>30']
ss_reqs_d = { sel: True for sel in ss_reqs }
ss_selection = selection.require(**ss_reqs_d)
bl_reqs_d = { sel: True for sel in bl_reqs }
BL = selection.require(**bl_reqs_d)
sr_reqs_d = { sel: True for sel in sr_reqs }
SR = selection.require(**sr_reqs_d)
cutflow = Cutflow(output, ev, weight=weight)
cutflow_reqs_d = {}
for req in sr_reqs:
cutflow_reqs_d.update({req: True})
cutflow.addRow( req, selection.require(**cutflow_reqs_d) )
# first, make a few super inclusive plots
output['PV_npvs'].fill(dataset=dataset, multiplicity=ev.PV[ss_selection].npvs, weight=weight.weight()[ss_selection])
output['PV_npvsGood'].fill(dataset=dataset, multiplicity=ev.PV[ss_selection].npvsGood, weight=weight.weight()[ss_selection])
output['N_jet'].fill(dataset=dataset, multiplicity=ak.num(jet)[ss_selection], weight=weight.weight()[ss_selection])
output['N_b'].fill(dataset=dataset, multiplicity=ak.num(btag)[ss_selection], weight=weight.weight()[ss_selection])
output['N_central'].fill(dataset=dataset, multiplicity=ak.num(central)[ss_selection], weight=weight.weight()[ss_selection])
output['N_ele'].fill(dataset=dataset, multiplicity=ak.num(electron)[ss_selection], weight=weight.weight()[ss_selection])
output['N_mu'].fill(dataset=dataset, multiplicity=ak.num(electron)[ss_selection], weight=weight.weight()[ss_selection])
output['N_fwd'].fill(dataset=dataset, multiplicity=ak.num(fwd)[ss_selection], weight=weight.weight()[ss_selection])
output['nLepFromTop'].fill(dataset=dataset, multiplicity=ev[BL].nLepFromTop, weight=weight.weight()[BL])
output['nLepFromTau'].fill(dataset=dataset, multiplicity=ev.nLepFromTau[BL], weight=weight.weight()[BL])
output['nLepFromZ'].fill(dataset=dataset, multiplicity=ev.nLepFromZ[BL], weight=weight.weight()[BL])
output['nLepFromW'].fill(dataset=dataset, multiplicity=ev.nLepFromW[BL], weight=weight.weight()[BL])
output['nGenTau'].fill(dataset=dataset, multiplicity=ev.nGenTau[BL], weight=weight.weight()[BL])
output['nGenL'].fill(dataset=dataset, multiplicity=ak.num(ev.GenL[BL], axis=1), weight=weight.weight()[BL])
output['chargeFlip_vs_nonprompt'].fill(dataset=dataset, n1=n_chargeflip[ss_selection], n2=n_nonprompt[ss_selection], n_ele=ak.num(electron)[ss_selection], weight=weight.weight()[ss_selection])
output['MET'].fill(
dataset = dataset,
pt = ev.MET[ss_selection].pt,
phi = ev.MET[ss_selection].phi,
weight = weight.weight()[ss_selection]
)
output['lead_gen_lep'].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_gen_lep[BL].pt)),
eta = ak.to_numpy(ak.flatten(leading_gen_lep[BL].eta)),
phi = ak.to_numpy(ak.flatten(leading_gen_lep[BL].phi)),
weight = weight.weight()[BL]
)
output['trail_gen_lep'].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(trailing_gen_lep[BL].pt)),
eta = ak.to_numpy(ak.flatten(trailing_gen_lep[BL].eta)),
phi = ak.to_numpy(ak.flatten(trailing_gen_lep[BL].phi)),
weight = weight.weight()[BL]
)
output['lead_lep'].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_lepton[BL].pt)),
eta = ak.to_numpy(ak.flatten(leading_lepton[BL].eta)),
phi = ak.to_numpy(ak.flatten(leading_lepton[BL].phi)),
weight = weight.weight()[BL]
)
output['trail_lep'].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(trailing_lepton[BL].pt)),
eta = ak.to_numpy(ak.flatten(trailing_lepton[BL].eta)),
phi = ak.to_numpy(ak.flatten(trailing_lepton[BL].phi)),
weight = weight.weight()[BL]
)
output['j1'].fill(
dataset = dataset,
pt = ak.flatten(jet.pt_nom[:, 0:1][BL]),
eta = ak.flatten(jet.eta[:, 0:1][BL]),
phi = ak.flatten(jet.phi[:, 0:1][BL]),
weight = weight.weight()[BL]
)
output['j2'].fill(
dataset = dataset,
pt = ak.flatten(jet[:, 1:2][BL].pt_nom),
eta = ak.flatten(jet[:, 1:2][BL].eta),
phi = ak.flatten(jet[:, 1:2][BL].phi),
weight = weight.weight()[BL]
)
output['j3'].fill(
dataset = dataset,
pt = ak.flatten(jet[:, 2:3][BL].pt_nom),
eta = ak.flatten(jet[:, 2:3][BL].eta),
phi = ak.flatten(jet[:, 2:3][BL].phi),
weight = weight.weight()[BL]
)
return output
def process(self, events):
np.random.seed(
10
) # sets seed so values from random distributions are reproducible (JER corrections)
output = self.accumulator.identity()
self.sample_name = events.metadata['dataset']
## make event weights
# data or MC distinction made internally
mu_evt_weights = MCWeights.get_event_weights(
events, year=args.year, corrections=self.corrections)
el_evt_weights = MCWeights.get_event_weights(
events, year=args.year, corrections=self.corrections)
## initialize selections and regions
selection = PackedSelection()
regions = {
'Muon': {
'lep_and_filter_pass', 'passing_jets', 'jets_3', 'tight_MU',
'btag_pass', 'semilep'
},
'Electron': {
'lep_and_filter_pass', 'passing_jets', 'jets_3', 'tight_EL',
'btag_pass', 'semilep'
},
}
# get all passing leptons
lep_and_filter_pass = objsel.select_leptons(events, year=args.year)
selection.add('lep_and_filter_pass', lep_and_filter_pass
) # add passing leptons requirement to all systematics
## add different selections
## muons
tight_mu_sel = ak.sum(events['Muon']['TIGHTMU'], axis=1) == 1
selection.add('tight_MU',
tight_mu_sel) # one muon passing TIGHT criteria
## electrons
tight_el_sel = ak.sum(events['Electron']['TIGHTEL'], axis=1) == 1
selection.add('tight_EL',
tight_el_sel) # one electron passing TIGHT criteria
### apply lepton SFs to MC (only applicable to tight leptons)
if 'LeptonSF' in self.corrections.keys():
tight_muons = events['Muon'][tight_mu_sel][(
events['Muon'][tight_mu_sel]['TIGHTMU'] == True)]
muSFs_dict = MCWeights.get_lepton_sf(
year=args.year,
lepton='Muons',
corrections=self.corrections['LeptonSF'],
pt=ak.flatten(tight_muons['pt']),
eta=ak.flatten(tight_muons['eta']))
mu_reco_cen = np.ones(len(events))
mu_reco_err = np.zeros(len(events))
mu_trig_cen = np.ones(len(events))
mu_trig_err = np.zeros(len(events))
mu_reco_cen[tight_mu_sel] = muSFs_dict['RECO_CEN']
mu_reco_err[tight_mu_sel] = muSFs_dict['RECO_ERR']
mu_trig_cen[tight_mu_sel] = muSFs_dict['TRIG_CEN']
mu_trig_err[tight_mu_sel] = muSFs_dict['TRIG_ERR']
mu_evt_weights.add('Lep_RECO',
mu_reco_cen,
mu_reco_err,
mu_reco_err,
shift=True)
mu_evt_weights.add('Lep_TRIG',
mu_trig_cen,
mu_trig_err,
mu_trig_err,
shift=True)
tight_electrons = events['Electron'][tight_el_sel][(
events['Electron'][tight_el_sel]['TIGHTEL'] == True)]
elSFs_dict = MCWeights.get_lepton_sf(
year=args.year,
lepton='Electrons',
corrections=self.corrections['LeptonSF'],
pt=ak.flatten(tight_electrons['pt']),
eta=ak.flatten(tight_electrons['etaSC']))
el_reco_cen = np.ones(len(events))
el_reco_err = np.zeros(len(events))
el_trig_cen = np.ones(len(events))
el_trig_err = np.zeros(len(events))
el_reco_cen[tight_el_sel] = elSFs_dict['RECO_CEN']
el_reco_err[tight_el_sel] = elSFs_dict['RECO_ERR']
el_trig_cen[tight_el_sel] = elSFs_dict['TRIG_CEN']
el_trig_err[tight_el_sel] = elSFs_dict['TRIG_ERR']
el_evt_weights.add('Lep_RECO',
el_reco_cen,
el_reco_err,
el_reco_err,
shift=True)
el_evt_weights.add('Lep_TRIG',
el_trig_cen,
el_trig_err,
el_trig_err,
shift=True)
## build corrected jets and MET
events['Jet'], events['MET'] = IDJet.process_jets(
events, args.year, self.corrections['JetCor'])
# jet selection
passing_jets = objsel.jets_selection(events, year=args.year)
selection.add('passing_jets', passing_jets)
selection.add('jets_3', ak.num(events['SelectedJets']) == 3)
selection.add('btag_pass',
ak.sum(events['SelectedJets'][btag_wps[0]], axis=1) >= 2)
events['SelectedJets'] = events['SelectedJets'][ak.argsort(
events['SelectedJets']['btagDeepB'], ascending=False
)] if btaggers[0] == 'DeepCSV' else events['SelectedJets'][ak.argsort(
events['SelectedJets']['btagDeepFlavB'], ascending=False)]
if self.corrections['BTagSF'] == True:
#set_trace()
deepcsv_cen = np.ones(len(events))
threeJets_cut = selection.require(lep_and_filter_pass=True,
passing_jets=True,
jets_3=True)
deepcsv_3j_wts = self.corrections['BTag_Constructors']['DeepCSV'][
'3Jets'].get_scale_factor(
jets=events['SelectedJets'][threeJets_cut],
passing_cut='DeepCSV' + wps_to_use[0])
deepcsv_cen[threeJets_cut] = ak.prod(deepcsv_3j_wts['central'],
axis=1)
# make dict of btag weights
btag_weights = {
'DeepCSV_CEN': deepcsv_cen,
}
# find gen level particles for ttbar system
genpsel.select(events, mode='NORMAL')
selection.add('semilep', ak.num(events['SL']) > 0)
if 'NNLO_Rewt' in self.corrections.keys():
nnlo_wts = MCWeights.get_nnlo_weights(
self.corrections['NNLO_Rewt'], events)
mu_evt_weights.add(
'%s_reweighting' % self.corrections['NNLO_Rewt']['Var'],
nnlo_wts)
el_evt_weights.add(
'%s_reweighting' % self.corrections['NNLO_Rewt']['Var'],
nnlo_wts)
## fill hists for each region
for lepton in regions.keys():
evt_weights = mu_evt_weights if lepton == 'Muon' else el_evt_weights
cut = selection.all(*regions[lepton])
#set_trace()
if cut.sum() > 0:
leptype = 'MU' if lepton == 'Muon' else 'EL'
if 'loose_or_tight_%s' % leptype in regions[lepton]:
leptons = events[lepton][cut][(
(events[lepton][cut]['TIGHT%s' % leptype] == True) |
(events[lepton][cut]['LOOSE%s' % leptype] == True))]
elif 'tight_%s' % leptype in regions[lepton]:
leptons = events[lepton][cut][(
events[lepton][cut]['TIGHT%s' % leptype] == True)]
elif 'loose_%s' % leptype in regions[lepton]:
leptons = events[lepton][cut][(
events[lepton][cut]['LOOSE%s' % leptype] == True)]
else:
raise ValueError(
"Not sure what lepton type to choose for event")
# get jets and MET
jets, met = events['SelectedJets'][cut], events['SelectedMET'][
cut]
# find matched permutations
mp = ttmatcher.best_match(gen_hyp=events['SL'][cut],
jets=jets,
leptons=leptons,
met=met)
# find best permutations
best_perms = ttpermutator.find_best_permutations(
jets=jets, leptons=leptons, MET=met, btagWP=btag_wps[0])
valid_perms = ak.num(best_perms['TTbar'].pt) > 0
# compare matched per to best perm
bp_status = np.zeros(
cut.size, dtype=int
) # 0 == '' (no gen matching), 1 == 'right', 2 == 'matchable', 3 == 'unmatchable', 4 == 'sl_tau', 5 == 'noslep'
perm_cat_array = compare_matched_best_perms(mp,
best_perms,
njets='3Jets')
bp_status[cut] = perm_cat_array
if ak.any(ak.num(events['SL']['Lepton'].pdgId) != 1):
raise ValueError(
"Number of leptons is incorrect for classifying tau+jets events"
)
sl_tau_evts = ak.where(
np.abs(events['SL']['Lepton'].pdgId) == 15)[0]
bp_status[sl_tau_evts] = 4
## create MT regions
MT = make_vars.MT(leptons, met)
MTHigh = ak.flatten(MT[valid_perms] >= MTcut)
wts = (evt_weights.weight() *
btag_weights['%s_CEN' %
btaggers[0]])[cut][valid_perms][MTHigh]
output = self.make_3j_categories(
acc=output,
leptype=lepton,
permarray=bp_status[cut][valid_perms][MTHigh],
genttbar=events['SL'][cut][valid_perms][MTHigh],
bp=best_perms[valid_perms][MTHigh],
evt_wts=wts)
return output
def process(self, events):
events = events[
ak.num(events.Jet) >
0] #corrects for rare case where there isn't a single jet in event
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet) >= 0
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
# cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
### For FCNC, we want electron -> tightTTH
electron = Collections(ev, "Electron", "tightFCNC").get()
fakeableelectron = Collections(ev, "Electron", "fakeableFCNC").get()
muon = Collections(ev, "Muon", "tightFCNC").get()
fakeablemuon = Collections(ev, "Muon", "fakeableFCNC").get()
##Jets
Jets = events.Jet
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
lepton = fakeablemuon #ak.concatenate([fakeablemuon, fakeableelectron], axis=1)
mt_lep_met = mt(lepton.pt, lepton.phi, ev.MET.pt, ev.MET.phi)
min_mt_lep_met = ak.min(mt_lep_met, axis=1)
selection = PackedSelection()
selection.add('MET<20', (ev.MET.pt < 20))
selection.add('mt<20', min_mt_lep_met < 20)
#selection.add('MET<19', (ev.MET.pt<19) )
selection_reqs = ['MET<20', 'mt<20'] #, 'MET<19']
fcnc_reqs_d = {sel: True for sel in selection_reqs}
fcnc_selection = selection.require(**fcnc_reqs_d)
# define the weight
weight = Weights(len(ev))
if not dataset == 'MuonEG':
# generator weight
weight.add("weight", ev.genWeight)
jets = getJets(
ev, maxEta=2.4, minPt=25, pt_var='pt'
) #& (ak.num(jets[~match(jets, fakeablemuon, deltaRCut=1.0)])>=1)
single_muon_sel = (ak.num(muon) == 1) & (ak.num(fakeablemuon) == 1) | (
ak.num(muon) == 0) & (ak.num(fakeablemuon) == 1)
single_electron_sel = (ak.num(electron) == 1) & (
ak.num(fakeableelectron)
== 1) | (ak.num(electron) == 0) & (ak.num(fakeableelectron) == 1)
fcnc_muon_sel = (ak.num(
jets[~match(jets, fakeablemuon, deltaRCut=1.0)]) >=
1) & fcnc_selection & single_muon_sel
fcnc_electron_sel = (ak.num(
jets[~match(jets, fakeableelectron, deltaRCut=1.0)]) >=
1) & fcnc_selection & single_electron_sel
tight_muon_sel = (ak.num(muon) == 1) & fcnc_muon_sel
loose_muon_sel = (ak.num(fakeablemuon) == 1) & fcnc_muon_sel
tight_electron_sel = (ak.num(electron) == 1) & fcnc_electron_sel
loose_electron_sel = (ak.num(fakeableelectron)
== 1) & fcnc_electron_sel
output['single_mu_fakeable'].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(fakeablemuon[loose_muon_sel].conePt)),
eta=np.abs(
ak.to_numpy(ak.flatten(fakeablemuon[loose_muon_sel].eta))))
output['single_mu'].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(muon[tight_muon_sel].conePt)),
eta=np.abs(ak.to_numpy(ak.flatten(muon[tight_muon_sel].eta))))
output['single_e_fakeable'].fill(
dataset=dataset,
pt=ak.to_numpy(
ak.flatten(fakeableelectron[loose_electron_sel].conePt)),
eta=np.abs(
ak.to_numpy(
ak.flatten(fakeableelectron[loose_electron_sel].eta))))
output['single_e'].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(electron[tight_electron_sel].conePt)),
eta=np.abs(
ak.to_numpy(ak.flatten(electron[tight_electron_sel].eta))))
if self.debug:
#create pandas dataframe for debugging
passed_events = ev[tight_muon_sel]
passed_muons = muon[tight_muon_sel]
event_p = ak.to_pandas(passed_events[["event"]])
event_p["MET_PT"] = passed_events["MET"]["pt"]
event_p["mt"] = min_mt_lep_met[tight_muon_sel]
event_p["num_tight_mu"] = ak.to_numpy(ak.num(muon)[tight_muon_sel])
event_p["num_loose_mu"] = ak.num(fakeablemuon)[tight_muon_sel]
muon_p = ak.to_pandas(
ak.flatten(passed_muons)[[
"pt", "conePt", "eta", "dz", "dxy", "ptErrRel",
"miniPFRelIso_all", "jetRelIsoV2", "jetRelIso",
"jetPtRelv2"
]])
#convert to numpy array for the output
events_array = pd.concat([muon_p, event_p], axis=1)
events_to_add = [6886009]
for e in events_to_add:
tmp_event = ev[ev.event == e]
added_event = ak.to_pandas(tmp_event[["event"]])
added_event["MET_PT"] = tmp_event["MET"]["pt"]
added_event["mt"] = min_mt_lep_met[ev.event == e]
added_event["num_tight_mu"] = ak.to_numpy(
ak.num(muon)[ev.event == e])
added_event["num_loose_mu"] = ak.to_numpy(
ak.num(fakeablemuon)[ev.event == e])
add_muon = ak.to_pandas(
ak.flatten(muon[ev.event == e])[[
"pt", "conePt", "eta", "dz", "dxy", "ptErrRel",
"miniPFRelIso_all", "jetRelIsoV2", "jetRelIso",
"jetPtRelv2"
]])
add_concat = pd.concat([add_muon, added_event], axis=1)
events_array = pd.concat([events_array, add_concat], axis=0)
output['muons_df'] += processor.column_accumulator(
events_array.to_numpy())
return output
def process(self, events):
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet)>=2
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Electrons
electron = Collections(ev, "Electron", "tightFCNC", 0, self.year).get()
electron = electron[(electron.pt > 15) & (np.abs(electron.eta) < 2.4)]
electron = electron[(electron.genPartIdx >= 0)]
electron = electron[(np.abs(electron.matched_gen.pdgId)==11)] #from here on all leptons are gen-matched
electron = electron[( (electron.genPartFlav==1) | (electron.genPartFlav==15) )] #and now they are all prompt
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[leading_electron_idx]
trailing_electron_idx = ak.singletons(ak.argmin(electron.pt, axis=1))
trailing_electron = electron[trailing_electron_idx]
leading_parent = find_first_parent(leading_electron.matched_gen)
trailing_parent = find_first_parent(trailing_electron.matched_gen)
is_flipped = ( ( (electron.matched_gen.pdgId*(-1) == electron.pdgId) | (find_first_parent(electron.matched_gen)*(-1) == electron.pdgId) ) & (np.abs(electron.pdgId) == 11) )
flipped_electron = electron[is_flipped]
flipped_electron = flipped_electron[(ak.fill_none(flipped_electron.pt, 0)>0)]
flipped_electron = flipped_electron[~(ak.is_none(flipped_electron))]
n_flips = ak.num(flipped_electron)
##Muons
muon = Collections(ev, "Muon", "tightFCNC").get()
muon = muon[(muon.pt > 15) & (np.abs(muon.eta) < 2.4)]
muon = muon[(muon.genPartIdx >= 0)]
muon = muon[(np.abs(muon.matched_gen.pdgId)==13)] #from here, all muons are gen-matched
muon = muon[( (muon.genPartFlav==1) | (muon.genPartFlav==15) )] #and now they are all prompt
##Leptons
lepton = ak.concatenate([muon, electron], axis=1)
SSlepton = (ak.sum(lepton.charge, axis=1) != 0) & (ak.num(lepton)==2)
OSlepton = (ak.sum(lepton.charge, axis=1) == 0) & (ak.num(lepton)==2)
emulepton = (ak.num(electron) == 1) & (ak.num(muon) == 1)
no_mumu = (ak.num(muon) <= 1)
leading_lepton_idx = ak.singletons(ak.argmax(lepton.pt, axis=1))
leading_lepton = lepton[leading_lepton_idx]
trailing_lepton_idx = ak.singletons(ak.argmin(lepton.pt, axis=1))
trailing_lepton = lepton[trailing_lepton_idx]
#jets
jet = getJets(ev, minPt=40, maxEta=2.4, pt_var='pt')
jet = jet[ak.argsort(jet.pt, ascending=False)] # need to sort wrt smeared and recorrected jet pt
jet = jet[~match(jet, muon, deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(jet, electron, deltaRCut=0.4)]
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
# setting up the various weights
weight = Weights( len(ev) )
weight2 = Weights( len(ev))
if not dataset=='MuonEG':
# generator weight
weight.add("weight", ev.genWeight)
weight2.add("weight", ev.genWeight)
weight2.add("charge flip", self.charge_flip_ratio.flip_weight(electron))
#selections
filters = getFilters(ev, year=self.year, dataset=dataset)
ss = (SSlepton)
os = (OSlepton)
jet_all = (ak.num(jet) >= 2)
diele = (ak.num(electron) == 2)
emu = (emulepton)
flips = (n_flips == 1)
no_flips = (n_flips == 0)
nmm = no_mumu
selection = PackedSelection()
selection.add('filter', (filters) )
selection.add('ss', ss )
selection.add('os', os )
selection.add('jet', jet_all )
selection.add('ee', diele)
selection.add('emu', emu)
selection.add('flip', flips)
selection.add('nflip', no_flips)
selection.add('no_mumu', nmm)
bl_reqs = ['filter'] + ['jet']
bl_reqs_d = { sel: True for sel in bl_reqs }
baseline = selection.require(**bl_reqs_d)
f_reqs = bl_reqs + ['flip'] + ['ss'] + ['ee']
f_reqs_d = {sel: True for sel in f_reqs}
flip_sel = selection.require(**f_reqs_d)
f2_reqs = bl_reqs + ['flip'] + ['ss'] + ['emu']
f2_reqs_d = {sel: True for sel in f2_reqs}
flip_sel2 = selection.require(**f2_reqs_d)
f3_reqs = bl_reqs + ['flip'] + ['ss'] + ['no_mumu']
f3_reqs_d = {sel: True for sel in f3_reqs}
flip_sel3 = selection.require(**f3_reqs_d)
nf_reqs = bl_reqs + ['nflip'] + ['os'] + ['ee']
nf_reqs_d = {sel: True for sel in nf_reqs}
n_flip_sel = selection.require(**nf_reqs_d)
nf2_reqs = bl_reqs + ['nflip'] + ['os'] + ['emu']
nf2_reqs_d = {sel: True for sel in nf2_reqs}
n_flip_sel2 = selection.require(**nf2_reqs_d)
nf3_reqs = bl_reqs + ['nflip'] + ['os'] + ['no_mumu']
nf3_reqs_d = {sel: True for sel in nf3_reqs}
n_flip_sel3 = selection.require(**nf3_reqs_d)
s_reqs = bl_reqs + ['ss'] + ['no_mumu']
s_reqs_d = { sel: True for sel in s_reqs }
ss_sel = selection.require(**s_reqs_d)
o_reqs = bl_reqs + ['os'] + ['no_mumu']
o_reqs_d = {sel: True for sel in o_reqs }
os_sel = selection.require(**o_reqs_d)
ees_reqs = bl_reqs + ['ss'] + ['ee']
ees_reqs_d = { sel: True for sel in ees_reqs }
eess_sel = selection.require(**ees_reqs_d)
eeo_reqs = bl_reqs + ['os'] + ['ee']
eeo_reqs_d = {sel: True for sel in eeo_reqs }
eeos_sel = selection.require(**eeo_reqs_d)
ems_reqs = bl_reqs + ['ss'] + ['emu']
ems_reqs_d = { sel: True for sel in ems_reqs }
emss_sel = selection.require(**ems_reqs_d)
emo_reqs = bl_reqs + ['os'] + ['emu']
emo_reqs_d = {sel: True for sel in emo_reqs }
emos_sel = selection.require(**emo_reqs_d)
#outputs
output['N_jet'].fill(dataset=dataset, multiplicity=ak.num(jet)[baseline], weight=weight.weight()[baseline])
output['N_ele'].fill(dataset=dataset, multiplicity=ak.num(lepton)[ss_sel], weight=weight.weight()[ss_sel])
output['N_ele2'].fill(dataset=dataset, multiplicity=ak.num(lepton)[os_sel], weight=weight2.weight()[os_sel])
output['electron_flips'].fill(dataset=dataset, multiplicity = n_flips[flip_sel], weight=weight.weight()[flip_sel])
output['electron_flips2'].fill(dataset=dataset, multiplicity = n_flips[n_flip_sel], weight=weight2.weight()[n_flip_sel])
output['electron_flips3'].fill(dataset=dataset, multiplicity = n_flips[flip_sel2], weight=weight.weight()[flip_sel2])
output['electron_flips4'].fill(dataset=dataset, multiplicity = n_flips[n_flip_sel2], weight=weight2.weight()[n_flip_sel2])
output["electron"].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_electron[flip_sel3].pt)),
eta = np.abs(ak.to_numpy(ak.flatten(leading_electron[flip_sel3].eta))),
weight = weight.weight()[flip_sel3]
)
output["electron2"].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_electron[n_flip_sel3].pt)),
eta = np.abs(ak.to_numpy(ak.flatten(leading_electron[n_flip_sel3].eta))),
weight = weight2.weight()[n_flip_sel3]
)
output["flipped_electron"].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_electron[flip_sel].pt)),
eta = np.abs(ak.to_numpy(ak.flatten(leading_electron[flip_sel].eta))),
weight = weight.weight()[flip_sel]
)
output["flipped_electron2"].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_electron[n_flip_sel].pt)),
eta = np.abs(ak.to_numpy(ak.flatten(leading_electron[n_flip_sel].eta))),
weight = weight2.weight()[n_flip_sel]
)
output["flipped_electron3"].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_electron[flip_sel2].pt)),
eta = np.abs(ak.to_numpy(ak.flatten(leading_electron[flip_sel2].eta))),
weight = weight.weight()[flip_sel2]
)
output["flipped_electron4"].fill(
dataset = dataset,
pt = ak.to_numpy(ak.flatten(leading_electron[n_flip_sel2].pt)),
eta = np.abs(ak.to_numpy(ak.flatten(leading_electron[n_flip_sel2].eta))),
weight = weight2.weight()[n_flip_sel2]
)
#output["lepton_parent"].fill(
# dataset = dataset,
# pdgID = np.abs(ak.to_numpy(ak.flatten(leading_parent[ss_sel]))),
# weight = weight.weight()[ss_sel]
#)
#
#output["lepton_parent2"].fill(
# dataset = dataset,
# pdgID = np.abs(ak.to_numpy(ak.flatten(trailing_parent[ss_sel]))),
# weight = weight.weight()[ss_sel]
#)
return output
class Selection:
def __init__(self, **kwargs):
'''
kwargs should be:
ele (loose and tight)
mu
jets: all, central, forward, b-tag
met
'''
self.__dict__.update(kwargs)
## not yet sure whether this should go here, or later
#self.filters = getFilters(self.events, year=self.year, dataset=self.dataset)
def dilep_baseline(self, omit=[], cutflow=None, tight=False, SS=True):
'''
give it a cutflow object if you want it to be filed.
cuts in the omit list will not be applied
'''
self.selection = PackedSelection()
is_dilep = ((ak.num(self.ele) + ak.num(self.mu))==2)
pos_charge = ((ak.sum(self.ele.pdgId, axis=1) + ak.sum(self.mu.pdgId, axis=1))<0)
neg_charge = ((ak.sum(self.ele.pdgId, axis=1) + ak.sum(self.mu.pdgId, axis=1))>0)
lep0pt = ((ak.num(self.ele[(self.ele.pt>25)]) + ak.num(self.mu[(self.mu.pt>25)]))>0)
lep1pt = ((ak.num(self.ele[(self.ele.pt>20)]) + ak.num(self.mu[(self.mu.pt>20)]))>1)
lepveto = ((ak.num(self.ele_veto) + ak.num(self.mu_veto))==2)
dimu = choose(self.mu, 2)
diele = choose(self.ele, 2)
dilep = cross(self.mu, self.ele)
if SS:
is_SS = ( ak.any((dimu['0'].charge * dimu['1'].charge)>0, axis=1) | \
ak.any((diele['0'].charge * diele['1'].charge)>0, axis=1) | \
ak.any((dilep['0'].charge * dilep['1'].charge)>0, axis=1) )
else:
is_OS = ( ak.any((dimu['0'].charge * dimu['1'].charge)<0, axis=1) | \
ak.any((diele['0'].charge * diele['1'].charge)<0, axis=1) | \
ak.any((dilep['0'].charge * dilep['1'].charge)<0, axis=1) )
lepton = ak.concatenate([self.ele, self.mu], axis=1)
lepton_pdgId_pt_ordered = ak.fill_none(
ak.pad_none(
lepton[ak.argsort(lepton.pt, ascending=False)].pdgId, 2, clip=True),
0)
triggers = getTriggers(self.events,
ak.flatten(lepton_pdgId_pt_ordered[:,0:1]),
ak.flatten(lepton_pdgId_pt_ordered[:,1:2]), year=self.year, dataset=self.dataset)
ht = ak.sum(self.jet_all.pt, axis=1)
st = self.met.pt + ht + ak.sum(self.mu.pt, axis=1) + ak.sum(self.ele.pt, axis=1)
self.selection.add('lepveto', lepveto)
self.selection.add('dilep', is_dilep)
#self.selection.add('filter', self.filters)
self.selection.add('trigger', triggers)
self.selection.add('p_T(lep0)>25', lep0pt)
self.selection.add('p_T(lep1)>20', lep1pt)
if SS:
self.selection.add('SS', is_SS )
else:
self.selection.add('OS', is_OS )
self.selection.add('N_jet>3', (ak.num(self.jet_all)>3) )
self.selection.add('N_jet>4', (ak.num(self.jet_all)>4) )
self.selection.add('N_central>2', (ak.num(self.jet_central)>2) )
self.selection.add('N_central>3', (ak.num(self.jet_central)>3) )
self.selection.add('N_btag>0', (ak.num(self.jet_btag)>0) )
self.selection.add('N_fwd>0', (ak.num(self.jet_fwd)>0) )
self.selection.add('MET>30', (self.met.pt>30) )
self.selection.add('MET>50', (self.met.pt>50) )
self.selection.add('ST>600', (st>600) )
ss_reqs = [
# 'filter',
'lepveto',
'dilep',
'p_T(lep0)>25',
'p_T(lep1)>20',
'trigger',
'SS' if SS else 'OS',
'N_jet>3',
'N_central>2',
'N_btag>0',
'MET>30',
'N_fwd>0',
]
if tight:
ss_reqs += [
'N_jet>4',
'N_central>3',
'ST>600',
'MET>50',
#'delta_eta',
]
ss_reqs_d = { sel: True for sel in ss_reqs if not sel in omit }
ss_selection = self.selection.require(**ss_reqs_d)
if cutflow:
#
cutflow_reqs_d = {}
for req in ss_reqs:
cutflow_reqs_d.update({req: True})
cutflow.addRow( req, self.selection.require(**cutflow_reqs_d) )
return ss_selection
def trilep_baseline(self, omit=[], cutflow=None, tight=False):
'''
give it a cutflow object if you want it to be filed.
cuts in the omit list will not be applied
'''
self.selection = PackedSelection()
is_trilep = ((ak.num(self.ele) + ak.num(self.mu))==3)
pos_charge = ((ak.sum(self.ele.pdgId, axis=1) + ak.sum(self.mu.pdgId, axis=1))<0)
neg_charge = ((ak.sum(self.ele.pdgId, axis=1) + ak.sum(self.mu.pdgId, axis=1))>0)
lep0pt = ((ak.num(self.ele[(self.ele.pt>25)]) + ak.num(self.mu[(self.mu.pt>25)]))>0)
lep1pt = ((ak.num(self.ele[(self.ele.pt>20)]) + ak.num(self.mu[(self.mu.pt>20)]))>1)
lepveto = ((ak.num(self.ele_veto) + ak.num(self.mu_veto))==3)
dimu = choose(self.mu, 2)
diele = choose(self.ele, 2)
dilep = cross(self.mu, self.ele)
OS_dimu = dimu[(dimu['0'].charge*dimu['1'].charge < 0)]
OS_diele = diele[(diele['0'].charge*diele['1'].charge < 0)]
offZ = (ak.all(abs(OS_dimu.mass-91.2)>10, axis=1) & ak.all(abs(OS_diele.mass-91.2)>10, axis=1))
lepton = ak.concatenate([self.ele, self.mu], axis=1)
lepton_pdgId_pt_ordered = ak.fill_none(ak.pad_none(lepton[ak.argsort(lepton.pt, ascending=False)].pdgId, 2, clip=True), 0)
triggers = getTriggers(self.events,
ak.flatten(lepton_pdgId_pt_ordered[:,0:1]),
ak.flatten(lepton_pdgId_pt_ordered[:,1:2]), year=self.year, dataset=self.dataset)
ht = ak.sum(self.jet_all.pt, axis=1)
st = self.met.pt + ht + ak.sum(self.mu.pt, axis=1) + ak.sum(self.ele.pt, axis=1)
self.selection.add('lepveto', lepveto)
self.selection.add('trilep', is_trilep)
self.selection.add('filter', self.filters)
self.selection.add('trigger', triggers)
self.selection.add('p_T(lep0)>25', lep0pt)
self.selection.add('p_T(lep1)>20', lep1pt)
self.selection.add('N_jet>2', (ak.num(self.jet_all)>2) )
self.selection.add('N_jet>3', (ak.num(self.jet_all)>3) )
self.selection.add('N_central>1', (ak.num(self.jet_central)>1) )
self.selection.add('N_central>2', (ak.num(self.jet_central)>2) )
self.selection.add('N_btag>0', (ak.num(self.jet_btag)>0) )
self.selection.add('N_fwd>0', (ak.num(self.jet_fwd)>0) )
self.selection.add('MET>50', (self.met.pt>50) )
self.selection.add('ST>600', (st>600) )
self.selection.add('offZ', offZ )
reqs = [
'filter',
'lepveto',
'trilep',
'p_T(lep0)>25',
'p_T(lep1)>20',
'trigger',
'offZ',
'MET>50',
'N_jet>2',
'N_central>1',
'N_btag>0',
'N_fwd>0',
]
if tight:
reqs += [
'N_jet>3',
'N_central>2',
'ST>600',
#'MET>50',
#'delta_eta',
]
reqs_d = { sel: True for sel in reqs if not sel in omit }
selection = self.selection.require(**reqs_d)
self.reqs = [ sel for sel in reqs if not sel in omit ]
if cutflow:
#
cutflow_reqs_d = {}
for req in reqs:
cutflow_reqs_d.update({req: True})
cutflow.addRow( req, self.selection.require(**cutflow_reqs_d) )
return selection
def process(self, events):
output = self._accumulator.identity()
dataset_name = events.metadata['dataset']
output["total_events"][dataset_name] += events.__len__()
# HLT selection
HLT_mask = []
if year == "2016":
if "SingleMuon" in dataset_name: #this does not work, as the name of file which is under processing is unknown
if "2016B2" in dataset_name:
HLT_mask = events.HLT.IsoMu24 | events.HLT.IsoTkMu24 | events.HLT.Mu50
else:
HLT_mask = events.HLT.IsoMu24 | events.HLT.IsoTkMu24 | events.HLT.Mu50 | events.HLT.TkMu50
else: #https://twiki.cern.ch/twiki/bin/view/CMS/HLTPathsRunIIList
if "2016B2" in dataset_name:
HLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
elif "2016H" in dataset_name:
HLT_mask = events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
else:
HLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
if year == "2017":
if "SingleMuon" in dataset_name:
if "2017B" in dataset_name:
HLT_mask = events.HLT.IsoMu27 | events.HLT.Mu50
else:
HLT_mask = events.HLT.IsoMu27 | events.HLT.Mu50 | events.HLT.OldMu100 | events.HLT.TkMu100
else:
HLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
if year == "2018":
if "SingleMuon" in dataset_name:
HLT_mask = events.HLT.IsoMu24 | events.HLT.Mu50 | events.HLT.OldMu100 | events.HLT.TkMu100
else:
HLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
# Require 3 jets
jet_mask = (events.Jet.pt > 30.) & (abs(events.Jet.eta) <
2.5) & (events.Jet.isTight)
event_mask = (awk.sum(jet_mask, axis=1) >= 3)
event_mask = event_mask & HLT_mask
events_3j = events[event_mask]
# Reduce jet mask to only events with 3 good jets
jet_mask = jet_mask[event_mask]
# Array of the jets to consider for trijet resonance
selected_jets = events_3j.Jet[jet_mask][:, :3]
# Pairs of jets
#pairs = awk.argcombinations(selected_jets, 2)
#jet_i, jet_j = awk.unzip(pairs)
pairs = [(0, 1), (1, 2), (2, 0)]
jet_i, jet_j = zip(*pairs) # Returns [0, 1, 2] , [1, 2, 0]
m_ij = (selected_jets[:, jet_i] + selected_jets[:, jet_j]).mass
dR_ij = selected_jets[:, jet_i].delta_r(selected_jets[:, jet_j])
dEta_ij = abs(selected_jets[:, jet_i].eta -
selected_jets[:, jet_j].eta)
jet_k = [2, 0, 1]
dR_i_jk = selected_jets[:, jet_i].delta_r(selected_jets[:, jet_j] +
selected_jets[:, jet_k])
dEta_i_jk = abs(selected_jets[:, jet_i].eta -
(selected_jets[:, jet_j] +
selected_jets[:, jet_k]).eta)
dPhi_i_jk = abs(selected_jets[:, jet_i].phi -
(selected_jets[:, jet_j] +
selected_jets[:, jet_k]).phi)
m3j = selected_jets.sum().mass
pt_i_overM = selected_jets.pt / m3j
m_01_overM = m_ij[:, 0] / m3j
m_12_overM = m_ij[:, 1] / m3j
m_20_overM = m_ij[:, 2] / m3j
dPtoverM_0_12 = abs(selected_jets[:, 0].pt -
(selected_jets[:, 1] +
selected_jets[:, 2]).pt) / m3j
dPtoverM_1_20 = abs(selected_jets[:, 1].pt -
(selected_jets[:, 2] +
selected_jets[:, 0]).pt) / m3j
dPtoverM_2_01 = abs(selected_jets[:, 2].pt -
(selected_jets[:, 0] +
selected_jets[:, 1]).pt) / m3j
# Event selection masks
selection_masks = {}
# Pre-selection
selection = PackedSelection()
selection.add("Dummy", m3j > 000)
sel_mask = selection.require(
**{name: True
for name in selection.names})
selection_masks["Pre-selection"] = sel_mask
# HLT_trigger (this is already done at the beginning)
# if year == "2016":
# JetHLT_mask = []
# if "2016B2" in dataset_name:
# JetHLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
# elif "2016H" in dataset_name:
# JetHLT_mask = events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
# else:
# JetHLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
# selection_masks["JetHLT"] = JetHLT_mask[event_mask]
# if year == "2017":
# JetHLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
# selection_masks["JetHLT"] = JetHLT_mask[event_mask]
# if year == "2018":
# JetHLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
# selection_masks["JetHLT"] = JetHLT_mask[event_mask]
# Fill histograms
for selection, selection_mask in selection_masks.items():
output["mjjj"].fill(dataset=dataset_name,
selection=selection,
mjjj=m3j[selection_mask])
output["m_ij"].fill(dataset=dataset_name,
selection=selection,
m_01=m_ij[:, 0][selection_mask],
m_12=m_ij[:, 1][selection_mask],
m_20=m_ij[:, 2][selection_mask])
output["dR_ij"].fill(dataset=dataset_name,
selection=selection,
dR_01=dR_ij[:, 0][selection_mask],
dR_12=dR_ij[:, 1][selection_mask],
dR_20=dR_ij[:, 2][selection_mask])
output["dEta_ij"].fill(dataset=dataset_name,
selection=selection,
dEta_01=dEta_ij[:, 0][selection_mask],
dEta_12=dEta_ij[:, 1][selection_mask],
dEta_20=dEta_ij[:, 2][selection_mask])
output["moverM_ij"].fill(dataset=dataset_name,
selection=selection,
moverM_01=m_01_overM[selection_mask],
moverM_12=m_12_overM[selection_mask],
moverM_20=m_20_overM[selection_mask])
output["pt_i"].fill(dataset=dataset_name,
selection=selection,
pt_0=selected_jets[:, 0][selection_mask].pt,
pt_1=selected_jets[:, 1][selection_mask].pt,
pt_2=selected_jets[:, 2][selection_mask].pt)
output["eta_i"].fill(dataset=dataset_name,
selection=selection,
eta_0=selected_jets[:, 0][selection_mask].eta,
eta_1=selected_jets[:, 1][selection_mask].eta,
eta_2=selected_jets[:, 2][selection_mask].eta)
output["ptoverM_i"].fill(dataset=dataset_name,
selection=selection,
ptoverM_0=pt_i_overM[:,
0][selection_mask],
ptoverM_1=pt_i_overM[:,
1][selection_mask],
ptoverM_2=pt_i_overM[:,
2][selection_mask])
output["dR_i_jk"].fill(dataset=dataset_name,
selection=selection,
dR_0_12=dR_i_jk[:, 0][selection_mask],
dR_1_20=dR_i_jk[:, 1][selection_mask],
dR_2_01=dR_i_jk[:, 2][selection_mask])
output["dEta_i_jk"].fill(dataset=dataset_name,
selection=selection,
dEta_0_12=dEta_i_jk[:, 0][selection_mask],
dEta_1_20=dEta_i_jk[:, 1][selection_mask],
dEta_2_01=dEta_i_jk[:, 2][selection_mask])
output["dPhi_i_jk"].fill(dataset=dataset_name,
selection=selection,
dPhi_0_12=dPhi_i_jk[:, 0][selection_mask],
dPhi_1_20=dPhi_i_jk[:, 1][selection_mask],
dPhi_2_01=dPhi_i_jk[:, 2][selection_mask])
output["dPtoverM_i_jk"].fill(
dataset=dataset_name,
selection=selection,
dPtoverM_0_12=dPtoverM_0_12[selection_mask],
dPtoverM_1_20=dPtoverM_1_20[selection_mask],
dPtoverM_2_01=dPtoverM_2_01[selection_mask])
pt_i_overM_2fill = pt_i_overM[selection_mask]
dR_ij_2fill = dR_ij[selection_mask]
dEta_ij_2fill = dEta_ij[selection_mask]
dR_i_jk_2fill = dR_i_jk[selection_mask]
dEta_i_jk_2fill = dEta_i_jk[selection_mask]
dPhi_i_jk_2fill = dPhi_i_jk[selection_mask]
dPtoverM_0_12_2fill = dPtoverM_0_12[selection_mask]
dPtoverM_1_20_2fill = dPtoverM_1_20[selection_mask]
dPtoverM_2_01_2fill = dPtoverM_2_01[selection_mask]
selected_jets_2fill = selected_jets[selection_mask]
max_pt_overM_2fill = awk.max(pt_i_overM_2fill, axis=1)
min_pt_overM_2fill = awk.min(pt_i_overM_2fill, axis=1)
max_dR_2fill = awk.max(dR_ij_2fill, axis=1)
max_dEta_2fill = awk.max(dEta_ij_2fill, axis=1)
min_dR_2fill = awk.min(dR_ij_2fill, axis=1)
min_dEta_2fill = awk.min(dEta_ij_2fill, axis=1)
min_pt_2fill = awk.min(selected_jets_2fill.pt, axis=1)
max_eta_2fill = awk.max(abs(selected_jets_2fill.eta), axis=1)
max_dR_i_jk_2fill = awk.max(dR_i_jk_2fill, axis=1)
min_dR_i_jk_2fill = awk.min(dR_i_jk_2fill, axis=1)
max_dEta_i_jk_2fill = awk.max(dEta_i_jk_2fill, axis=1)
min_dEta_i_jk_2fill = awk.min(dEta_i_jk_2fill, axis=1)
max_dPhi_i_jk_2fill = awk.max(dPhi_i_jk_2fill, axis=1)
min_dPhi_i_jk_2fill = awk.min(dPhi_i_jk_2fill, axis=1)
max_dPtoverM_i_jk_2fill = []
min_dPtoverM_i_jk_2fill = []
for pair in zip(dPtoverM_0_12_2fill, dPtoverM_1_20_2fill,
dPtoverM_2_01_2fill):
max_dPtoverM_i_jk_2fill.append(max(pair))
min_dPtoverM_i_jk_2fill.append(min(pair))
max_dPtoverM_i_jk_2fill = np.array(max_dPtoverM_i_jk_2fill)
min_dPtoverM_i_jk_2fill = np.array(min_dPtoverM_i_jk_2fill)
max_pt_overM_2fill = awk.fill_none(max_pt_overM_2fill, -99)
min_pt_overM_2fill = awk.fill_none(min_pt_overM_2fill, -99)
max_dR_2fill = awk.fill_none(max_dR_2fill, -99)
max_dEta_2fill = awk.fill_none(max_dEta_2fill, -99)
min_dR_2fill = awk.fill_none(min_dR_2fill, -99)
min_dEta_2fill = awk.fill_none(min_dEta_2fill, -99)
min_pt_2fill = awk.fill_none(min_pt_2fill, -99)
max_eta_2fill = awk.fill_none(max_eta_2fill, -99)
max_dR_i_jk_2fill = awk.fill_none(max_dR_i_jk_2fill, -99)
min_dR_i_jk_2fill = awk.fill_none(min_dR_i_jk_2fill, -99)
max_dEta_i_jk_2fill = awk.fill_none(max_dEta_i_jk_2fill, -99)
min_dEta_i_jk_2fill = awk.fill_none(min_dEta_i_jk_2fill, -99)
max_dPhi_i_jk_2fill = awk.fill_none(max_dPhi_i_jk_2fill, -99)
min_dPhi_i_jk_2fill = awk.fill_none(min_dPhi_i_jk_2fill, -99)
output["max_dR"].fill(dataset=dataset_name,
selection=selection,
max_dR=max_dR_2fill)
output["max_dEta"].fill(dataset=dataset_name,
selection=selection,
max_dEta=max_dEta_2fill)
output["min_dR"].fill(dataset=dataset_name,
selection=selection,
min_dR=min_dR_2fill)
output["min_dEta"].fill(dataset=dataset_name,
selection=selection,
min_dEta=min_dEta_2fill)
output["min_pt"].fill(dataset=dataset_name,
selection=selection,
min_pt=min_pt_2fill)
output["max_eta"].fill(dataset=dataset_name,
selection=selection,
max_eta=max_eta_2fill)
output["max_ptoverM"].fill(dataset=dataset_name,
selection=selection,
max_ptoverM=max_pt_overM_2fill)
output["min_ptoverM"].fill(dataset=dataset_name,
selection=selection,
min_ptoverM=min_pt_overM_2fill)
output["max_dR_j_jj"].fill(dataset=dataset_name,
selection=selection,
max_dR_j_jj=max_dR_i_jk_2fill)
output["max_dEta_j_jj"].fill(dataset=dataset_name,
selection=selection,
max_dEta_j_jj=max_dEta_i_jk_2fill)
output["max_dPhi_j_jj"].fill(dataset=dataset_name,
selection=selection,
max_dPhi_j_jj=max_dPhi_i_jk_2fill)
output["max_dPtoverM_j_jj"].fill(
dataset=dataset_name,
selection=selection,
max_dPtoverM_j_jj=max_dPtoverM_i_jk_2fill)
output["min_dR_j_jj"].fill(dataset=dataset_name,
selection=selection,
min_dR_j_jj=min_dR_i_jk_2fill)
output["min_dEta_j_jj"].fill(dataset=dataset_name,
selection=selection,
min_dEta_j_jj=min_dEta_i_jk_2fill)
output["min_dPhi_j_jj"].fill(dataset=dataset_name,
selection=selection,
min_dPhi_j_jj=min_dPhi_i_jk_2fill)
output["min_dPtoverM_j_jj"].fill(
dataset=dataset_name,
selection=selection,
min_dPtoverM_j_jj=min_dPtoverM_i_jk_2fill)
return output
def process(self, events):
output = self._accumulator.identity()
dataset_name = events.metadata['dataset']
output["total_events"][dataset_name] += events.__len__()
# Initialize dict accumulators, if have not been initialized
for jet in [0, 1, 2]:
if dataset_name not in output[f"eta_{jet}_final"].keys():
output[f"eta_{jet}_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"ptoverM_{jet}_final"].keys():
output[f"ptoverM_{jet}_final"][dataset_name] = processor.column_accumulator(np.array([]))
for pair in [(0, 1), (1, 2), (2, 0)]:
if dataset_name not in output[f"dEta_{pair[0]}{pair[1]}_final"].keys():
output[f"dEta_{pair[0]}{pair[1]}_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dR_{pair[0]}{pair[1]}_final"].keys():
output[f"dR_{pair[0]}{pair[1]}_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"moverM_{pair[0]}{pair[1]}_final"].keys():
output[f"moverM_{pair[0]}{pair[1]}_final"][dataset_name] = processor.column_accumulator(np.array([]))
for pair in [(0, 1, 2), (1, 2, 0), (2, 0, 1)]:
if dataset_name not in output[f"dR_{pair[0]}_{pair[1]}{pair[2]}_final"].keys():
output[f"dR_{pair[0]}_{pair[1]}{pair[2]}_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dEta_{pair[0]}_{pair[1]}{pair[2]}_final"].keys():
output[f"dEta_{pair[0]}_{pair[1]}{pair[2]}_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"Phi_{pair[0]}_{pair[1]}{pair[2]}_final"].keys():
output[f"Phi_{pair[0]}_{pair[1]}{pair[2]}_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dPtoverM_{pair[0]}_{pair[1]}{pair[2]}_final"].keys():
output[f"dPtoverM_{pair[0]}_{pair[1]}{pair[2]}_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"ptoverM_max_final"].keys():
output[f"ptoverM_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"ptoverM_min_final"].keys():
output[f"ptoverM_min_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"eta_max_final"].keys():
output[f"eta_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dR_max_final"].keys():
output[f"dR_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dR_min_final"].keys():
output[f"dR_min_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dEta_max_final"].keys():
output[f"dEta_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dEta_min_final"].keys():
output[f"dEta_min_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dR_j_jj_max_final"].keys():
output[f"dR_j_jj_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dR_j_jj_min_final"].keys():
output[f"dR_j_jj_min_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dEta_j_jj_max_final"].keys():
output[f"dEta_j_jj_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dEta_j_jj_min_final"].keys():
output[f"dEta_j_jj_min_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dPhi_j_jj_max_final"].keys():
output[f"dPhi_j_jj_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dPhi_j_jj_min_final"].keys():
output[f"dPhi_j_jj_min_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dPtoverM_j_jj_max_final"].keys():
output[f"dPtoverM_j_jj_max_final"][dataset_name] = processor.column_accumulator(np.array([]))
if dataset_name not in output[f"dPtoverM_j_jj_min_final"].keys():
output[f"dPtoverM_j_jj_min_final"][dataset_name] = processor.column_accumulator(np.array([]))
# HLT selection
HLT_mask = []
if year == "2016":
if "SingleMuon" in dataset_name: #this does not work, as the name of file which is under processing is unknown
if "2016B2" in dataset_name:
HLT_mask = events.HLT.IsoMu24 | events.HLT.IsoTkMu24 | events.HLT.Mu50
else:
HLT_mask = events.HLT.IsoMu24 | events.HLT.IsoTkMu24 | events.HLT.Mu50 | events.HLT.TkMu50
else: #https://twiki.cern.ch/twiki/bin/view/CMS/HLTPathsRunIIList
if "2016B2" in dataset_name:
HLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
elif "2016H" in dataset_name:
HLT_mask = events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
else:
HLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
if year == "2017":
if "SingleMuon" in dataset_name:
if "2017B" in dataset_name:
HLT_mask = events.HLT.IsoMu27 | events.HLT.Mu50
else:
HLT_mask = events.HLT.IsoMu27 | events.HLT.Mu50 | events.HLT.OldMu100 | events.HLT.TkMu100
else:
HLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
if year == "2018":
if "SingleMuon" in dataset_name:
HLT_mask = events.HLT.IsoMu24 | events.HLT.Mu50 | events.HLT.OldMu100 | events.HLT.TkMu100
else:
HLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
# Require 3 jets
jet_mask = (events.Jet.pt > 30.) & (abs(events.Jet.eta) < 2.5) & (events.Jet.isTight)
event_mask = (awk.sum(jet_mask, axis=1) >= 3)
event_mask = event_mask & HLT_mask
events_3j = events[event_mask]
# Reduce jet mask to only events with 3 good jets
jet_mask = jet_mask[event_mask]
# Array of the jets to consider for trijet resonance
selected_jets = events_3j.Jet[jet_mask][:, :3]
# Pairs of jets
#pairs = awk.argcombinations(selected_jets, 2)
#jet_i, jet_j = awk.unzip(pairs)
pairs = [(0, 1), (1, 2), (2, 0)]
jet_i, jet_j = zip(*pairs) # Returns [0, 1, 2] , [1, 2, 0]
m_ij = (selected_jets[:, jet_i] + selected_jets[:, jet_j]).mass
dR_ij = selected_jets[:, jet_i].delta_r(selected_jets[:, jet_j])
dEta_ij = abs(selected_jets[:, jet_i].eta - selected_jets[:, jet_j].eta)
jet_k = [2, 0, 1]
dR_i_jk = selected_jets[:, jet_i].delta_r(selected_jets[:, jet_j] + selected_jets[:, jet_k])
dEta_i_jk = abs(selected_jets[:, jet_i].eta - (selected_jets[:, jet_j] + selected_jets[:, jet_k]).eta)
dPhi_i_jk = abs(selected_jets[:, jet_i].phi - (selected_jets[:, jet_j] + selected_jets[:, jet_k]).phi)
m3j = selected_jets.sum().mass
pt_i_overM = selected_jets.pt / m3j
m_01_overM = m_ij[:,0] / m3j
m_12_overM = m_ij[:,1] / m3j
m_20_overM = m_ij[:,2] / m3j
dPtoverM_0_12 = abs(selected_jets[:, 0].pt - (selected_jets[:, 1] + selected_jets[:, 2]).pt) / m3j
dPtoverM_1_20 = abs(selected_jets[:, 1].pt - (selected_jets[:, 2] + selected_jets[:, 0]).pt) / m3j
dPtoverM_2_01 = abs(selected_jets[:, 2].pt - (selected_jets[:, 0] + selected_jets[:, 1]).pt) / m3j
# Event selection masks
# selection_masks = {}
# Pre-selection
selection = PackedSelection()
selection.add("Dummy", m3j > 000)
sel_mask = selection.require(**{name: True for name in selection.names})
# selection_masks["Pre-selection"] = sel_mask
output["selected_events"][dataset_name] += events_3j[sel_mask].__len__()
for jet in [0, 1, 2]:
output[f"eta_{jet}_final"][dataset_name] += processor.column_accumulator(np.array(selected_jets[:, jet][sel_mask].eta))
output[f"ptoverM_{jet}_final"][dataset_name] += processor.column_accumulator(np.array(pt_i_overM[:, jet][sel_mask]))
for pair in [(0, 1), (1, 2), (2, 0)]:
output[f"dEta_{pair[0]}{pair[1]}_final"][dataset_name] += processor.column_accumulator(np.array(dEta_ij[:, pair[0]][sel_mask]))
output[f"dR_{pair[0]}{pair[1]}_final"][dataset_name] += processor.column_accumulator(np.array(dR_ij[:, pair[0]][sel_mask]))
output[f"moverM_01_final"][dataset_name] += processor.column_accumulator(np.array(m_01_overM[sel_mask]))
output[f"moverM_12_final"][dataset_name] += processor.column_accumulator(np.array(m_12_overM[sel_mask]))
output[f"moverM_20_final"][dataset_name] += processor.column_accumulator(np.array(m_20_overM[sel_mask]))
for pair in [(0, 1, 2), (1, 2, 0), (2, 0, 1)]:
output[f"dR_{pair[0]}_{pair[1]}{pair[2]}_final"][dataset_name] += processor.column_accumulator(np.array(dR_i_jk[:, pair[0]][sel_mask]))
output[f"dEta_{pair[0]}_{pair[1]}{pair[2]}_final"][dataset_name] += processor.column_accumulator(np.array(dEta_i_jk[:, pair[0]][sel_mask]))
output[f"Phi_{pair[0]}_{pair[1]}{pair[2]}_final"][dataset_name] += processor.column_accumulator(np.array(dPhi_i_jk[:, pair[0]][sel_mask]))
output[f"dPtoverM_0_12_final"][dataset_name] += processor.column_accumulator(np.array(dPtoverM_0_12[sel_mask]))
output[f"dPtoverM_1_20_final"][dataset_name] += processor.column_accumulator(np.array(dPtoverM_1_20[sel_mask]))
output[f"dPtoverM_2_01_final"][dataset_name] += processor.column_accumulator(np.array(dPtoverM_2_01[sel_mask]))
max_pt_overM_2fill = awk.max(pt_i_overM[sel_mask], axis=1)
min_pt_overM_2fill = awk.min(pt_i_overM[sel_mask], axis=1)
max_dR_2fill = awk.max(dR_ij[sel_mask], axis=1)
max_dEta_2fill = awk.max(dEta_ij[sel_mask], axis=1)
min_dR_2fill = awk.min(dR_ij[sel_mask], axis=1)
min_dEta_2fill = awk.min(dEta_ij[sel_mask], axis=1)
min_pt_2fill = awk.min(selected_jets[sel_mask].pt, axis=1)
max_eta_2fill = awk.max(abs(selected_jets[sel_mask].eta), axis=1)
max_dR_i_jk_2fill = awk.max(dR_i_jk[sel_mask], axis=1)
min_dR_i_jk_2fill = awk.min(dR_i_jk[sel_mask], axis=1)
max_dEta_i_jk_2fill = awk.max(dEta_i_jk[sel_mask], axis=1)
min_dEta_i_jk_2fill = awk.min(dEta_i_jk[sel_mask], axis=1)
max_dPhi_i_jk_2fill = awk.max(dPhi_i_jk[sel_mask], axis=1)
min_dPhi_i_jk_2fill = awk.min(dPhi_i_jk[sel_mask], axis=1)
max_dPtoverM_i_jk_2fill = []
min_dPtoverM_i_jk_2fill = []
dPtoverM_0_12_2fill = dPtoverM_0_12[sel_mask]
dPtoverM_1_20_2fill = dPtoverM_1_20[sel_mask]
dPtoverM_2_01_2fill = dPtoverM_2_01[sel_mask]
for pair in zip(dPtoverM_0_12_2fill, dPtoverM_1_20_2fill, dPtoverM_2_01_2fill):
max_dPtoverM_i_jk_2fill.append(max(pair))
min_dPtoverM_i_jk_2fill.append(min(pair))
max_pt_overM_2fill = awk.fill_none(max_pt_overM_2fill, -99)
min_pt_overM_2fill = awk.fill_none(min_pt_overM_2fill, -99)
max_dR_2fill = awk.fill_none(max_dR_2fill, -99)
max_dEta_2fill = awk.fill_none(max_dEta_2fill, -99)
min_dR_2fill = awk.fill_none(min_dR_2fill, -99)
min_dEta_2fill = awk.fill_none(min_dEta_2fill, -99)
min_pt_2fill = awk.fill_none(min_pt_2fill, -99)
max_eta_2fill = awk.fill_none(max_eta_2fill, -99)
max_dR_i_jk_2fill = awk.fill_none(max_dR_i_jk_2fill, -99)
min_dR_i_jk_2fill = awk.fill_none(min_dR_i_jk_2fill, -99)
max_dEta_i_jk_2fill = awk.fill_none(max_dEta_i_jk_2fill, -99)
min_dEta_i_jk_2fill = awk.fill_none(min_dEta_i_jk_2fill, -99)
max_dPhi_i_jk_2fill = awk.fill_none(max_dPhi_i_jk_2fill, -99)
min_dPhi_i_jk_2fill = awk.fill_none(min_dPhi_i_jk_2fill, -99)
output[f"ptoverM_max_final"][dataset_name] += processor.column_accumulator(np.array(max_pt_overM_2fill))
output[f"ptoverM_min_final"][dataset_name] += processor.column_accumulator(np.array(min_pt_overM_2fill))
output[f"eta_max_final"][dataset_name] += processor.column_accumulator(np.array(max_eta_2fill))
output[f"dR_max_final"][dataset_name] += processor.column_accumulator(np.array(max_dR_2fill))
output[f"dR_min_final"][dataset_name] += processor.column_accumulator(np.array(min_dR_2fill))
output[f"dEta_max_final"][dataset_name] += processor.column_accumulator(np.array(max_dEta_2fill))
output[f"dEta_min_final"][dataset_name] += processor.column_accumulator(np.array(min_dEta_2fill))
output[f"dR_j_jj_max_final"][dataset_name] += processor.column_accumulator(np.array(max_dR_i_jk_2fill))
output[f"dR_j_jj_min_final"][dataset_name] += processor.column_accumulator(np.array(min_dR_i_jk_2fill))
output[f"dEta_j_jj_max_final"][dataset_name] += processor.column_accumulator(np.array(max_dEta_i_jk_2fill))
output[f"dEta_j_jj_min_final"][dataset_name] += processor.column_accumulator(np.array(min_dEta_i_jk_2fill))
output[f"dPhi_j_jj_max_final"][dataset_name] += processor.column_accumulator(np.array(max_dPhi_i_jk_2fill))
output[f"dPhi_j_jj_min_final"][dataset_name] += processor.column_accumulator(np.array(min_dPhi_i_jk_2fill))
output[f"dPtoverM_j_jj_max_final"][dataset_name] += processor.column_accumulator(np.array(max_dPtoverM_i_jk_2fill))
output[f"dPtoverM_j_jj_min_final"][dataset_name] += processor.column_accumulator(np.array(min_dPtoverM_i_jk_2fill))
return output
def process(self, events):
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet) >= 0
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Muons
muon = ev.Muon
## Electrons
electron = ev.Electron
## Merge electrons and muons - this should work better now in ak1
dilepton = cross(muon, electron)
SSlepton = ak.any((dilepton['0'].charge * dilepton['1'].charge) > 0,
axis=1)
lepton = ak.concatenate([muon, electron], axis=1)
leading_lepton_idx = ak.singletons(ak.argmax(lepton.pt, axis=1))
leading_lepton = lepton[leading_lepton_idx]
trailing_lepton_idx = ak.singletons(ak.argmin(lepton.pt, axis=1))
trailing_lepton = lepton[trailing_lepton_idx]
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
# define the weight
weight = Weights(len(ev))
if not dataset == 'MuonEG':
# generator weight
weight.add("weight", ev.genWeight)
filters = getFilters(ev, year=self.year, dataset=dataset)
dilep = ((ak.num(electron) + ak.num(muon)) == 2)
selection = PackedSelection()
selection.add('dilep', dilep)
selection.add('filter', (filters))
bl_reqs = ['dilep', 'filter']
bl_reqs_d = {sel: True for sel in bl_reqs}
baseline = selection.require(**bl_reqs_d)
output['N_ele'].fill(dataset=dataset,
multiplicity=ak.num(electron)[baseline],
weight=weight.weight()[baseline])
output['N_mu'].fill(dataset=dataset,
multiplicity=ak.num(muon)[baseline],
weight=weight.weight()[baseline])
output['lead_lep'].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_lepton[baseline].pt)),
eta=ak.to_numpy(ak.flatten(leading_lepton[baseline].eta)),
phi=ak.to_numpy(ak.flatten(leading_lepton[baseline].phi)),
weight=weight.weight()[baseline])
return output
def process(self, events):
output = self.accumulator.identity()
# use a very loose preselection to filter the events
presel = ak.num(events.Jet) > 2
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Muons
muon = Collections(ev, "Muon", "tightTTH").get()
vetomuon = Collections(ev, "Muon", "vetoTTH").get()
dimuon = choose(muon, 2)
SSmuon = ak.any((dimuon['0'].charge * dimuon['1'].charge) > 0, axis=1)
OSmuon = ak.any((dimuon['0'].charge * dimuon['1'].charge) < 0, axis=1)
leading_muon_idx = ak.singletons(ak.argmax(muon.pt, axis=1))
leading_muon = muon[leading_muon_idx]
## Electrons
electron = Collections(ev, "Electron", "tightTTH").get()
vetoelectron = Collections(ev, "Electron", "vetoTTH").get()
dielectron = choose(electron, 2)
SSelectron = ak.any(
(dielectron['0'].charge * dielectron['1'].charge) > 0, axis=1)
OSelectron = ak.any(
(dielectron['0'].charge * dielectron['1'].charge) < 0, axis=1)
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[leading_electron_idx]
## Merge electrons and muons - this should work better now in ak1
lepton = ak.concatenate([muon, electron], axis=1)
dilepton = cross(muon, electron)
SSlepton = ak.any((dilepton['0'].charge * dilepton['1'].charge) > 0,
axis=1)
OSlepton = ak.any((dilepton['0'].charge * dilepton['1'].charge) < 0,
axis=1)
leading_lepton_idx = ak.singletons(ak.argmax(lepton.pt, axis=1))
leading_lepton = lepton[leading_lepton_idx]
trailing_lepton_idx = ak.singletons(ak.argmin(lepton.pt, axis=1))
trailing_lepton = lepton[trailing_lepton_idx]
## Jets
jet = getJets(ev, minPt=25, maxEta=4.7, pt_var='pt_nom')
jet = jet[ak.argsort(
jet.pt_nom, ascending=False
)] # need to sort wrt smeared and recorrected jet pt
jet = jet[~match(jet, muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(
jet, electron,
deltaRCut=0.4)] # remove jets that overlap with electrons
central = jet[(abs(jet.eta) < 2.4)]
btag = getBTagsDeepFlavB(
jet, year=self.year) # should study working point for DeepJet
light = getBTagsDeepFlavB(jet, year=self.year, invert=True)
fwd = getFwdJet(light)
fwd_noPU = getFwdJet(light, puId=False)
## forward jets
high_p_fwd = fwd[ak.singletons(ak.argmax(
fwd.p, axis=1))] # highest momentum spectator
high_pt_fwd = fwd[ak.singletons(ak.argmax(
fwd.pt_nom, axis=1))] # highest transverse momentum spectator
high_eta_fwd = fwd[ak.singletons(ak.argmax(abs(
fwd.eta), axis=1))] # most forward spectator
## Get the two leading b-jets in terms of btag score
high_score_btag = central[ak.argsort(central.btagDeepFlavB)][:, :2]
jf = cross(high_p_fwd, jet)
mjf = (jf['0'] + jf['1']).mass
deltaEta = abs(high_p_fwd.eta -
jf[ak.singletons(ak.argmax(mjf, axis=1))]['1'].eta)
deltaEtaMax = ak.max(deltaEta, axis=1)
mjf_max = ak.max(mjf, axis=1)
jj = choose(jet, 2)
mjj_max = ak.max((jj['0'] + jj['1']).mass, axis=1)
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
## other variables
ht = ak.sum(jet.pt, axis=1)
st = met_pt + ht + ak.sum(muon.pt, axis=1) + ak.sum(electron.pt,
axis=1)
ht_central = ak.sum(central.pt, axis=1)
## event selectors
filters = getFilters(ev, year=self.year, dataset=dataset)
triggers = getTriggers(ev, year=self.year, dataset=dataset)
dilep = ((ak.num(electron) == 1) & (ak.num(muon) == 1))
lep0pt = ((ak.num(electron[(electron.pt > 25)]) +
ak.num(muon[(muon.pt > 25)])) > 0)
lep1pt = ((ak.num(electron[(electron.pt > 20)]) +
ak.num(muon[(muon.pt > 20)])) > 1)
lepveto = ((ak.num(vetoelectron) + ak.num(vetomuon)) == 2)
# define the weight
weight = Weights(len(ev))
if not dataset == 'MuonEG':
# lumi weight
weight.add("weight", ev.weight * cfg['lumi'][self.year])
# PU weight - not in the babies...
weight.add("PU",
ev.puWeight,
weightUp=ev.puWeightUp,
weightDown=ev.puWeightDown,
shift=False)
# b-tag SFs
weight.add("btag", self.btagSF.Method1a(btag, light))
# lepton SFs
weight.add("lepton", self.leptonSF.get(electron, muon))
selection = PackedSelection()
selection.add('lepveto', lepveto)
selection.add('dilep', dilep)
selection.add('trigger', (triggers))
selection.add('filter', (filters))
selection.add('p_T(lep0)>25', lep0pt)
selection.add('p_T(lep1)>20', lep1pt)
selection.add('OS', OSlepton)
selection.add('N_btag=2', (ak.num(btag) == 2))
selection.add('N_jet>2', (ak.num(jet) >= 3))
selection.add('N_central>1', (ak.num(central) >= 2))
selection.add('N_fwd>0', (ak.num(fwd) >= 1))
selection.add('MET>30', (ev.MET.pt > 30))
os_reqs = [
'lepveto', 'dilep', 'trigger', 'filter', 'p_T(lep0)>25',
'p_T(lep1)>20', 'OS'
]
bl_reqs = os_reqs + [
'N_btag=2', 'N_jet>2', 'N_central>1', 'N_fwd>0', 'MET>30'
]
os_reqs_d = {sel: True for sel in os_reqs}
os_selection = selection.require(**os_reqs_d)
bl_reqs_d = {sel: True for sel in bl_reqs}
BL = selection.require(**bl_reqs_d)
cutflow = Cutflow(output, ev, weight=weight)
cutflow_reqs_d = {}
for req in bl_reqs:
cutflow_reqs_d.update({req: True})
cutflow.addRow(req, selection.require(**cutflow_reqs_d))
# first, make a few super inclusive plots
output['PV_npvs'].fill(dataset=dataset,
multiplicity=ev.PV[os_selection].npvs,
weight=weight.weight()[os_selection])
output['PV_npvsGood'].fill(dataset=dataset,
multiplicity=ev.PV[os_selection].npvsGood,
weight=weight.weight()[os_selection])
output['N_jet'].fill(dataset=dataset,
multiplicity=ak.num(jet)[os_selection],
weight=weight.weight()[os_selection])
output['N_b'].fill(dataset=dataset,
multiplicity=ak.num(btag)[os_selection],
weight=weight.weight()[os_selection])
output['N_central'].fill(dataset=dataset,
multiplicity=ak.num(central)[os_selection],
weight=weight.weight()[os_selection])
output['N_ele'].fill(dataset=dataset,
multiplicity=ak.num(electron)[os_selection],
weight=weight.weight()[os_selection])
output['N_mu'].fill(dataset=dataset,
multiplicity=ak.num(electron)[os_selection],
weight=weight.weight()[os_selection])
output['N_fwd'].fill(dataset=dataset,
multiplicity=ak.num(fwd)[os_selection],
weight=weight.weight()[os_selection])
output['MET'].fill(dataset=dataset,
pt=ev.MET[os_selection].pt,
phi=ev.MET[os_selection].phi,
weight=weight.weight()[os_selection])
output['electron'].fill(dataset=dataset,
pt=ak.to_numpy(ak.flatten(electron[BL].pt)),
eta=ak.to_numpy(ak.flatten(electron[BL].eta)),
phi=ak.to_numpy(ak.flatten(electron[BL].phi)),
weight=weight.weight()[BL])
output['muon'].fill(dataset=dataset,
pt=ak.to_numpy(ak.flatten(muon[BL].pt)),
eta=ak.to_numpy(ak.flatten(muon[BL].eta)),
phi=ak.to_numpy(ak.flatten(muon[BL].phi)),
weight=weight.weight()[BL])
output['lead_lep'].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_lepton[BL].pt)),
eta=ak.to_numpy(ak.flatten(leading_lepton[BL].eta)),
phi=ak.to_numpy(ak.flatten(leading_lepton[BL].phi)),
weight=weight.weight()[BL])
output['trail_lep'].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(trailing_lepton[BL].pt)),
eta=ak.to_numpy(ak.flatten(trailing_lepton[BL].eta)),
phi=ak.to_numpy(ak.flatten(trailing_lepton[BL].phi)),
weight=weight.weight()[BL])
output['fwd_jet'].fill(dataset=dataset,
pt=ak.flatten(high_p_fwd[BL].pt_nom),
eta=ak.flatten(high_p_fwd[BL].eta),
phi=ak.flatten(high_p_fwd[BL].phi),
weight=weight.weight()[BL])
output['b1'].fill(dataset=dataset,
pt=ak.flatten(high_score_btag[:, 0:1][BL].pt_nom),
eta=ak.flatten(high_score_btag[:, 0:1][BL].eta),
phi=ak.flatten(high_score_btag[:, 0:1][BL].phi),
weight=weight.weight()[BL])
output['b2'].fill(dataset=dataset,
pt=ak.flatten(high_score_btag[:, 1:2][BL].pt_nom),
eta=ak.flatten(high_score_btag[:, 1:2][BL].eta),
phi=ak.flatten(high_score_btag[:, 1:2][BL].phi),
weight=weight.weight()[BL])
output['j1'].fill(dataset=dataset,
pt=ak.flatten(jet.pt_nom[:, 0:1][BL]),
eta=ak.flatten(jet.eta[:, 0:1][BL]),
phi=ak.flatten(jet.phi[:, 0:1][BL]),
weight=weight.weight()[BL])
output['j2'].fill(dataset=dataset,
pt=ak.flatten(jet[:, 1:2][BL].pt_nom),
eta=ak.flatten(jet[:, 1:2][BL].eta),
phi=ak.flatten(jet[:, 1:2][BL].phi),
weight=weight.weight()[BL])
output['j3'].fill(dataset=dataset,
pt=ak.flatten(jet[:, 2:3][BL].pt_nom),
eta=ak.flatten(jet[:, 2:3][BL].eta),
phi=ak.flatten(jet[:, 2:3][BL].phi),
weight=weight.weight()[BL])
# Now, take care of systematic unceratinties
if not dataset == 'MuonEG':
alljets = getJets(ev, minPt=0, maxEta=4.7)
alljets = alljets[(alljets.jetId > 1)]
for var in self.variations:
# get the collections that change with the variations
jet = getPtEtaPhi(alljets, pt_var=var)
jet = jet[(jet.pt > 25)]
jet = jet[~match(
jet, muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(
jet, electron,
deltaRCut=0.4)] # remove jets that overlap with electrons
central = jet[(abs(jet.eta) < 2.4)]
btag = getBTagsDeepFlavB(
jet,
year=self.year) # should study working point for DeepJet
light = getBTagsDeepFlavB(jet, year=self.year, invert=True)
fwd = getFwdJet(light)
fwd_noPU = getFwdJet(light, puId=False)
## forward jets
high_p_fwd = fwd[ak.singletons(ak.argmax(
fwd.p, axis=1))] # highest momentum spectator
high_pt_fwd = fwd[ak.singletons(ak.argmax(
fwd.pt, axis=1))] # highest transverse momentum spectator
high_eta_fwd = fwd[ak.singletons(
ak.argmax(abs(fwd.eta), axis=1))] # most forward spectator
## Get the two leading b-jets in terms of btag score
high_score_btag = central[ak.argsort(
central.btagDeepFlavB)][:, :2]
# get the modified selection -> more difficult
selection.add('N_jet>2_' + var,
(ak.num(jet.pt) >= 3)) # stupid bug here...
selection.add('N_btag=2_' + var, (ak.num(btag) == 2))
selection.add('N_central>1_' + var, (ak.num(central) >= 2))
selection.add('N_fwd>0_' + var, (ak.num(fwd) >= 1))
selection.add('MET>30_' + var, (getattr(ev.MET, var) > 30))
## Don't change the selection for now...
bl_reqs = os_reqs + [
'N_jet>2_' + var, 'MET>30_' + var, 'N_btag=2_' + var,
'N_central>1_' + var, 'N_fwd>0_' + var
]
bl_reqs_d = {sel: True for sel in bl_reqs}
BL = selection.require(**bl_reqs_d)
# the OS selection remains unchanged
output['N_jet_' + var].fill(
dataset=dataset,
multiplicity=ak.num(jet)[os_selection],
weight=weight.weight()[os_selection])
output['N_fwd_' + var].fill(
dataset=dataset,
multiplicity=ak.num(fwd)[os_selection],
weight=weight.weight()[os_selection])
output['N_b_' + var].fill(
dataset=dataset,
multiplicity=ak.num(btag)[os_selection],
weight=weight.weight()[os_selection])
output['N_central_' + var].fill(
dataset=dataset,
multiplicity=ak.num(central)[os_selection],
weight=weight.weight()[os_selection])
# We don't need to redo all plots with variations. E.g., just add uncertainties to the jet plots.
output['j1_' + var].fill(dataset=dataset,
pt=ak.flatten(jet.pt[:, 0:1][BL]),
eta=ak.flatten(jet.eta[:, 0:1][BL]),
phi=ak.flatten(jet.phi[:, 0:1][BL]),
weight=weight.weight()[BL])
output['b1_' + var].fill(
dataset=dataset,
pt=ak.flatten(high_score_btag[:, 0:1].pt[:, 0:1][BL]),
eta=ak.flatten(high_score_btag[:, 0:1].eta[:, 0:1][BL]),
phi=ak.flatten(high_score_btag[:, 0:1].phi[:, 0:1][BL]),
weight=weight.weight()[BL])
output['fwd_jet_' + var].fill(
dataset=dataset,
pt=ak.flatten(high_p_fwd[BL].pt),
eta=ak.flatten(high_p_fwd[BL].eta),
phi=ak.flatten(high_p_fwd[BL].phi),
weight=weight.weight()[BL])
output['MET_' + var].fill(dataset=dataset,
pt=getattr(ev.MET,
var)[os_selection],
phi=ev.MET[os_selection].phi,
weight=weight.weight()[os_selection])
return output
def process(self, events):
output = self._accumulator.identity()
dataset_name = events.metadata['dataset']
output["total_events"][dataset_name] += events.__len__()
# HLT selection
HLT_mask = []
if year == "2016":
if "SingleMuon" in dataset_name:
if "2016B2" in dataset_name:
HLT_mask = events.HLT.IsoMu24 | events.HLT.IsoTkMu24 | events.HLT.Mu50
else:
HLT_mask = events.HLT.IsoMu24 | events.HLT.IsoTkMu24 | events.HLT.Mu50 | events.HLT.TkMu50
else: #https://twiki.cern.ch/twiki/bin/view/CMS/HLTPathsRunIIList
if "2016B2" in dataset_name:
HLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
elif "2016H" in dataset_name:
HLT_mask = events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
else:
HLT_mask = events.HLT.PFHT800 | events.HLT.PFHT900 | events.HLT.AK8PFJet450 | events.HLT.AK8PFJet500 | events.HLT.PFJet500 | events.HLT.CaloJet500_NoJetID
if year == "2017":
if "SingleMuon" in dataset_name:
if "2017B" in dataset_name:
HLT_mask = events.HLT.IsoMu27 | events.HLT.Mu50
else:
HLT_mask = events.HLT.IsoMu27 | events.HLT.Mu50 | events.HLT.OldMu100 | events.HLT.TkMu100
else:
HLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
if year == "2018":
if "SingleMuon" in dataset_name:
HLT_mask = events.HLT.IsoMu24 | events.HLT.Mu50 | events.HLT.OldMu100 | events.HLT.TkMu100
else:
HLT_mask = events.HLT.PFHT1050 | events.HLT.AK8PFJet500 | events.HLT.AK8PFJet550 | events.HLT.CaloJet500_NoJetID | events.HLT.CaloJet550_NoJetID | events.HLT.PFJet500
# Require 3 jets
jet_mask = (events.Jet.pt > 30.) & (abs(events.Jet.eta) <
2.5) & (events.Jet.isTight)
event_mask = (awk.sum(jet_mask, axis=1) >= 3)
event_mask = event_mask & HLT_mask
events_3j = events[event_mask]
# Reduce jet mask to only events with 3 good jets
jet_mask = jet_mask[event_mask]
# Array of the jets to consider for trijet resonance
selected_jets = events_3j.Jet[jet_mask][:, :3]
# Pairs of jets
pairs = [(0, 1), (1, 2), (2, 0)]
jet_i, jet_j = zip(*pairs) # Returns [0, 1, 2] , [1, 2, 0]
m_ij = (selected_jets[:, jet_i] + selected_jets[:, jet_j]).mass
dR_ij = selected_jets[:, jet_i].delta_r(selected_jets[:, jet_j])
dEta_ij = abs(selected_jets[:, jet_i].eta -
selected_jets[:, jet_j].eta)
max_dR = awk.max(dR_ij, axis=1)
max_dEta = awk.max(dEta_ij, axis=1)
min_dR = awk.min(dR_ij, axis=1)
min_dEta = awk.min(dEta_ij, axis=1)
min_pT = awk.min(selected_jets.pt, axis=1)
max_eta = abs(awk.max(selected_jets.eta, axis=1))
jet_k = [2, 0, 1]
dR_i_jk = selected_jets[:, jet_i].delta_r(selected_jets[:, jet_j] +
selected_jets[:, jet_k])
dEta_i_jk = abs(selected_jets[:, jet_i].eta -
(selected_jets[:, jet_j] +
selected_jets[:, jet_k]).eta)
dPhi_i_jk = abs(selected_jets[:, jet_i].phi -
(selected_jets[:, jet_j] +
selected_jets[:, jet_k]).phi)
dPt_i_jk = abs(selected_jets[:, jet_i].pt -
(selected_jets[:, jet_j] + selected_jets[:, jet_k]).pt)
max_dPhi_jjj = awk.max(dPhi_i_jk, axis=1)
m3j = selected_jets.sum().mass
pt_i_overM = selected_jets.pt / m3j
max_pt_overM = awk.max(pt_i_overM, axis=1)
min_pt_overM = awk.min(pt_i_overM, axis=1)
m_01_overM = m_ij[:, 0] / m3j
m_12_overM = m_ij[:, 1] / m3j
m_20_overM = m_ij[:, 2] / m3j
for pt_cut in range(30, 1150, 5):
cut_name = f"min_pT_cut{pt_cut}".format(pt_cut)
selection = PackedSelection()
selection.add("MinJetPt_cut", min_pT > pt_cut)
sel_mask = selection.require(
**{name: True
for name in selection.names})
output[f"N_min_pT_cut{pt_cut}".format(
pt_cut)][dataset_name] += events_3j[sel_mask].__len__()
for eta_cut in np.arange(0, 2.5, 0.05):
cut_name = f"max_eta_cut{eta_cut}".format(eta_cut)
selection = PackedSelection()
selection.add("MaxJetEta_cut", max_eta < eta_cut)
sel_mask = selection.require(
**{name: True
for name in selection.names})
output[f"N_max_eta_cut{eta_cut}".format(
eta_cut)][dataset_name] += events_3j[sel_mask].__len__()
for dEta_max_cut in np.arange(0, 5, 0.1):
cut_name = f"dEta_max_cut{dEta_max_cut}".format(dEta_max_cut)
selection = PackedSelection()
selection.add("MaxJJdEta_cut", max_dEta < dEta_max_cut)
sel_mask = selection.require(
**{name: True
for name in selection.names})
output[f"N_dEta_jj_max_cut{dEta_max_cut}".format(
dEta_max_cut)][dataset_name] += events_3j[sel_mask].__len__()
for dR_min_cut in np.arange(0, 5, 0.1):
cut_name = f"dR_min_cut{dR_min_cut}".format(dR_min_cut)
selection = PackedSelection()
selection.add("MinJJdR_cut", min_dR > dR_min_cut)
sel_mask = selection.require(
**{name: True
for name in selection.names})
output[f"N_dR_jj_min_cut{dR_min_cut}".format(
dR_min_cut)][dataset_name] += events_3j[sel_mask].__len__()
#min cut for the variable dPhi_jjj_max
# for dPhi_jjj_max_min_cut in range(0,6,0.1):
# cut_name = f"dPhi_jjj_max_min_cut{dPhi_jjj_max_min_cut}".format(dPhi_jjj_max_min_cut)
# selections[cut_name] = PackedSelection()
# selections[cut_name].add("j_jj_dPhi_max_cut", min_dR > dPhi_jjj_max_min_cut)
# selection_items[cut_name] = []
# selection_items[cut_name].append("j_jj_dPhi_max_cut")
# sel_mask = HLT_mask & selections[cut_name].require(**{name: True for name in selection_items[cut_name]})
# output[f"N_dPhi_jjj_max_min_cut{dPhi_jjj_max_min_cut}".format(dPhi_jjj_max_min_cut)][dataset_name] += events_3j[sel_mask].__len__()
# for dPhi_jjj_min_max_cut in range(0,6,0.1):
# cut_name = f"dPhi_jjj_max_min_cut{dPhi_jjj_max_min_cut}".format(dPhi_jjj_max_min_cut)
# selections[cut_name] = PackedSelection()
# selections[cut_name].add("j_jj_dPhi_max_cut", min_dR > dPhi_jjj_max_min_cut)
# selection_items[cut_name] = []
# selection_items[cut_name].append("j_jj_dPhi_max_cut")
# sel_mask = HLT_mask & selections[cut_name].require(**{name: True for name in selection_items[cut_name]})
# output[f"N_dPhi_jjj_max_min_cut{dPhi_jjj_max_min_cut}".format(dPhi_jjj_max_min_cut)][dataset_name] += events_3j[sel_mask].__len__()
return output
def process(self, events):
output = self.accumulator.identity()
output['total']['all'] += len(events)
# use a very loose preselection to filter the events
presel = ak.num(events.Jet) > 2
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
## Muons
muon = Collections(ev, "Muon", "vetoTTH").get()
tightmuon = Collections(ev, "Muon", "tightTTH").get()
dimuon = choose(muon, 2)
SSmuon = ak.any((dimuon['0'].charge * dimuon['1'].charge) > 0, axis=1)
leading_muon_idx = ak.singletons(ak.argmax(muon.pt, axis=1))
leading_muon = muon[leading_muon_idx]
## Electrons
electron = Collections(ev, "Electron", "vetoTTH").get()
tightelectron = Collections(ev, "Electron", "tightTTH").get()
dielectron = choose(electron, 2)
SSelectron = ak.any(
(dielectron['0'].charge * dielectron['1'].charge) > 0, axis=1)
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[leading_electron_idx]
## Merge electrons and muons - this should work better now in ak1
dilepton = cross(muon, electron)
SSlepton = ak.any((dilepton['0'].charge * dilepton['1'].charge) > 0,
axis=1)
lepton = ak.concatenate([muon, electron], axis=1)
leading_lepton_idx = ak.singletons(ak.argmax(lepton.pt, axis=1))
leading_lepton = lepton[leading_lepton_idx]
trailing_lepton_idx = ak.singletons(ak.argmin(lepton.pt, axis=1))
trailing_lepton = lepton[trailing_lepton_idx]
dilepton_mass = (leading_lepton + trailing_lepton).mass
dilepton_pt = (leading_lepton + trailing_lepton).pt
dilepton_dR = delta_r(leading_lepton, trailing_lepton)
mt_lep_met = mt(lepton.pt, lepton.phi, ev.MET.pt, ev.MET.phi)
min_mt_lep_met = ak.min(mt_lep_met, axis=1)
## Jets
jet = getJets(ev, minPt=25, maxEta=4.7, pt_var='pt_nom')
jet = jet[ak.argsort(
jet.pt_nom, ascending=False
)] # need to sort wrt smeared and recorrected jet pt
jet = jet[~match(jet, muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(
jet, electron,
deltaRCut=0.4)] # remove jets that overlap with electrons
central = jet[(abs(jet.eta) < 2.4)]
btag = getBTagsDeepFlavB(
jet, year=self.year) # should study working point for DeepJet
light = getBTagsDeepFlavB(jet, year=self.year, invert=True)
fwd = getFwdJet(light)
fwd_noPU = getFwdJet(light, puId=False)
tau = getTaus(ev)
track = getIsoTracks(ev)
## forward jets
j_fwd = fwd[ak.singletons(ak.argmax(
fwd.p, axis=1))] # highest momentum spectator
high_score_btag = central[ak.argsort(central.btagDeepFlavB)][:, :2]
bl = cross(lepton, high_score_btag)
bl_dR = delta_r(bl['0'], bl['1'])
min_bl_dR = ak.min(bl_dR, axis=1)
jf = cross(j_fwd, jet)
mjf = (jf['0'] + jf['1']).mass
j_fwd2 = jf[ak.singletons(
ak.argmax(mjf, axis=1)
)]['1'] # this is the jet that forms the largest invariant mass with j_fwd
delta_eta = ak.fill_none(
ak.pad_none(abs(j_fwd2.eta - j_fwd.eta), 1, clip=True), 0)
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
## other variables
ht = ak.sum(jet.pt, axis=1)
st = met_pt + ht + ak.sum(muon.pt, axis=1) + ak.sum(electron.pt,
axis=1)
## event selectors
filters = getFilters(ev, year=self.year, dataset=dataset)
dilep = ((ak.num(tightelectron) + ak.num(tightmuon)) == 2)
lep0pt = ((ak.num(electron[(electron.pt > 25)]) +
ak.num(muon[(muon.pt > 25)])) > 0)
lep1pt = ((ak.num(electron[(electron.pt > 20)]) +
ak.num(muon[(muon.pt > 20)])) > 1)
lepveto = ((ak.num(electron) + ak.num(muon)) == 2)
selection = PackedSelection()
selection.add('lepveto', lepveto)
selection.add('dilep', dilep)
selection.add('filter', (filters))
selection.add('p_T(lep0)>25', lep0pt)
selection.add('p_T(lep1)>20', lep1pt)
selection.add('SS', (SSlepton | SSelectron | SSmuon))
selection.add('N_jet>3', (ak.num(jet) >= 4))
selection.add('N_central>2', (ak.num(central) >= 3))
selection.add('N_btag>0', (ak.num(btag) >= 1))
selection.add('N_fwd>0', (ak.num(fwd) >= 1))
#ss_reqs = ['lepveto', 'dilep', 'filter', 'p_T(lep0)>25', 'p_T(lep1)>20', 'SS']
ss_reqs = [
'lepveto', 'dilep', 'filter', 'p_T(lep0)>25', 'p_T(lep1)>20', 'SS'
]
#bl_reqs = ss_reqs + ['N_jet>3', 'N_central>2', 'N_btag>0', 'N_fwd>0']
bl_reqs = ss_reqs + ['N_jet>3', 'N_central>2', 'N_btag>0']
ss_reqs_d = {sel: True for sel in ss_reqs}
ss_selection = selection.require(**ss_reqs_d)
bl_reqs_d = {sel: True for sel in bl_reqs}
BL = selection.require(**bl_reqs_d)
weight = Weights(len(ev))
if not dataset == 'MuonEG':
# lumi weight
weight.add("weight", ev.weight)
# PU weight - not in the babies...
weight.add("PU",
ev.puWeight,
weightUp=ev.puWeightUp,
weightDown=ev.puWeightDown,
shift=False)
# b-tag SFs
weight.add("btag", self.btagSF.Method1a(btag, light))
# lepton SFs
weight.add("lepton", self.leptonSF.get(electron, muon))
#cutflow = Cutflow(output, ev, weight=weight)
#cutflow_reqs_d = {}
#for req in bl_reqs:
# cutflow_reqs_d.update({req: True})
# cutflow.addRow( req, selection.require(**cutflow_reqs_d) )
labels = {
'topW_v3': 0,
'TTW': 1,
'TTZ': 2,
'TTH': 3,
'ttbar': 4,
'ttbar1l_MG': 4
}
if dataset in labels:
label_mult = labels[dataset]
else:
label_mult = 5
label = np.ones(len(ev[BL])) * label_mult
output["n_lep"] += processor.column_accumulator(
ak.to_numpy((ak.num(electron) + ak.num(muon))[BL]))
output["n_lep_tight"] += processor.column_accumulator(
ak.to_numpy((ak.num(tightelectron) + ak.num(tightmuon))[BL]))
output["lead_lep_pt"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(leading_lepton[BL].pt, axis=1)))
output["lead_lep_eta"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(leading_lepton[BL].eta, axis=1)))
output["lead_lep_phi"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(leading_lepton[BL].phi, axis=1)))
output["lead_lep_charge"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(leading_lepton[BL].charge, axis=1)))
output["sublead_lep_pt"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(trailing_lepton[BL].pt, axis=1)))
output["sublead_lep_eta"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(trailing_lepton[BL].eta, axis=1)))
output["sublead_lep_phi"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(trailing_lepton[BL].phi, axis=1)))
output["sublead_lep_charge"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(trailing_lepton[BL].charge, axis=1)))
output["lead_jet_pt"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(jet[:, 0:1][BL].pt, axis=1)))
output["lead_jet_eta"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(jet[:, 0:1][BL].eta, axis=1)))
output["lead_jet_phi"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(jet[:, 0:1][BL].phi, axis=1)))
output["sublead_jet_pt"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(jet[:, 1:2][BL].pt, axis=1)))
output["sublead_jet_eta"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(jet[:, 1:2][BL].eta, axis=1)))
output["sublead_jet_phi"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(jet[:, 1:2][BL].phi, axis=1)))
output["lead_btag_pt"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(high_score_btag[:, 0:1][BL].pt, axis=1)))
output["lead_btag_eta"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(high_score_btag[:, 0:1][BL].eta, axis=1)))
output["lead_btag_phi"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(high_score_btag[:, 0:1][BL].phi, axis=1)))
output["sublead_btag_pt"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(high_score_btag[:, 1:2][BL].pt, axis=1)))
output["sublead_btag_eta"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(high_score_btag[:, 1:2][BL].eta, axis=1)))
output["sublead_btag_phi"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(high_score_btag[:, 1:2][BL].phi, axis=1)))
output["fwd_jet_p"] += processor.column_accumulator(
ak.to_numpy(
ak.flatten(ak.fill_none(ak.pad_none(j_fwd[BL].p, 1, clip=True),
0),
axis=1)))
output["fwd_jet_pt"] += processor.column_accumulator(
ak.to_numpy(
ak.flatten(ak.fill_none(
ak.pad_none(j_fwd[BL].pt, 1, clip=True), 0),
axis=1)))
output["fwd_jet_eta"] += processor.column_accumulator(
ak.to_numpy(
ak.flatten(ak.fill_none(
ak.pad_none(j_fwd[BL].eta, 1, clip=True), 0),
axis=1)))
output["fwd_jet_phi"] += processor.column_accumulator(
ak.to_numpy(
ak.flatten(ak.fill_none(
ak.pad_none(j_fwd[BL].phi, 1, clip=True), 0),
axis=1)))
output["mjj_max"] += processor.column_accumulator(
ak.to_numpy(ak.fill_none(ak.max(mjf[BL], axis=1), 0)))
output["delta_eta_jj"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(delta_eta[BL], axis=1)))
output["met"] += processor.column_accumulator(ak.to_numpy(met_pt[BL]))
output["ht"] += processor.column_accumulator(ak.to_numpy(ht[BL]))
output["st"] += processor.column_accumulator(ak.to_numpy(st[BL]))
output["n_jet"] += processor.column_accumulator(
ak.to_numpy(ak.num(jet[BL])))
output["n_btag"] += processor.column_accumulator(
ak.to_numpy(ak.num(btag[BL])))
output["n_fwd"] += processor.column_accumulator(
ak.to_numpy(ak.num(fwd[BL])))
output["n_central"] += processor.column_accumulator(
ak.to_numpy(ak.num(central[BL])))
output["n_tau"] += processor.column_accumulator(
ak.to_numpy(ak.num(tau[BL])))
output["n_track"] += processor.column_accumulator(
ak.to_numpy(ak.num(track[BL])))
output["dilepton_pt"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(dilepton_pt[BL], axis=1)))
output["dilepton_mass"] += processor.column_accumulator(
ak.to_numpy(ak.flatten(dilepton_mass[BL], axis=1)))
output["min_bl_dR"] += processor.column_accumulator(
ak.to_numpy(min_bl_dR[BL]))
output["min_mt_lep_met"] += processor.column_accumulator(
ak.to_numpy(min_mt_lep_met[BL]))
output["label"] += processor.column_accumulator(label)
output["weight"] += processor.column_accumulator(weight.weight()[BL])
output["presel"]["all"] += len(ev[ss_selection])
output["sel"]["all"] += len(ev[BL])
return output
def process(self, events):
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet) >= 2
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
## Electrons
electron = Collections(ev, "Electron", "tight").get()
electron = electron[(electron.pt > 20) & (abs(electron.eta) < 2.4)]
electron = electron[((electron.genPartIdx >= 0) &
(np.abs(electron.matched_gen.pdgId) == 11)
)] #from here on all leptons are gen-matched
##Muons
muon = Collections(ev, "Muon", "tight").get()
muon = muon[(muon.pt > 20) & (abs(muon.eta) < 2.4)]
muon = muon[((muon.genPartIdx >= 0) &
(np.abs(muon.matched_gen.pdgId) == 13))]
##Leptons
lepton = ak.concatenate([muon, electron], axis=1)
SSlepton = (ak.sum(lepton.charge, axis=1) != 0) & (ak.num(lepton) == 2)
OSlepton = (ak.sum(lepton.charge, axis=1) == 0) & (ak.num(lepton) == 2)
leading_lepton_idx = ak.singletons(ak.argmax(lepton.pt, axis=1))
leading_lepton = lepton[leading_lepton_idx]
#jets
jet = getJets(ev, minPt=40, maxEta=2.4, pt_var='pt')
jet = jet[ak.argsort(
jet.pt, ascending=False
)] # need to sort wrt smeared and recorrected jet pt
jet = jet[~match(jet, muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(jet, electron, deltaRCut=0.4)]
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
# setting up the various weights
weight = Weights(len(ev))
weight2 = Weights(len(ev))
if not dataset == 'MuonEG':
# generator weight
weight.add("weight", ev.genWeight)
weight2.add("weight", ev.genWeight)
weight2.add("charge flip",
self.charge_flip_ratio.flip_weight(electron))
#selections
filters = getFilters(ev, year=self.year, dataset=dataset)
ss = (SSlepton)
os = (OSlepton)
jet_all = (ak.num(jet) >= 2)
selection = PackedSelection()
selection.add('filter', (filters))
selection.add('ss', ss)
selection.add('os', os)
selection.add('jet', jet_all)
bl_reqs = ['filter', 'jet']
bl_reqs_d = {sel: True for sel in bl_reqs}
baseline = selection.require(**bl_reqs_d)
s_reqs = bl_reqs + ['ss']
s_reqs_d = {sel: True for sel in s_reqs}
ss_sel = selection.require(**s_reqs_d)
o_reqs = bl_reqs + ['os']
o_reqs_d = {sel: True for sel in o_reqs}
os_sel = selection.require(**o_reqs_d)
#outputs
output['N_jet'].fill(dataset=dataset,
multiplicity=ak.num(jet)[baseline],
weight=weight.weight()[baseline])
output['N_ele'].fill(dataset=dataset,
multiplicity=ak.num(lepton)[ss_sel],
weight=weight.weight()[ss_sel])
output['N_ele2'].fill(dataset=dataset,
multiplicity=ak.num(lepton)[os_sel],
weight=weight2.weight()[os_sel])
output["electron"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_lepton[ss_sel].pt)),
eta=abs(ak.to_numpy(ak.flatten(leading_lepton[ss_sel].eta))),
phi=ak.to_numpy(ak.flatten(leading_lepton[ss_sel].phi)),
weight=weight.weight()[ss_sel])
output["electron2"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_lepton[os_sel].pt)),
eta=abs(ak.to_numpy(ak.flatten(leading_lepton[os_sel].eta))),
phi=ak.to_numpy(ak.flatten(leading_lepton[os_sel].phi)),
weight=weight2.weight()[os_sel])
return output
def process(self, events):
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet) > 0
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
#output['totalEvents']['all'] += len(events)
#output['skimmedEvents']['all'] += len(ev)
if self.year == 2018:
triggers = ev.HLT.Ele23_Ele12_CaloIdL_TrackIdL_IsoVL
elif self.year == 2017:
triggers = ev.HLT.Ele23_Ele12_CaloIdL_TrackIdL_IsoVL
elif self.year == 2016:
triggers = ev.HLT.Ele23_Ele12_CaloIdL_TrackIdL_IsoVL_DZ
if self.year == 2018:
lumimask = LumiMask(
'processors/Cert_314472-325175_13TeV_Legacy2018_Collisions18_JSON.txt'
)
## Electrons
electron = Collections(ev, "Electron", "tight").get()
electron = electron[(electron.pt > 25) & (np.abs(electron.eta) < 2.4)]
loose_electron = Collections(ev, "Electron", "veto").get()
loose_electron = loose_electron[(loose_electron.pt > 25)
& (np.abs(loose_electron.eta) < 2.4)]
SSelectron = (ak.sum(electron.charge, axis=1) != 0) & (ak.num(electron)
== 2)
OSelectron = (ak.sum(electron.charge, axis=1) == 0) & (ak.num(electron)
== 2)
dielectron = choose(electron, 2)
dielectron_mass = (dielectron['0'] + dielectron['1']).mass
dielectron_pt = (dielectron['0'] + dielectron['1']).pt
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[(leading_electron_idx)]
leading_electron = leading_electron[(leading_electron.pt > 30)]
trailing_electron_idx = ak.singletons(ak.argmin(electron.pt, axis=1))
trailing_electron = electron[trailing_electron_idx]
##Muons
loose_muon = Collections(ev, "Muon", "veto").get()
loose_muon = loose_muon[(loose_muon.pt > 20)
& (np.abs(loose_muon.eta) < 2.4)]
#jets
jet = getJets(ev, minPt=40, maxEta=2.4, pt_var='pt', UL=False)
jet = jet[ak.argsort(
jet.pt, ascending=False
)] # need to sort wrt smeared and recorrected jet pt
jet = jet[~match(jet, loose_muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(
jet, electron,
deltaRCut=0.4)] # remove jets that overlap with electrons
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
#selections
filters = getFilters(ev, year=self.year, dataset=dataset)
mask = lumimask(ev.run, ev.luminosityBlock)
ss = (SSelectron)
os = (OSelectron)
mass = (ak.min(np.abs(dielectron_mass - 91.2), axis=1) < 15)
lead_electron = (ak.min(leading_electron.pt, axis=1) > 30)
jet1 = (ak.num(jet) >= 1)
jet2 = (ak.num(jet) >= 2)
num_loose = ((ak.num(loose_electron) == 2) & (ak.num(loose_muon) == 0))
selection = PackedSelection()
selection.add('filter', (filters))
selection.add('mask', (mask))
selection.add('ss', ss)
selection.add('os', os)
selection.add('mass', mass)
selection.add('leading', lead_electron)
selection.add('triggers', triggers)
selection.add('one jet', jet1)
selection.add('two jets', jet2)
selection.add('num_loose', num_loose)
bl_reqs = ['filter'] + ['mass'] + ['mask'] + ['triggers'] + [
'leading'
] + ['num_loose']
#bl_reqs = ['filter'] + ['mass'] + ['triggers'] + ['leading'] + ['num_loose']
bl_reqs_d = {sel: True for sel in bl_reqs}
baseline = selection.require(**bl_reqs_d)
s_reqs = bl_reqs + ['ss']
s_reqs_d = {sel: True for sel in s_reqs}
ss_sel = selection.require(**s_reqs_d)
o_reqs = bl_reqs + ['os']
o_reqs_d = {sel: True for sel in o_reqs}
os_sel = selection.require(**o_reqs_d)
j1s_reqs = s_reqs + ['one jet']
j1s_reqs_d = {sel: True for sel in j1s_reqs}
j1ss_sel = selection.require(**j1s_reqs_d)
j1o_reqs = o_reqs + ['one jet']
j1o_reqs_d = {sel: True for sel in j1o_reqs}
j1os_sel = selection.require(**j1o_reqs_d)
j2s_reqs = s_reqs + ['two jets']
j2s_reqs_d = {sel: True for sel in j2s_reqs}
j2ss_sel = selection.require(**j2s_reqs_d)
j2o_reqs = o_reqs + ['two jets']
j2o_reqs_d = {sel: True for sel in j2o_reqs}
j2os_sel = selection.require(**j2o_reqs_d)
output["N_jet"].fill(
dataset=dataset,
multiplicity=ak.num(jet)[os_sel],
)
return output
def process(self, events):
output = self.accumulator.identity()
# we can use a very loose preselection to filter the events. nothing is done with this presel, though
presel = ak.num(events.Jet) > 0
if self.year == 2016:
lumimask = LumiMask(
'../data/lumi/Cert_271036-284044_13TeV_Legacy2016_Collisions16_JSON.txt'
)
if self.year == 2017:
lumimask = LumiMask(
'../data/lumi/Cert_294927-306462_13TeV_UL2017_Collisions17_GoldenJSON.txt'
)
if self.year == 2018:
lumimask = LumiMask(
'../data/lumi/Cert_314472-325175_13TeV_Legacy2018_Collisions18_JSON.txt'
)
ev = events[presel]
dataset = ev.metadata['dataset']
# load the config - probably not needed anymore
cfg = loadConfig()
output['totalEvents']['all'] += len(events)
output['skimmedEvents']['all'] += len(ev)
if self.year == 2018:
triggers = ev.HLT.Ele23_Ele12_CaloIdL_TrackIdL_IsoVL
elif self.year == 2017:
triggers = ev.HLT.Ele23_Ele12_CaloIdL_TrackIdL_IsoVL
elif self.year == 2016:
triggers = ev.HLT.Ele23_Ele12_CaloIdL_TrackIdL_IsoVL_DZ
## Electrons
electron = Collections(ev, "Electron", "tightFCNC", 0, self.year).get()
electron = electron[(electron.pt > 25) & (np.abs(electron.eta) < 2.4)]
loose_electron = Collections(ev, "Electron", "looseFCNC", 0,
self.year).get()
loose_electron = loose_electron[(loose_electron.pt > 25)
& (np.abs(loose_electron.eta) < 2.4)]
SSelectron = (ak.sum(electron.charge, axis=1) != 0) & (ak.num(electron)
== 2)
OSelectron = (ak.sum(electron.charge, axis=1) == 0) & (ak.num(electron)
== 2)
dielectron = choose(electron, 2)
dielectron_mass = (dielectron['0'] + dielectron['1']).mass
dielectron_pt = (dielectron['0'] + dielectron['1']).pt
leading_electron_idx = ak.singletons(ak.argmax(electron.pt, axis=1))
leading_electron = electron[(leading_electron_idx)]
leading_electron = leading_electron[(leading_electron.pt > 30)]
trailing_electron_idx = ak.singletons(ak.argmin(electron.pt, axis=1))
trailing_electron = electron[trailing_electron_idx]
##Muons
loose_muon = Collections(ev, "Muon", "looseFCNC", 0, self.year).get()
loose_muon = loose_muon[(loose_muon.pt > 20)
& (np.abs(loose_muon.eta) < 2.4)]
#jets
jet = getJets(ev, minPt=40, maxEta=2.4, pt_var='pt')
jet = jet[~match(jet, loose_muon,
deltaRCut=0.4)] # remove jets that overlap with muons
jet = jet[~match(
jet, electron,
deltaRCut=0.4)] # remove jets that overlap with electrons
## MET -> can switch to puppi MET
met_pt = ev.MET.pt
met_phi = ev.MET.phi
#weights
weight = Weights(len(ev))
weight2 = Weights(len(ev))
weight2.add("charge flip",
self.charge_flip_ratio.flip_weight(electron))
#selections
filters = getFilters(ev, year=self.year, dataset=dataset, UL=False)
mask = lumimask(ev.run, ev.luminosityBlock)
ss = (SSelectron)
os = (OSelectron)
mass = (ak.min(np.abs(dielectron_mass - 91.2), axis=1) < 15)
lead_electron = (ak.min(leading_electron.pt, axis=1) > 30)
jet1 = (ak.num(jet) >= 1)
jet2 = (ak.num(jet) >= 2)
num_loose = ((ak.num(loose_electron) == 2) & (ak.num(loose_muon) == 0))
selection = PackedSelection()
selection.add('filter', (filters))
selection.add('mask', (mask))
selection.add('ss', ss)
selection.add('os', os)
selection.add('mass', mass)
selection.add('leading', lead_electron)
selection.add('triggers', triggers)
selection.add('one jet', jet1)
selection.add('two jets', jet2)
selection.add('num_loose', num_loose)
bl_reqs = ['filter'] + ['triggers'] + ['mask']
bl_reqs_d = {sel: True for sel in bl_reqs}
baseline = selection.require(**bl_reqs_d)
s_reqs = bl_reqs + ['ss'] + ['mass'] + ['num_loose'] + ['leading']
s_reqs_d = {sel: True for sel in s_reqs}
ss_sel = selection.require(**s_reqs_d)
o_reqs = bl_reqs + ['os'] + ['mass'] + ['num_loose'] + ['leading']
o_reqs_d = {sel: True for sel in o_reqs}
os_sel = selection.require(**o_reqs_d)
j1s_reqs = s_reqs + ['one jet']
j1s_reqs_d = {sel: True for sel in j1s_reqs}
j1ss_sel = selection.require(**j1s_reqs_d)
j1o_reqs = o_reqs + ['one jet']
j1o_reqs_d = {sel: True for sel in j1o_reqs}
j1os_sel = selection.require(**j1o_reqs_d)
j2s_reqs = s_reqs + ['two jets']
j2s_reqs_d = {sel: True for sel in j2s_reqs}
j2ss_sel = selection.require(**j2s_reqs_d)
j2o_reqs = o_reqs + ['two jets']
j2o_reqs_d = {sel: True for sel in j2o_reqs}
j2os_sel = selection.require(**j2o_reqs_d)
#outputs
output["electron_data1"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_electron[os_sel].pt)),
eta=ak.to_numpy(ak.flatten(leading_electron[os_sel].eta)),
phi=ak.to_numpy(ak.flatten(leading_electron[os_sel].phi)),
weight=weight2.weight()[os_sel])
output["electron_data2"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(trailing_electron[os_sel].pt)),
eta=ak.to_numpy(ak.flatten(trailing_electron[os_sel].eta)),
phi=ak.to_numpy(ak.flatten(trailing_electron[os_sel].phi)),
weight=weight2.weight()[os_sel])
output["electron_data3"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_electron[j1os_sel].pt)),
eta=ak.to_numpy(ak.flatten(leading_electron[j1os_sel].eta)),
phi=ak.to_numpy(ak.flatten(leading_electron[j1os_sel].phi)),
weight=weight2.weight()[j1os_sel])
output["electron_data4"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(trailing_electron[j1os_sel].pt)),
eta=ak.to_numpy(ak.flatten(trailing_electron[j1os_sel].eta)),
phi=ak.to_numpy(ak.flatten(trailing_electron[j1os_sel].phi)),
weight=weight2.weight()[j1os_sel])
output["electron_data5"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_electron[j2os_sel].pt)),
eta=ak.to_numpy(ak.flatten(leading_electron[j2os_sel].eta)),
phi=ak.to_numpy(ak.flatten(leading_electron[j2os_sel].phi)),
weight=weight2.weight()[j2os_sel])
output["electron_data6"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(trailing_electron[j2os_sel].pt)),
eta=ak.to_numpy(ak.flatten(trailing_electron[j2os_sel].eta)),
phi=ak.to_numpy(ak.flatten(trailing_electron[j2os_sel].phi)),
weight=weight2.weight()[j2os_sel])
output["electron_data7"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_electron[ss_sel].pt)),
eta=ak.to_numpy(ak.flatten(leading_electron[ss_sel].eta)),
phi=ak.to_numpy(ak.flatten(leading_electron[ss_sel].phi)),
weight=weight.weight()[ss_sel])
output["electron_data8"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(trailing_electron[ss_sel].pt)),
eta=ak.to_numpy(ak.flatten(trailing_electron[ss_sel].eta)),
phi=ak.to_numpy(ak.flatten(trailing_electron[ss_sel].phi)),
weight=weight.weight()[ss_sel])
output["electron_data9"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_electron[j1ss_sel].pt)),
eta=ak.to_numpy(ak.flatten(leading_electron[j1ss_sel].eta)),
phi=ak.to_numpy(ak.flatten(leading_electron[j1ss_sel].phi)),
weight=weight.weight()[j1ss_sel])
output["electron_data10"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(trailing_electron[j1ss_sel].pt)),
eta=ak.to_numpy(ak.flatten(trailing_electron[j1ss_sel].eta)),
phi=ak.to_numpy(ak.flatten(trailing_electron[j1ss_sel].phi)),
weight=weight.weight()[j1ss_sel])
output["electron_data11"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(leading_electron[j2ss_sel].pt)),
eta=ak.to_numpy(ak.flatten(leading_electron[j2ss_sel].eta)),
phi=ak.to_numpy(ak.flatten(leading_electron[j2ss_sel].phi)),
weight=weight.weight()[j2ss_sel])
output["electron_data12"].fill(
dataset=dataset,
pt=ak.to_numpy(ak.flatten(trailing_electron[j2ss_sel].pt)),
eta=ak.to_numpy(ak.flatten(trailing_electron[j2ss_sel].eta)),
phi=ak.to_numpy(ak.flatten(trailing_electron[j2ss_sel].phi)),
weight=weight.weight()[j2ss_sel])
output["dilep_mass1"].fill(
dataset=dataset,
mass=ak.to_numpy(ak.flatten(dielectron_mass[os_sel])),
pt=ak.to_numpy(ak.flatten(dielectron_pt[os_sel])),
weight=weight2.weight()[os_sel])
output["dilep_mass2"].fill(
dataset=dataset,
mass=ak.to_numpy(ak.flatten(dielectron_mass[j1os_sel])),
pt=ak.to_numpy(ak.flatten(dielectron_pt[j1os_sel])),
weight=weight2.weight()[j1os_sel])
output["dilep_mass3"].fill(
dataset=dataset,
mass=ak.to_numpy(ak.flatten(dielectron_mass[j2os_sel])),
pt=ak.to_numpy(ak.flatten(dielectron_pt[j2os_sel])),
weight=weight2.weight()[j2os_sel])
output["dilep_mass4"].fill(
dataset=dataset,
mass=ak.to_numpy(ak.flatten(dielectron_mass[ss_sel])),
pt=ak.to_numpy(ak.flatten(dielectron_pt[ss_sel])),
weight=weight.weight()[ss_sel])
output["dilep_mass5"].fill(
dataset=dataset,
mass=ak.to_numpy(ak.flatten(dielectron_mass[j1ss_sel])),
pt=ak.to_numpy(ak.flatten(dielectron_pt[j1ss_sel])),
weight=weight.weight()[j1ss_sel])
output["dilep_mass6"].fill(
dataset=dataset,
mass=ak.to_numpy(ak.flatten(dielectron_mass[j2ss_sel])),
pt=ak.to_numpy(ak.flatten(dielectron_pt[j2ss_sel])),
weight=weight.weight()[j2ss_sel])
output["MET"].fill(dataset=dataset,
pt=met_pt[os_sel],
weight=weight2.weight()[os_sel])
output["MET2"].fill(dataset=dataset,
pt=met_pt[j1os_sel],
weight=weight2.weight()[j1os_sel])
output["MET3"].fill(dataset=dataset,
pt=met_pt[j2os_sel],
weight=weight2.weight()[j2os_sel])
output["MET4"].fill(dataset=dataset,
pt=met_pt[ss_sel],
weight=weight.weight()[ss_sel])
output["MET5"].fill(dataset=dataset,
pt=met_pt[j1ss_sel],
weight=weight.weight()[j1ss_sel])
output["MET6"].fill(dataset=dataset,
pt=met_pt[j2ss_sel],
weight=weight.weight()[j2ss_sel])
output["N_jet"].fill(dataset=dataset,
multiplicity=ak.num(jet)[os_sel],
weight=weight2.weight()[os_sel])
output["N_jet2"].fill(dataset=dataset,
multiplicity=ak.num(jet)[j1os_sel],
weight=weight2.weight()[j1os_sel])
output["N_jet3"].fill(dataset=dataset,
multiplicity=ak.num(jet)[j2os_sel],
weight=weight2.weight()[j2os_sel])
output["N_jet4"].fill(dataset=dataset,
multiplicity=ak.num(jet)[ss_sel],
weight=weight.weight()[ss_sel])
output["N_jet5"].fill(dataset=dataset,
multiplicity=ak.num(jet)[j1ss_sel],
weight=weight.weight()[j1ss_sel])
output["N_jet6"].fill(dataset=dataset,
multiplicity=ak.num(jet)[j2ss_sel],
weight=weight.weight()[j2ss_sel])
output["PV_npvsGood"].fill(dataset=dataset,
multiplicity=ev.PV[os_sel].npvsGood,
weight=weight2.weight()[os_sel])
output["PV_npvsGood2"].fill(dataset=dataset,
multiplicity=ev.PV[j1os_sel].npvsGood,
weight=weight2.weight()[j1os_sel])
output["PV_npvsGood3"].fill(dataset=dataset,
multiplicity=ev.PV[j2os_sel].npvsGood,
weight=weight2.weight()[j2os_sel])
output["PV_npvsGood4"].fill(dataset=dataset,
multiplicity=ev.PV[ss_sel].npvsGood,
weight=weight.weight()[ss_sel])
output["PV_npvsGood5"].fill(dataset=dataset,
multiplicity=ev.PV[j1ss_sel].npvsGood,
weight=weight.weight()[j1ss_sel])
output["PV_npvsGood6"].fill(dataset=dataset,
multiplicity=ev.PV[j2ss_sel].npvsGood,
weight=weight.weight()[j2ss_sel])
return output
