def buildTable(lams, cells, func):
results = {}
for cell in cells:
newCell = Kali.CellID(func(cell))
if lams.has_key(newCell):
results[cell] = lams[newCell]
return results
def fakeCell(c):
"""
Define 'fake' cell (useful) for grouping
"""
i = c.row()
j = c.col()
z = math.sqrt((i - 31.5)**2 + (j - 31.5)**2)
z -= 5
z = abs(z)
z = int(z / 4)
nc = Kali.CellID(c.calo(), c.area(), z, 1)
return nc
def mergefiles(filelist, outputfile):
"""
The only one interesting function: merge Kali-databases
"""
try:
import KaliCalo.Kali as Kali
histos = Kali.HistoMap()
histos.read(filelist)
histos.save(outputfile)
except:
return -1
return 0
import os
os.system(' ls -al ')
parser = POOLCatalogParser()
lfns = parser.lfns()
pfns = parser.pfns()
print 'LFNs: ', lfns
print 'PFNS: ', pfns
## try to do something useful:
import KaliCalo.Kali as Kali
from KaliCalo.Pi0FillHistos import fillDataBase
from GaudiKernel.SystemOfUnits import GeV
lambdas = Kali.LambdaMap()
histos, lambdas, badfiles = fillDataBase(lambdas,
pfns,
dbase_name='kali_db.gz',
Unit=GeV)
os.system(" ls -al ")
# =============================================================================
# The END
# =============================================================================
def fakeCell(cell):
nc = Kali.CellID(cell.calo(), cell.area(), cell.row(), 1)
return nc
def analyse(histomap, lambdamap, fitted=False, manager=None, nHistos=60):
"""
Analyse the histograms
"""
histomap = getHistosFromDB(histomap)
if not fitted:
histomap = Task.fitHistos(histomap, manager, nHistos)
keys = histomap.keys()
keys.sort()
for key in Cells.Zones:
if key in keys: keys.remove(key)
keys.insert(0, key)
import sets
badCells = sets.Set()
notFitCells = sets.Set()
lowCells = sets.Set()
ikey = 0
nFits = 0
iCnt = Kali.Counter()
for key in keys:
ikey += 1
if max(histomap[key].entries()) < 100:
lowCells.add(key)
continue
hs = histomap[key]
if key in Cells.Zones:
r0 = FitUtils.getPi0Params(hs[0])
r1 = FitUtils.getPi0Params(hs[1])
r2 = FitUtils.getPi0Params(hs[2])
if r0 and r1 and r2:
print ' MASS : %-24.24s %-24.24s %-24.24s ' % (r0[1], r1[1],
r2[1]), key
print ' SIGMA: %-24.24s %-24.24s %-24.24s ' % (r0[2], r1[2],
r2[2]), key
print ' NUM0 : %-24.24s %-24.24s %-24.24s ' % (r0[0], r1[0],
r2[0]), key
print ' S/B : %-24.24s %-24.24s %-24.24s ' % (FitUtils.s2b(
hs[0]), FitUtils.s2b(hs[1]), FitUtils.s2b(hs[2])), key
print ' '
else:
print 'No fit information available for ZONE: ', key
continue
result = None
if hasattr(hs, 'result'): result = hs.result
if not result:
## print ' no reliable fits for ',key
notFitCells.add(key)
continue
iCnt += result[-1]
r0 = FitUtils.getPi0Params(hs[0])
r1 = FitUtils.getPi0Params(hs[1])
r2 = FitUtils.getPi0Params(hs[2])
r0 = r0 if 0 <= result[0] else None
r1 = r1 if 0 <= result[1] else None
r2 = r2 if 0 <= result[2] else None
if r0 and r0[2].value() < 5:
print r0, hs[0].GetName()
if not r0 and not r1 and not r2:
notFitCells.add(key)
continue
alpha = 1.0
mass = Kali.VE(135, 135**2)
sigma = Kali.VE(12, 12**2)
corr = Kali.VE(1, 1**2)
num = Kali.VE(1, 1**2)
Used = False
if r0 and (6 <= r0[2].value() <= 20) and (15 <= abs(r0[0].value())):
m = r0[1]
deltam = r0[1] - 135.0
corr1 = 1.0 - alpha * deltam / r0[1]
s = r0[2]
if 0.00001 < corr1.error() < 0.50:
corr = corr.mean(corr1)
mass = mass.mean(m)
if s.error() > 0.00001: sigma = sigma.mean(s)
else: sigma.setValue(s.value())
num = abs(r0[0])
Used = True
use = FitUtils.badParam(mass, sigma, num, corr) or not Used
if r1 and use:
m = r1[1]
deltam = r1[1] - 135.0
corr2 = 1.0 - alpha * deltam / r1[1]
s = r1[2]
if 0.00001 < corr2.error() < 0.50:
corr = corr.mean(corr2)
mass = mass.mean(m)
if s.error() > 0.00001: sigma = sigma.mean(s)
num += abs(r1[0])
Used = True
use = FitUtils.badParam(mass, sigma, num, corr) or not Used
if r2 and use:
m = r2[1]
deltam = r2[1] - 135.0
hc = hs.kappas()
kappa = hc[1]
corr3 = 1.0 - alpha * deltam / r2[1] / (1.0 - kappa)
s = r2[2]
if 0.00001 < corr3.error() < 0.50:
corr = corr.mean(corr3)
mass = mass.mean(m)
if s.error() > 0.00001: sigma = sigma.mean(s)
num += abs(r2[0])
if not Used:
badCells.add(key)
continue
lams = lambdamap[key]
## do not allow too large corrections per one step
if corr.value() < 0.90: corr.setValue(0.90)
if corr.value() > 1.10: corr.setValue(1.10)
lam = corr * lams[-1]
lams.append(lam.value())
nFits = Kali.VE(iCnt.flagMean(), iCnt.flagRMS())
mx = 100.0 / len(histomap)
if random.uniform(0, 1.0) < mx and r0 and r1 and r2:
print ' '
print ' KEY :', key, ikey, len(
histomap), result, iCnt.flag(), nFits
print ' NUM0 : %-24.24s %-24.24s %-24.24s ' % (r0[0], r1[0],
r2[0])
print ' MASS : %-24.24s %-24.24s %-24.24s ' % (
r0[1], r1[1], r2[1]), corr, [
'%.3f' % l for l in lambdamap[key]
]
print ' SIGMA : %-24.24s %-24.24s %-24.24s ' % (r0[2], r1[2],
r2[2])
print ' S/B : %-24.24s %-24.24s %-24.24s ' % (FitUtils.s2b(
hs[0]), FitUtils.s2b(hs[1]), FitUtils.s2b(hs[2]))
return (badCells, lowCells, notFitCells)
#!/usr/bin/env python
import KaliCalo.Kali as Kali
histos = Kali.HistoMap()
lambdas = Kali.LambdaMap()
mydir = '/afs/cern.ch/user/i/ibelyaev/cmtuser/DaVinci_v25r2p1/Phys/KaliCalo/cmt/'
histos.read(mydir + 'histos_db.gz')
from KaliCalo.Analyse import analyse
analyse(histos, lambdas)
lambdas.save('lambdas_db.gz')