def unpack(self, file_index):
treeevtorig = uproot.open(self.l_root[file_index])[self.n_treeevt]
dfevtorig = treeevtorig.pandas.df(branches=self.v_evt)
dfevtorig = selectdfrunlist(dfevtorig, self.runlist, "run_number")
dfevtorig = selectdfquery(dfevtorig, self.s_cen_unp)
dfevtorig = dfevtorig.reset_index(drop=True)
pickle.dump(dfevtorig, openfile(self.l_evtorig[file_index], "wb"), protocol=4)
dfevt = selectdfquery(dfevtorig, self.s_good_evt_unp)
dfevt = dfevt.reset_index(drop=True)
pickle.dump(dfevt, openfile(self.l_evt[file_index], "wb"), protocol=4)
treereco = uproot.open(self.l_root[file_index])[self.n_treereco]
dfreco = treereco.pandas.df(branches=self.v_all)
dfreco = selectdfrunlist(dfreco, self.runlist, "run_number")
dfreco = selectdfquery(dfreco, self.s_reco_unp)
dfreco = pd.merge(dfreco, dfevt, on=self.v_evtmatch)
isselacc = selectfidacc(dfreco.pt_cand.values, dfreco.y_cand.values)
dfreco = dfreco[np.array(isselacc, dtype=bool)]
if self.b_trackcuts is not None:
dfreco = filter_bit_df(dfreco, self.v_bitvar, self.b_trackcuts)
dfreco[self.v_isstd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_std), dtype=int)
dfreco = dfreco.reset_index(drop=True)
if self.mcordata == "mc":
dfreco[self.v_ismcsignal] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsig), dtype=int)
dfreco[self.v_ismcprompt] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsigprompt), dtype=int)
dfreco[self.v_ismcfd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsigfd), dtype=int)
dfreco[self.v_ismcbkg] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcbkg), dtype=int)
pickle.dump(dfreco, openfile(self.l_reco[file_index], "wb"), protocol=4)
if self.mcordata == "mc":
treegen = uproot.open(self.l_root[file_index])[self.n_treegen]
dfgen = treegen.pandas.df(branches=self.v_gen)
dfgen = selectdfrunlist(dfgen, self.runlist, "run_number")
dfgen = pd.merge(dfgen, dfevtorig, on=self.v_evtmatch)
dfgen = selectdfquery(dfgen, self.s_gen_unp)
dfgen[self.v_isstd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_std), dtype=int)
dfgen[self.v_ismcsignal] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsig), dtype=int)
dfgen[self.v_ismcprompt] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigprompt), dtype=int)
dfgen[self.v_ismcfd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigfd), dtype=int)
dfgen[self.v_ismcbkg] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcbkg), dtype=int)
dfgen = dfgen.reset_index(drop=True)
pickle.dump(dfgen, openfile(self.l_gen[file_index], "wb"), protocol=4)
def skimmer(filein, filevt, fileout, skimming_sel, var_evt_match, param_case,
presel_reco, sel_cent, skimming2_dotrackpid, runlist):
df = pickle.load(open(filein, "rb"))
dfevt = pickle.load(open(filevt, "rb"))
if "Evt" not in filein:
df = pd.merge(df, dfevt, on=var_evt_match)
if skimming_sel is not None:
df = df.query(skimming_sel)
if runlist is not None:
array_run = df.run_number.values
isgoodrun = select_runs(runlist, array_run)
df = df[np.array(isgoodrun, dtype=bool)]
if "Reco" in filein:
if skimming2_dotrackpid is True:
df = filter_df_cand(df, param_case, 'presel_track_pid')
if presel_reco is not None:
df = df.query(presel_reco)
array_pt = df.pt_cand.values
array_y = df.y_cand.values
isselacc = selectfidacc(array_pt, array_y)
df = df[np.array(isselacc, dtype=bool)]
if sel_cent is not None:
df = df.query(sel_cent)
df.to_pickle(fileout)
def unpack(self, file_index):
treeevtorig = uproot.open(self.l_root[file_index])[self.n_treeevt]
dfevtorig = treeevtorig.pandas.df(branches=self.v_evt)
dfevtorig = selectdfrunlist(dfevtorig, self.runlist, "run_number")
dfevtorig = selectdfquery(dfevtorig, self.s_cen_unp)
dfevtorig = dfevtorig.reset_index(drop=True)
pickle.dump(dfevtorig, openfile(self.l_evtorig[file_index], "wb"), protocol=4)
dfevt = selectdfquery(dfevtorig, self.s_good_evt_unp)
dfevt = dfevt.reset_index(drop=True)
pickle.dump(dfevt, openfile(self.l_evt[file_index], "wb"), protocol=4)
treereco = uproot.open(self.l_root[file_index])[self.n_treereco]
dfreco = treereco.pandas.df(branches=self.v_all)
dfreco = selectdfrunlist(dfreco, self.runlist, "run_number")
dfreco = selectdfquery(dfreco, self.s_reco_unp)
dfreco = pd.merge(dfreco, dfevt, on=self.v_evtmatch)
isselacc = selectfidacc(dfreco.pt_cand.values, dfreco.y_cand.values)
dfreco = dfreco[np.array(isselacc, dtype=bool)]
arraysub = [0 for ival in range(len(dfreco))]
n_tracklets_corr = dfreco["n_tracklets_corr"].values
n_tracklets_corr_sub = None
for iprong in range(self.nprongs):
spdhits_thisprong = dfreco["spdhits_prong%s" % iprong].values
ntrackletsthisprong = [1 if spdhits_thisprong[index] == 3 else 0 \
for index in range(len(dfreco))]
arraysub = np.add(ntrackletsthisprong, arraysub)
n_tracklets_corr_sub = np.subtract(n_tracklets_corr, arraysub)
dfreco["n_tracklets_corr_sub"] = n_tracklets_corr_sub
if self.b_trackcuts is not None:
dfreco = filter_bit_df(dfreco, self.v_bitvar, self.b_trackcuts)
dfreco[self.v_isstd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_std), dtype=int)
dfreco = dfreco.reset_index(drop=True)
if self.mcordata == "mc":
dfreco[self.v_ismcsignal] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsig), dtype=int)
dfreco[self.v_ismcprompt] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsigprompt), dtype=int)
dfreco[self.v_ismcfd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsigfd), dtype=int)
dfreco[self.v_ismcbkg] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcbkg), dtype=int)
pickle.dump(dfreco, openfile(self.l_reco[file_index], "wb"), protocol=4)
if self.mcordata == "mc":
treegen = uproot.open(self.l_root[file_index])[self.n_treegen]
dfgen = treegen.pandas.df(branches=self.v_gen)
dfgen = selectdfrunlist(dfgen, self.runlist, "run_number")
dfgen = pd.merge(dfgen, dfevtorig, on=self.v_evtmatch)
dfgen = selectdfquery(dfgen, self.s_gen_unp)
dfgen[self.v_isstd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_std), dtype=int)
dfgen[self.v_ismcsignal] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsig), dtype=int)
dfgen[self.v_ismcprompt] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigprompt), dtype=int)
dfgen[self.v_ismcfd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigfd), dtype=int)
dfgen[self.v_ismcbkg] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcbkg), dtype=int)
dfgen = dfgen.reset_index(drop=True)
pickle.dump(dfgen, openfile(self.l_gen[file_index], "wb"), protocol=4)
def unpack(self, file_index):
treeevtorig = uproot.open(self.l_root[file_index])[self.n_treeevt]
try:
dfevtorig = treeevtorig.pandas.df(branches=self.v_evt)
except Exception as e: # pylint: disable=broad-except
print('Missing variable in the event root tree', str(e))
print('Missing variable in the candidate root tree')
print('I am sorry, I am dying ...\n \n \n')
sys.exit()
dfevtorig = selectdfquery(dfevtorig, self.s_cen_unp)
dfevtorig = dfevtorig.reset_index(drop=True)
pickle.dump(dfevtorig,
openfile(self.l_evtorig[file_index], "wb"),
protocol=4)
dfevt = selectdfquery(dfevtorig, self.s_good_evt_unp)
dfevt = dfevt.reset_index(drop=True)
pickle.dump(dfevt, openfile(self.l_evt[file_index], "wb"), protocol=4)
treereco = uproot.open(self.l_root[file_index])[self.n_treereco]
try:
dfreco = treereco.pandas.df(branches=self.v_all)
except Exception as e: # pylint: disable=broad-except
print('Missing variable in the candidate root tree')
print('I am sorry, I am dying ...\n \n \n')
sys.exit()
dfreco = selectdfquery(dfreco, self.s_reco_unp)
dfreco = pd.merge(dfreco, dfevt, on=self.v_evtmatch)
isselacc = selectfidacc(dfreco.pt_cand.values, dfreco.y_cand.values)
dfreco = dfreco[np.array(isselacc, dtype=bool)]
arraysub = [0 for ival in range(len(dfreco))]
n_tracklets_corr = dfreco["n_tracklets_corr"].values
n_tracklets_corr_shm = dfreco["n_tracklets_corr_shm"].values
n_tracklets_corr_sub = None
n_tracklets_corr_shm_sub = None
for iprong in range(self.nprongs):
if self.prongformultsub[iprong] == 0:
continue
#print("considering prong %d for sub" % iprong)
spdhits_thisprong = dfreco["spdhits_prong%s" % iprong].values
ntrackletsthisprong = [1 if spdhits_thisprong[index] == 3 else 0 \
for index in range(len(dfreco))]
arraysub = np.add(ntrackletsthisprong, arraysub)
n_tracklets_corr_sub = np.subtract(n_tracklets_corr, arraysub)
n_tracklets_corr_shm_sub = np.subtract(n_tracklets_corr_shm, arraysub)
dfreco["n_tracklets_corr_sub"] = n_tracklets_corr_sub
dfreco["n_tracklets_corr_shm_sub"] = n_tracklets_corr_shm_sub
if self.b_trackcuts is not None:
dfreco = filter_bit_df(dfreco, self.v_bitvar, self.b_trackcuts)
dfreco[self.v_isstd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_std),
dtype=int)
dfreco = dfreco.reset_index(drop=True)
if self.mcordata == "mc":
dfreco[self.v_ismcsignal] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcsig),
dtype=int)
dfreco[self.v_ismcprompt] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcsigprompt),
dtype=int)
dfreco[self.v_ismcfd] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcsigfd),
dtype=int)
dfreco[self.v_ismcbkg] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcbkg),
dtype=int)
pickle.dump(dfreco,
openfile(self.l_reco[file_index], "wb"),
protocol=4)
if self.mcordata == "mc":
treegen = uproot.open(self.l_root[file_index])[self.n_treegen]
dfgen = treegen.pandas.df(branches=self.v_gen)
dfgen = pd.merge(dfgen, dfevtorig, on=self.v_evtmatch)
dfgen = selectdfquery(dfgen, self.s_gen_unp)
dfgen[self.v_isstd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_std),
dtype=int)
dfgen[self.v_ismcsignal] = np.array(tag_bit_df(
dfgen, self.v_bitvar, self.b_mcsig),
dtype=int)
dfgen[self.v_ismcprompt] = np.array(tag_bit_df(
dfgen, self.v_bitvar, self.b_mcsigprompt),
dtype=int)
dfgen[self.v_ismcfd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigfd),
dtype=int)
dfgen[self.v_ismcbkg] = np.array(tag_bit_df(
dfgen, self.v_bitvar, self.b_mcbkg),
dtype=int)
dfgen = dfgen.reset_index(drop=True)
pickle.dump(dfgen,
openfile(self.l_gen[file_index], "wb"),
protocol=4)
def unpack(self, file_index):
treeevtorig = uproot.open(self.l_root[file_index])[self.n_treeevt]
dfevtorig = treeevtorig.pandas.df(branches=self.v_evt)
dfevtorig = selectdfrunlist(dfevtorig, self.runlist, "run_number")
dfevtorig = selectdfquery(dfevtorig, self.s_cen_unp)
dfevtorig = dfevtorig.reset_index(drop=True)
pickle.dump(dfevtorig, openfile(self.l_evtorig[file_index], "wb"), protocol=4)
dfevt = selectdfquery(dfevtorig, self.s_good_evt_unp)
dfevt = dfevt.reset_index(drop=True)
pickle.dump(dfevt, openfile(self.l_evt[file_index], "wb"), protocol=4)
fileevtroot = TFile.Open(self.l_evtorigroot[file_index], "recreate")
hNorm = TH1F("hEvForNorm", ";;Normalisation", 2, 0.5, 2.5)
hNorm.GetXaxis().SetBinLabel(1, "normsalisation factor")
hNorm.GetXaxis().SetBinLabel(2, "selected events")
nselevt = 0
norm = 0
# Handle silent weird behaviour of Pandas if dataframe is empty
# Otherwise, if it is empty it might just silently return from this frunction for some
# reason and everything what follows would just be skipped.
if not dfevt.empty:
nselevt = len(dfevt.query("is_ev_rej==0"))
norm = getnormforselevt(dfevt)
hNorm.SetBinContent(1, norm)
hNorm.SetBinContent(2, nselevt)
hNorm.Write()
fileevtroot.Close()
treereco = uproot.open(self.l_root[file_index])[self.n_treereco]
dfreco = treereco.pandas.df(branches=self.v_all)
dfreco = selectdfrunlist(dfreco, self.runlist, "run_number")
dfreco = selectdfquery(dfreco, self.s_reco_unp)
dfreco = pd.merge(dfreco, dfevt, on=self.v_evtmatch)
isselacc = selectfidacc(dfreco.pt_cand.values, dfreco.y_cand.values)
dfreco = dfreco[np.array(isselacc, dtype=bool)]
if self.b_trackcuts is not None:
dfreco = filter_bit_df(dfreco, self.v_bitvar, self.b_trackcuts)
dfreco[self.v_isstd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_std), dtype=int)
dfreco = dfreco.reset_index(drop=True)
if self.mcordata == "mc":
dfreco[self.v_ismcsignal] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsig), dtype=int)
dfreco[self.v_ismcprompt] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsigprompt), dtype=int)
dfreco[self.v_ismcfd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcsigfd), dtype=int)
dfreco[self.v_ismcbkg] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_mcbkg), dtype=int)
pickle.dump(dfreco, openfile(self.l_reco[file_index], "wb"), protocol=4)
if self.mcordata == "mc":
treegen = uproot.open(self.l_root[file_index])[self.n_treegen]
dfgen = treegen.pandas.df(branches=self.v_gen)
dfgen = selectdfrunlist(dfgen, self.runlist, "run_number")
dfgen = pd.merge(dfgen, dfevtorig, on=self.v_evtmatch)
dfgen = selectdfquery(dfgen, self.s_gen_unp)
dfgen[self.v_isstd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_std), dtype=int)
dfgen[self.v_ismcsignal] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsig), dtype=int)
dfgen[self.v_ismcprompt] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigprompt), dtype=int)
dfgen[self.v_ismcfd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigfd), dtype=int)
dfgen[self.v_ismcbkg] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcbkg), dtype=int)
dfgen = dfgen.reset_index(drop=True)
pickle.dump(dfgen, openfile(self.l_gen[file_index], "wb"), protocol=4)
def unpack(self, file_index):
# Open root file and save event tree to dataframe
treeevtorig = uproot.open(self.l_root[file_index])[self.n_treeevt]
dfevtorig = treeevtorig.pandas.df(branches=self.v_evt)
# Only save events within the given run period & required centrality
dfevtorig = selectdfrunlist(dfevtorig, self.runlist, "run_number")
dfevtorig = selectdfquery(dfevtorig, self.s_cen_unp)
# Reset dataframe index and save to "original" pickle file
dfevtorig = dfevtorig.reset_index(drop=True)
dfevtorig.to_pickle(self.l_evtorig[file_index])
# Select "good" events and save to a second pickle file
dfevt = selectdfquery(dfevtorig, self.s_good_evt_unp)
dfevt = dfevt.reset_index(drop=True)
dfevt.to_pickle(self.l_evt[file_index])
# Open root file again, get the reconstructed tree into a dataframe
treereco = uproot.open(self.l_root[file_index])[self.n_treereco]
if not treereco:
print('Couldn\'t find tree %s in file %s' % \
(self.n_treereco, self.l_root[file_index]))
dfreco = treereco.pandas.df(branches=self.v_all)
# Only save events within the given run period & required cuts
dfreco = selectdfrunlist(dfreco, self.runlist, "run_number")
dfreco = selectdfquery(dfreco, self.s_reco_unp)
dfreco = pd.merge(dfreco, dfevt, on=self.v_evtmatch)
if 'Jet' not in self.case:
isselacc = selectfidacc(dfreco.pt_cand.values,
dfreco.y_cand.values)
dfreco = dfreco[np.array(isselacc, dtype=bool)]
if self.b_trackcuts is not None:
dfreco = filter_bit_df(dfreco, self.v_bitvar, self.b_trackcuts)
dfreco[self.v_isstd] = np.array(tag_bit_df(dfreco, self.v_bitvar,
self.b_std),
dtype=int)
dfreco = dfreco.reset_index(drop=True)
if self.mcordata == "mc":
dfreco[self.v_ismcsignal] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcsig),
dtype=int)
dfreco[self.v_ismcprompt] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcsigprompt),
dtype=int)
dfreco[self.v_ismcfd] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcsigfd),
dtype=int)
dfreco[self.v_ismcbkg] = np.array(tag_bit_df(
dfreco, self.v_bitvar, self.b_mcbkg),
dtype=int)
# Save reconstructed data to another pickle file
dfreco.to_pickle(self.l_reco[file_index])
if self.mcordata == "mc":
treegen = uproot.open(self.l_root[file_index])[self.n_treegen]
dfgen = treegen.pandas.df(branches=self.v_gen)
dfgen = selectdfrunlist(dfgen, self.runlist, "run_number")
dfgen = pd.merge(dfgen, dfevtorig, on=self.v_evtmatch)
dfgen = selectdfquery(dfgen, self.s_gen_unp)
dfgen[self.v_isstd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_std),
dtype=int)
dfgen[self.v_ismcsignal] = np.array(tag_bit_df(
dfgen, self.v_bitvar, self.b_mcsig),
dtype=int)
dfgen[self.v_ismcprompt] = np.array(tag_bit_df(
dfgen, self.v_bitvar, self.b_mcsigprompt),
dtype=int)
dfgen[self.v_ismcfd] = np.array(tag_bit_df(dfgen, self.v_bitvar,
self.b_mcsigfd),
dtype=int)
dfgen[self.v_ismcbkg] = np.array(tag_bit_df(
dfgen, self.v_bitvar, self.b_mcbkg),
dtype=int)
dfgen = dfgen.reset_index(drop=True)
dfgen.to_pickle(self.l_gen[file_index])