def get_ami_events(ami_client, dataset):
results = atlas_api.get_dataset_info(ami_client, dataset)
if len(results) != 1:
print "WARNING: %d results returned from AMI, expected 1" % len(
results)
return int(results[0]["totalEvents"])
def getDatasetInfo(dataset, debug=False):
client = pyAMI.client.Client('atlas')
AtlasAPI.init()
results = AtlasAPI.get_dataset_info(client, dataset)
if len(results) != 1:
raise SystemExit('\n***EXIT*** no valid results for dataset %s'%dataset)
eff = None
for name, value in results[0].iteritems():
if name=='totalEvents':
nevents = value
elif name=='crossSection':
xsec = float(value)*1e6 #xsec is in nb, hence the 1e6 factor to get it in fb
elif name=='datasetNumber':
dsid = value
elif name=='genFiltEff':
eff = float(value)
#if geFiltEff is not available, use aprox_GenFiltEff
if eff is None:
for name, value in results[0].iteritems():
if name=='approx_GenFiltEff':
eff = float(value)
print '%s %e %e %s'%(dsid, xsec, eff, nevents)
if debug:
for name, value in results[0].iteritems():
print ' %s %s'%((name+':').ljust(24),value)
print ''
return dsid, xsec, eff, nevents
def getSampleWeight(dsid):
fname = ""
files = os.listdir(file_loc)
for f in files:
if dsid in f: fname = f.strip("_ntupleOutput\.root\/").strip("user\.tholmes\.")
fname = "m" + fname
info = AtlasAPI.get_dataset_info(client, fname)[0]
xs = float(info.get('approx_crossSection'))*float(info.get('approx_GenFiltEff'))
return xs
def encodeDSInfo(s, ldn):
datasetinfo = AtlasAPI.get_dataset_info(s.client, ldn)
neventsAMI = float(datasetinfo[0]["totalEvents"])
crosssectionAMI = float(datasetinfo[0]["crossSection"])
if "approx_GenFiltEff" in datasetinfo[0].keys():
filtereffAMI = float(datasetinfo[0]["approx_GenFiltEff"])
else:
filtereffAMI = None
return {"events": neventsAMI, "xsec": crosssectionAMI, "fit_eff": filtereffAMI, "ldn": ldn}
def getNumberAmi(dsName):
try:
InfoDict = AtlasAPI.get_dataset_info(client, dsName)
except pyAMI.exception.Error as bla:
print "Failed to query DS %s on AMI: Exception was \"%s\"" % (dsName,
bla)
return -1
try:
n_Events = InfoDict[0]["totalEvents"]
except KeyError:
print "Unable to decipher the InfoDict: "
print InfoDict
return -1
return int(n_Events)
def getXS(dsid):
xs = getCrossSection(int(dsid), verbose=False)
if xs == -1:
#print "using pyami for dsid", dsid
fname = ""
files = os.listdir(file_loc)
for f in files:
if dsid in f: fname = f.strip("_ntupleOutput\.root\/").strip("user\.tholmes\.")
fname = "m" + fname
info = AtlasAPI.get_dataset_info(client, fname)[0]
#if (info.get('crossSection_unit') != 'nano barn'): print 'WARNING: Cross section listed in', info.get('crossSection_unit'),"!!"
xs = 1000*float(info.get('approx_crossSection'))*float(info.get('approx_GenFiltEff'))
return xs
def genParamsFromParents(client, datasetName, datasetNumber):
from pyAMI.atlas.api import get_dataset_info
approx_GenFiltEff = None
xsec = None
prov = pyAMI.atlas.api.get_dataset_prov(client, datasetName)
for parent in prov['node']:
# minbias overlays are also parents so need to
# check the channel number
if int(parent[u'logicalDatasetName'].split(".")[1]) == datasetNumber:
parentinfo = get_dataset_info(client,
parent[u'logicalDatasetName'])[0]
if parentinfo.has_key(u'approx_GenFiltEff'):
approx_GenFiltEff = parentinfo[u'approx_GenFiltEff']
pass
if parentinfo.has_key(u'crossSection'
) and parentinfo[u'crossSection'] != u'NULL':
xsec = float(parentinfo[u'crossSection'])
pass
if approx_GenFiltEff and xsec:
break
pass
return (xsec, approx_GenFiltEff)
import argparse
import json, ast
parser = argparse.ArgumentParser(description='Helper to get nFiles of a list of datasets')
# parser.add_argument('datasets', type=str, help='file containing list of datasets')
parser.add_argument('file', type=argparse.FileType('r'), nargs='+')
args = parser.parse_args()
client = pyAMI.client.Client(
'atlas'
)
print "INFO - Checking AMI status of datasets"
for f in args.file:
print "INFO - Checking %s" %f
for dataset in f:
dataset = dataset.rstrip()
if dataset[:1] == '#':
continue
if dataset[:2] == 'mc' or dataset[:4] == 'data':
try:
result = AtlasAPI.get_dataset_info(client, dataset = dataset)
result = ast.literal_eval(json.dumps(result))
print(result[0]['nFiles'])
except:
print("ERROR - Not in AMI - %s" %dataset)
for prov in dsProv["node"]:
if prov['dataType'] == "EVNT":
thisProvDSName = prov['logicalDatasetName']
thisProvDSID = thisProvDSName.split(".")[1]
if thisProvDSID == dsID:
print "\tUsing ",thisProvDSName
inputDS.append(thisProvDSName)
def getUnitSF(unit):
if unit == "nano barn":
return 1000
print "Unknown unit..."
return 1.0
fh_out=open(args.output,'w') if args.output!=None else None
for ds in inputDS:
dsList = AtlasAPI.get_dataset_info(client,dataset=ds)
dsInfo = dsList[0]
#print dsInfo['logicalDatasetName']
#print "\tcross section",dsInfo["crossSection_mean"]
#print "\tfilter Eff.",dsInfo["GenFiltEff_mean"]
print "totalEvents:",dsInfo['totalEvents']
unit = dsInfo['crossSection_unit']
getSF = getUnitSF(unit)
if fh_out==None:
print dsInfo['datasetNumber']," ",dsInfo['physicsShort']," ",float(dsInfo["crossSection_mean"])*getSF," 1. ",float(dsInfo["GenFiltEff_mean"])," 1."
else:
fh_out.write("%s\t%s\t%e\t1.\t%e\t1.\n"%(dsInfo['datasetNumber'],dsInfo['physicsShort'],float(dsInfo["crossSection_mean"])*getSF,float(dsInfo["GenFiltEff_mean"])))
def main():
# configurable options
config = parseCmdLine(sys.argv[1:])
if (config.baseline or config.official) and config.sample:
print "--baseline, --official and --sample are mutually exclusive"
sys.exit(1)
if (config.baseline or config.official
or config.sample) and config.grl != "":
print "--grl is incompatible with --baseline, --official and --sample"
sys.exit(1)
# AMI client connection
client = pyAMI.client.Client('atlas')
pyAMI.client.endpoint = config.server
pyAMI.atlas.api.init()
# consistency checks
if config.whichMC15 != '':
if config.whichMC15 == 'week1' and config.prefix != 'mc15_week1':
print 'prefix changed to mc15_week1 in agrement with whichMC15'
config.prefix = 'mc15_week1'
elif config.whichMC15 == '50ns' and config.prefix != 'mc15_13TeV':
print 'prefix changed to mc15_13TeV in agrement with whichMC15'
config.prefix = 'mc15_13TeV'
elif config.whichMC15 == '25ns' and config.prefix != 'mc15_13TeV':
print 'prefix changed to mc15_13TeV in agrement with whichMC15'
config.prefix = 'mc15_13TeV'
# data type is NTUP_SUSY for 2011/2012 and AOD for 2014 on
datatype = config.datatype
if 'mc11_' in config.prefix or 'mc12_' in config.prefix or 'data11_' in config.prefix or 'data12_' in config.prefix:
datatype = '%.merge.NTUP_SUSY%'
# make list of official datasets (baseline+alt)
officialids = []
if config.official or config.baseline or config.sample:
if 'mc12_8TeV' in config.prefix or 'mc14_8TeV' in config.prefix:
import mc12_8TeV_MCSampleList as mcsl
elif 'mc14_13TeV' in config.prefix:
import mc14_13TeV_MCSampleList as mcsl
elif 'mc15_13TeV' in config.prefix:
import mc15_13TeV_MCSampleList as mcsl
elif 'mc15_week1' in config.prefix:
import mc15_13TeV_week1_MCSampleList as mcsl
else:
print '--official is only supported for mc12_8TeV, mc14_8TeV, mc14_13TeV, mc15_13TeV and mc15_week1'
sys.exit(1)
if config.sample:
officialids = mcsl.__dict__[str(config.sample)]
else:
officialids = mcsl.__dict__["lbaseline"]
if config.official:
officialids += mcsl.__dict__["lalt"]
elif config.grl != "":
if not os.path.exists(config.grl):
print 'Couldnot find GRL', config.grl
sys.exit(1)
pass
doc = ET.parse(config.grl)
for item in doc.findall('./NamedLumiRange/Metadata'):
if item.attrib['Name'] == 'RunList':
for r in item.text.split(','):
officialids.append(int(r))
pass
# get all datasets matching prefix & tag and then filter them
from pyAMI.atlas.api import get_dataset_info, list_datasets
alldatasets = []
if config.whichMC15 != '':
prefix = config.prefix
if prefix == 'mc15_week1': prefix = 'mc15_13TeV'
for tag in mc15_rtags[config.whichMC15]:
dskey = prefix + datatype + tag + config.tag
print 'Querying AMI for datasets matching pattern', dskey
alldatasets += list_datasets(client, dskey)
else:
prefix = config.prefix
if prefix == 'mc15_week1': prefix = 'mc15_13TeV'
dskey = config.prefix + datatype + config.tag
print 'Querying AMI for datasets matching pattern', dskey
alldatasets = list_datasets(client, dskey)
acceptedDS = []
for DSlist in alldatasets:
dsname = DSlist['ldn']
cut = False
for filter in filters:
if filter in dsname.split('.')[2]: cut = True
if (config.official or config.baseline
or config.sample or config.grl != "") and not int(
dsname.split('.')[1]) in officialids:
cut = True
if config.signal:
cut = True
for pattern in lsignals:
if pattern in dsname: cut = False
if cut: continue
acceptedDS.append(dsname)
pass
acceptedDS.sort()
# get informations for all accepted datasets
dsinfos = []
for dsname in acceptedDS:
dsinfos.append(get_dataset_info(client, dsname)[0])
pass
# write file
coveredids = set()
if not (config.suffix == ""):
myoutputfile = 'datasets_' + config.suffix + '.txt'
else:
myoutputfile = 'datasets.txt'
fout = open(myoutputfile, 'w')
for info in dsinfos:
try:
dsname = info['logicalDatasetName']
if config.grl == "":
generatorString = info['generatorName']
version = info['version']
if badDataset(dsname, generatorString, version): continue
availability = info['prodsysStatus']
if config.onlyComplete and availability != u'ALL EVENTS AVAILABLE':
print 'Skip incomplete dataset', dsname, availability
continue
nFiles = int(info['nFiles'])
if nFiles > 0 and config.prefix.startswith('data'):
fout.write(dsname + '\n')
elif nFiles > 0:
period = 'MC'
xsec = 0.
effic = 1.
if info.has_key('period'):
period = info['period']
else:
datasetNumber = int(info[u'datasetNumber'])
coveredids.add(datasetNumber)
# confirmed with AMI team that this should be enought, no need
# to re-implement get_dataset_xsec_effic for PyAMI5
# there are sometime problems in the propagation of these
# properties to the xAOD/derived datasets so go back in
# parentage to find the information
xsec = info[u'crossSection']
if info.has_key(u'approx_GenFiltEff'):
effic = info[u'approx_GenFiltEff']
if config.datatype == '%TRUTH1%':
effic = 1
if ((xsec == u'NULL'
or not info.has_key(u'approx_GenFiltEff'))
and not (config.datatype == '%TRUTH1%')):
xsec, effic = genParamsFromParents(
client, dsname, datasetNumber)
if not xsec: xsec = 0
if not effic:
print 'No approx_GenFiltEff found for', dsname, 'set to 0 !!!!'
effic = 0
pass
nevts = info['totalEvents']
nfiles = info['nFiles']
if not dsname.endswith('/'): dsname += '/'
fout.write("%s %s %s %s %s %s\n" %
(dsname, nevts, nfiles, period, xsec, effic))
except KeyError as prop:
print 'Missing property', prop, 'for dataset ', dsname, 'in AMI, skip'
fout.close()
if len(coveredids) == 0:
if not config.prefix.startswith('data'):
print 'Could not extract any channel IDs from datasets found, this is OK for data but suspicious for MC'
else:
for id in officialids:
if not id in coveredids:
print 'No dataset found for channel ', id
pass
for a in open('stupid.txt'):
b = a.strip().replace('/','')
if len(b) == 0:
continue
#Take AOD instead of DAOD (not the same original number of events)
b = b.replace('DAOD_HIGG4D2', 'AOD')
b = b.replace('_p2419', '')
# b = b.rstrip(b[-6:])
# print b
id = b.split('.')[1]
if id in Ids:
continue
Ids.append(id)
infos = AtlasAPI.get_dataset_info(client, b)[0]
xsec = float(infos['crossSection']) * 1000
filtstr = infos['approx_GenFiltEff']
if filtstr.find('N/A')>=0:
# print 'No filter efficiency available'
# print infos
filt = 1
else:
filt = float(filtstr)
kfac = 1.0
nevts = float(infos['totalEvents'])
lumi = nevts/ 1000. / (xsec * kfac * filt)
print id, '\t', '%s' % int(nevts), '\t', '{0}*{1}*{2}'.format(xsec, kfac, filt), '\t', '{0:1.2f}'.format(lumi),'\t', b
runList.append(key.GetName().split('_')[2])
runList.sort()
with open(args.output,'wb') as csvfile:
cwriter = csv.writer(csvfile,delimiter=',')
cwriter.writerow(['run',
'luminosity',
'CBC selected',
'Initial',
'GRL',
'event cleaning',
'trigger',
'pT_lead',
'n_fatjet==3',
'n_fatjet==3 && b-tag',
'n_fatjet==4 && MJ < 600',
'nfatjet==4 && MJ < 600 && b-tag',
'n_fatjet >= 5 && MJ < 600',
'n_fatjet >= 5 && MJ < 600 && b-tag'])
for run in runList:
h = f.Get('h_cutflow_'+run)
row = [h.GetBinContent(i) for i in range(1,h.GetNbinsX()+1)]
row.insert(0,'')
row.insert(0,int(run))
dsName = 'data15_13TeV.00'+run+'.physics_Main.merge.DAOD_EXOT3.r7562_p2521_p2614'
d=AtlasAPI.get_dataset_info(client,dsName)[0]
totalEvents = int(d['totalEvents'])
print 'DSID: %s, AMI = %i, CBC = %i, Initial = %i' % (run,totalEvents,row[2],row[3])
cwriter.writerow(row)
runList = []
client = pyAMI.client.Client('atlas')
AtlasAPI.init()
for key in f.GetListOfKeys():
if 'h_cutflow' in key.GetName():
runList.append(key.GetName().split('_')[2])
runList.sort()
with open(args.output, 'wb') as csvfile:
cwriter = csv.writer(csvfile, delimiter=',')
cwriter.writerow([
'run', 'luminosity', 'CBC selected', 'Initial', 'GRL',
'event cleaning', 'trigger', 'pT_lead', 'n_fatjet==3',
'n_fatjet==3 && b-tag', 'n_fatjet==4 && MJ < 600',
'nfatjet==4 && MJ < 600 && b-tag', 'n_fatjet >= 5 && MJ < 600',
'n_fatjet >= 5 && MJ < 600 && b-tag'
])
for run in runList:
h = f.Get('h_cutflow_' + run)
row = [h.GetBinContent(i) for i in range(1, h.GetNbinsX() + 1)]
row.insert(0, '')
row.insert(0, int(run))
dsName = 'data15_13TeV.00' + run + '.physics_Main.merge.DAOD_EXOT3.r7562_p2521_p2614'
d = AtlasAPI.get_dataset_info(client, dsName)[0]
totalEvents = int(d['totalEvents'])
print 'DSID: %s, AMI = %i, CBC = %i, Initial = %i' % (run, totalEvents,
row[2], row[3])
cwriter.writerow(row)
def SampleHandler_QueryAmi(samples):
# set up an AMI client
# This is the basic minimum - and it will look for an encrypted file with your user credentals
# If it does not find that it will try for a VOMS proxy
# Make the encrypted file by running the amo command
# ami auth
# first.
# In the pyAMI doc you will find an example of how to get your program to request a
# the user to make a file.
# https://atlas-ami.cern.ch/AMI/pyAMI/examples/api.html
amiClient = pyAMI.client.Client('atlas')
# Extract from your mail
###############
# The quantities in the class MetaDataSample are pretty much all I need:
# * whether it is data or MC
# * the luminosity of the sample
# * the k-factor of the sample (only for MC)
# * the number of events in the sample
# * the cross section of the sample (only for MC)
# * the filter efficiency
data = ROOT.SH.MetaDataQuery()
data.messages = 'done by ami query'
# I am assuming that "samples" is a list of dataset names, and that
# the user already checked that they exist and are valid
for sample in samples:
sample_noscope = sample.split(':')[-1]
mydata = ROOT.SH.MetaDataSample(sample)
# The first question you ask is it data or mc.
# Actually you should be able to tell this without ambiguity from the name
# without going to the trouble of a request to AMI.
# description: 1 for data, 0 for MC, or -1 if this is not known.
mydata.source = 'https://atlas-ami.cern.ch/AMI/pyAMI/'
mydata.unknown = 0
if (sample.startswith("mc")):
mydata.isData = 0
pass
elif (sample.startswith("data")):
mydata.isData = 1
pass
else:
mydata.isData = -1
pass
# You are calling ths AMI functions with tid suffixes.
# AMI does not specifically catalogue TID datasets so
# I am stripping off the suffix.
# Normally uses should not be concerned with these datasets
# but only with the containers.
# However if you are really only interested in the output of a particular
# prodsys task then we can do it - but it would be more complex
# as we need to redo the event and cross section calculations
# just for those tasks.
if (sample.find("_tid")):
print("Stripping tid suffix from " + sample)
sample = sample.split("_tid")[0]
pass
# All datasets should have the number of events.
# have to convert this to a long int I suppose?
amiinfo = get_dataset_info(amiClient, sample_noscope)[0]
mydata.nevents = long(amiinfo['totalEvents'])
# AMI does not yet have a function for getting luminosity.
# It IS on the todo list, as luminosity info per run is available
# in COMA, and AMI has access to the information in principle
# So this is in part a place holder
# I do not know anything about k-factor. We have no such parameter sent to us.
# This should be taken up with the MC people I suppose.
if (mydata.isData == 1):
# get luminosity for the run
mydata.crossSection = -1
mydata.filterEfficiency = -1
pass
else:
mydata.luminosity = -1
# MC - can get cross-section and filter efficiency
xsec = float(amiinfo['approx_crossSection'])
effic = float(amiinfo['approx_GenFiltEff'])
# + conversion string to float.
mydata.crossSection = xsec
mydata.filterEfficiency = effic
if mydata.crossSection > 0 and mydata.filterEfficiency > 0:
mydata.luminosity = float(
float(mydata.nevents) /
(mydata.crossSection * mydata.filterEfficiency))
pass
pass
data.addSample(mydata)
# print "cross section = "+str(mydata.crossSection)+", filter efficiency = "+str(mydata.filterEfficiency)+", nEvents= "+str(mydata.nevents)
pass
return data
ROOT.SH.ScanDir().sampleDepth(0).samplePattern(args.eosDataSet).scanEOS(sh_all, base)
else:
raise Exception("What just happened?")
if args.xsecFromAMI:
import pyAMI.client
import pyAMI.atlas.api as AtlasAPI
from pyAMI.atlas.api import get_dataset_info
client = pyAMI.client.Client("atlas")
AtlasAPI.init()
evntName = line.rstrip()
if "merge.DAOD_EXOT3" in evntName:
evntName = evntName.replace("merge.DAOD_EXOT3", "evgen.EVNT")
evntName = evntName[0 : evntName.find("evgen.EVNT") + 16]
# d=AtlasAPI.get_dataset_info(client,line.rstrip())[0]
d = AtlasAPI.get_dataset_info(client, evntName)[0]
filtEff = 1
xsec = 1
if "genFiltEff" in d:
if d["genFiltEff"] != "NULL":
filtEff = float(d["genFiltEff"])
elif "GenFiltEff_mean" in d:
if d["GenFiltEff_mean"] != "NULL":
filtEff = float(d["GenFiltEff_mean"])
if "crossSection" in d:
if d["crossSection"] != "NULL":
xsec = float(d["crossSection"])
dsid = str(d["datasetNumber"])
sh_all.setMetaString(line.rstrip().rstrip("/"), "dsid", dsid)
sh_all.setMetaDouble(line.rstrip().rstrip("/"), "weight_xs", filtEff * xsec)
print(
# Don't check data
if did.startswith("00"): continue
# Get all possible AODs for the output ntuple
base_files = AtlasAPI.list_datasets(
client,
patterns=['mc15_13TeV.' + did + '%merge.AOD%' + tags + '%'],
type='AOD')
print "-------------------------------------------------------------------------"
print "Ntuple name:", d
print base_files[0]['ldn']
if len(base_files) > 1:
print "Found more than one matching AOD."
print "For file", f
print "using", base_files[0]['ldn']
info = AtlasAPI.get_dataset_info(client, base_files[0]['ldn'])[0]
n_real_events = info.get('totalEvents')
if float(n_real_events) < float(n_aod_events):
print "\033[91m Error in dsid", did
bad_dids += [did]
print "\t Real number of AOD events:", n_real_events
print "\t EventCountHist number: ", n_aod_events
print "\t Ratio: ", float(n_aod_events) / float(
n_real_events)
print '\033[0m'
print "All dids with bad weights:"
print bad_dids
raise Exception("What just happened?")
if args.xsecFromAMI:
import pyAMI.client
import pyAMI.atlas.api as AtlasAPI
from pyAMI.atlas.api import get_dataset_info
client = pyAMI.client.Client('atlas')
AtlasAPI.init()
evntName = line.rstrip()
if 'merge.DAOD_EXOT3' in evntName:
evntName = evntName.replace(
'merge.DAOD_EXOT3', 'evgen.EVNT')
evntName = evntName[0:evntName.
find('evgen.EVNT') +
16]
# d=AtlasAPI.get_dataset_info(client,line.rstrip())[0]
d = AtlasAPI.get_dataset_info(
client, evntName)[0]
filtEff = 1
xsec = 1
if 'genFiltEff' in d:
if d['genFiltEff'] != 'NULL':
filtEff = float(d['genFiltEff'])
elif 'GenFiltEff_mean' in d:
if d['GenFiltEff_mean'] != 'NULL':
filtEff = float(d['GenFiltEff_mean'])
if 'crossSection' in d:
if d['crossSection'] != 'NULL':
xsec = float(d['crossSection'])
dsid = str(d['datasetNumber'])
sh_all.setMetaString(line.rstrip().rstrip('/'),
'dsid', dsid)
sh_all.setMetaDouble(line.rstrip().rstrip('/'),
inDS = inDS.split(":")[1]
items = inDS.split(".")
tag = items[-1].split("_")[0]
evntDS = ".".join(items[:-3] + ['evgen', 'EVNT', tag])
tokens = dataset.split(".")
mystr = tokens[2].split("_")[-1]
numbers = ""
for character in mystr:
if character.isdigit() :
numbers = numbers+character
tag = "JZ{0}W".format(numbers)
info = AtlasAPI.get_dataset_info(client, evntDS)[0]
infoDict = {}
infoDict['crossSection'] = info['crossSection_max']
infoDict['filterEff'] = info['GenFiltEff_mean']
infoDict['nEvt'] = info['totalEvents']
print infoDict
mydict[tag] = infoDict
outfile = "pyAMIInfoForFiles_{0}.py".format(outFileTag)
w = open(outfile, "w")
w.write("amiInfoDict = {\n")
for key, val in sorted(mydict.items()) :
w.write("'{0}'".format(key)+" : "+ "{" )
for ikey,ival in val.items() :
thisProvDSID = thisProvDSName.split(".")[1]
if thisProvDSID == dsID:
print "\tUsing ", thisProvDSName
inputDS.append(thisProvDSName)
def getUnitSF(unit):
if unit == "nano barn":
return 1000
print "Unknown unit..."
return 1.0
fh_out = open(args.output, 'w') if args.output != None else None
for ds in inputDS:
dsList = AtlasAPI.get_dataset_info(client, dataset=ds)
dsInfo = dsList[0]
#print dsInfo['logicalDatasetName']
#print "\tcross section",dsInfo["crossSection_mean"]
#print "\tfilter Eff.",dsInfo["GenFiltEff_mean"]
print "totalEvents:", dsInfo['totalEvents']
unit = dsInfo['crossSection_unit']
getSF = getUnitSF(unit)
if fh_out == None:
print dsInfo['datasetNumber'], " ", dsInfo['physicsShort'], " ", float(
dsInfo["crossSection_mean"]) * getSF, " 1. ", float(
dsInfo["GenFiltEff_mean"]), " 1."
else:
fh_out.write("%s\t%s\t%e\t1.\t%e\t1.\n" %
(dsInfo['datasetNumber'], dsInfo['physicsShort'],
float(dsInfo["crossSection_mean"]) * getSF,
def main():
logging.basicConfig(format='%(levelname)s:%(message)s')
import time, datetime
from pytz import timezone
import argparse
try:
import pyAMI.client
import pyAMI.atlas.api as AtlasAPI
import pyAMI.config
except ImportError:
logging.error(
"Unable to find pyAMI client. Please try this command first: lsetup pyAMI"
)
return -1
extraFieldDefaults = {
} #{"approx_crossSection":None,"approx_GenFiltEff":1.0}
fieldDefaults = {"subprocessID": 0, "dataset_number": 0}
#populate the fieldDefaults ... for all, assume 'None'
for field in pyAMI.config.tables['datasets'].keys():
if str(field) == "cross_section":
continue #special exception because this field only present in
if str(field) in fieldDefaults.keys(): continue
if str(field).startswith("@"): continue
fieldDefaults[str(field)] = None
import commands
#check the voms proxy
status, out = commands.getstatusoutput("voms-proxy-info -fqan -exists")
if status != 0:
logging.error(
"Please renew your certificate with this command: voms-proxy-init -voms atlas"
)
return -1
try:
client = pyAMI.client.Client('atlas')
AtlasAPI.init()
except:
logging.error(
"Could not establish pyAMI session. Are you sure you have a valid certificate? Do: voms-proxy-init -voms atlas"
)
return -1
#need to collect the ami dataset parameter defaults
paramExplains = [] #for the help message only
paramUnits = dict()
paramDefaults = {}
res = client.execute('ListPhysicsParameterDefs', format='dom_object')
for r in res.get_rows(): #r is OrderedDict
explainString = "%s: %s" % (r[u'PARAMNAME'], r[u'DESCRIPTION'])
if r[u'UNITS'] != u'NULL':
explainString += " (units: %s)" % r[u'UNITS']
paramUnits[r[u'PARAMNAME']] = r[u'UNITS']
if r[u'HASDEFAULT'] == u'N': paramDefaults[str(r[u'PARAMNAME'])] = None
else:
explainString += " (default value = %s)" % r[u'DEFAULTVALUE']
if r[u'PARAMTYPE'] == u'number':
paramDefaults[str(r[u'PARAMNAME'])] = float(
r[u'DEFAULTVALUE']
) #FIXME: Assumes all parameters are floats
elif r[u'PARAMTYPE'] == u'string':
paramDefaults[str(r[u'PARAMNAME'])] = str(r[u'DEFAULTVALUE'])
paramExplains += [explainString]
paramDefaults["crossSection_pb"] = None
paramUnits["crossSection_pb"] = "pb"
paramExplains += [
"crossSection_pb: Same as crossSection except in pb units (units: pb)"
]
cern_time = timezone('UCT')
current_time = datetime.datetime.fromtimestamp(
time.time(), cern_time).strftime('%Y-%m-%d %H:%M:%S')
from argparse import RawTextHelpFormatter
parser = argparse.ArgumentParser(description=__doc__,
formatter_class=RawTextHelpFormatter)
parser.add_argument('--inDS',
nargs='+',
default=[""],
help="List of datasets to retrieve parameters for")
parser.add_argument(
'--inDsTxt',
default="",
help=
"Alternative to --inDS, can specify the datasets from an input file")
parser.add_argument(
'--fields',
nargs='+',
help=
"List of parameters to extract. Available parameters are: \n\n %s\n\nYou can also include any from:\n %s\nYou can also do keyword_xxx to add a bool branch for keywords"
% ("\n ".join(paramExplains),
", ".join(fieldDefaults.keys() + extraFieldDefaults.keys())),
default=["dataset_number", "crossSection", "kFactor", "genFiltEff"])
parser.add_argument(
'--timestamp',
default=current_time,
help=
"The timestamp to query parameters at, specified in Universal Central Time (UCT). If left blank, will take the current time"
)
parser.add_argument(
'--physicsGroups',
nargs='+',
default=["PMG,MCGN"],
help=
"Physics group from which to retrieve parameters, listed in order of priority (highest first). Default value is 'PMG,MCGN' (i.e. try to use PMG values, fallback on MCGN values if unavailable). Allowed groups are:\n PMG (this is the PMG's group name), BPHY, COSM, DAPR, EGAM, EXOT, FTAG, HIGG, HION, IDET, IDTR, JETM, LARG, MCGN (this is the AMI default group name), MDET, MUON, PHYS, REPR, SIMU, STDM, SUSY, TAUP, TCAL, TDAQ, THLT, TOPQ, TRIG, UPGR, VALI"
)
parser.add_argument(
'--oldTimestamp',
default="",
help=
"If specified, will instead display a diff between the old and new timestamp, showing explanation of any changed parameters"
)
parser.add_argument(
'--explainFields',
nargs='+',
default=[],
help=
"The fields you would like explained .. will appear as comment lines after each row in the output"
)
parser.add_argument(
'--explainInfo',
nargs='+',
default=['explanation', 'insert_time'],
help=
"Properties of the parameter you want to show in the explanation. Can list from: explanation, insert_time, physicsGroup, createdby. Default is: explanation,insert_time"
)
parser.add_argument(
'--outFile',
default=sys.stdout,
type=argparse.FileType('w'),
help="Where to print the output to. Leave blank to print to stdout")
parser.add_argument(
'--delim',
default="",
help=
"The delimiter character. Defaults to spaces leading to nice formatting table"
)
parser.add_argument('-v',
action='store_true',
help="Verbose output for debugging")
args = parser.parse_args()
if args.v: logging.getLogger().setLevel(logging.DEBUG)
else: logging.getLogger().setLevel(logging.INFO)
logging.debug(args.inDS)
logging.debug(args.fields)
logging.debug(args.timestamp)
if args.timestamp == "the dawn of time":
logging.error(
"Unfortunately we don't know any parameters from this time period... but we're working on it!"
)
return 9999
#split elements of fields by comma to get full list
args.fields = sum((y.split(',') for y in args.fields), [])
args.fields = [x.strip() for x in args.fields] #strips whitespace
#look for keyword_ fields, these are special ...
args.keywords = []
for f in args.fields:
if f.startswith("keyword_"):
k = f[8:]
#and then add each keyword to the extraFieldDefaults so it is recognised thusly
extraFieldDefaults["keyword_%s" % k] = bool(False)
args.keywords += [k]
#same for physics groups
args.physicsGroups = sum((y.split(',') for y in args.physicsGroups), [])
args.physicsGroups = [x.strip()
for x in args.physicsGroups] #strips whitespace
#same for explainFields and explainInfo
args.explainFields = sum((y.split(',') for y in args.explainFields), [])
args.explainFields = [x.strip()
for x in args.explainFields] #strips whitespace
args.explainInfo = sum((y.split(',') for y in args.explainInfo), [])
args.explainInfo = [x.strip()
for x in args.explainInfo] #strips whitespace
if args.inDsTxt != '': args.inDS = readDsFromFile(args.inDsTxt)
#and same for inDS
args.inDS = sum((y.split(',') for y in args.inDS), [])
args.inDS = [x.strip() for x in args.inDS] #strips whitespace
#1. check field values are allowed, we obtain default field values at same time..
#2. For each entry in inDS, if contains wildcard we obtain list of DS, otherwise check DS exists. During this time we obtain the datasetid and numEvents properties, incase we need them
#3. For each of these DS, get parameters from ami matching the timestamp. Organize into fields and index by subprocessID
#4. Output a line to our output file
#1.
#before adding all the ami parameters, identify which of provided fields are: 1). Obtained from list_datasets command (dsFields) 2). actual parameters
dsFields = [
x for x in args.fields
if x in fieldDefaults.keys() and x not in ["subprocessID", "ldn"]
]
extraFields = [x for x in args.fields if x in extraFieldDefaults.keys()]
paramFields = [x for x in args.fields if x in paramDefaults.keys()]
if len(paramFields) > 0 and args.physicsGroups == [""]:
logging.error(
"You must specify at least one physics group. See -h for allowed groups"
)
return -1
#combine paramDefaults with fieldDefaults
fieldDefaults.update(paramDefaults)
#and with extra fields
fieldDefaults.update(extraFieldDefaults)
for field in args.fields:
if field not in fieldDefaults:
logging.error("%s is not a recognised field. Allowed fields are:" %
field)
logging.error(fieldDefaults.keys())
return -1
if args.oldTimestamp != "":
logging.info("oldTimestamp option specified. Running in diff mode...")
args.explainFields = args.fields
args.explainInfo = [
"explanation", "insert_time", "physicsGroup", "createdby"
]
#2.
#replace all '*' with '%' and strip "/"
args.inDS = [ds.replace("*", "%") for ds in args.inDS]
args.inDS = [ds.rstrip("/") for ds in args.inDS]
if len(args.inDS) == 0 or (len(args.inDS) == 1 and args.inDS[0] == ""):
logging.error(
"No datasets provided. Please specify datasets with the --inDS or --inDsTxt options"
)
return -1
logging.info(
"Fetching list of datasets from AMI (this may take a few minutes)...")
#obtain list of datasets
res = AtlasAPI.list_datasets(
client,
patterns=args.inDS,
fields=dsFields + ['ldn'],
ami_status="VALID"
) #changed status from %, to only catch valid now: wb 08/2015
logging.info("...Found %d datasets matching your selection" % len(res))
if len(res) == 0:
return 0
#NOTE: Should we allow retrieval of the extra information: keyword, genfiltereff, approx crossection, .. these all come from GetDatasetInfo ami command
dataset_values = dict()
for r in res:
mydict = dict()
dataset_values[str(r['ldn'])] = mydict
for field in r.items():
if str(field[0]) == "ldn": continue
if str(field[0]) not in args.fields: continue
mydict[str(field[0])] = str(field[1])
#also if we have the 'extra fields or keywords' we will need to execute AtlasAPI.get_dataset_info ..
if len(extraFields) > 0 or len(args.keywords) > 0:
info_res = AtlasAPI.get_dataset_info(client, str(r['ldn']))
#print(info_res)
if len(info_res) == 0:
logging.error("Unable to retrieve dataset info for %s" %
str(r['ldn']))
return -1
for field in extraFields:
#ignore the keyword_ fields
if field.startswith("keyword_"): continue
mydict[field] = float(info_res[0][unicode(field)]) if isfloat(
info_res[0][unicode(field)]) else extraFieldDefaults[field]
for k in args.keywords:
mydict["keyword_%s" % k] = int(
(k in str(info_res[0][unicode('keyword')]).split(",")))
#sort dataset_values as well as possible
from collections import OrderedDict
sorted_values = OrderedDict()
for ds in args.inDS:
if ds in dataset_values.keys():
sorted_values[ds] = dataset_values[ds]
for ds in sorted(dataset_values):
if ds not in sorted_values.keys():
sorted_values[ds] = dataset_values[ds]
dataset_values = sorted_values
logging.debug(dataset_values)
#res = client.execute(['GetDatasetInfo
for ds in args.inDS:
if '%' not in ds and ds not in dataset_values.keys():
logging.warning("Unknown dataset: %s" % ds)
datasetsToQuery = ",".join(dataset_values.keys())
#if using inDsTxt, retain any comment or blank lines in structure of output
complete_values = OrderedDict()
if args.inDsTxt != "":
# read lines
commentcount = 0
import re
txt = open(args.inDsTxt)
for tmpLine in txt:
# remove \n
tmpLine = re.sub('\n', '', tmpLine)
# remove white spaces
tmpLine = tmpLine.strip()
# skip comment or empty
if tmpLine.startswith('#') or tmpLine == '':
complete_values['comment%d' % (commentcount)] = tmpLine
commentcount = commentcount + 1
continue
# append
tmpLine = tmpLine.rstrip("/")
if tmpLine in dataset_values.keys():
complete_values[tmpLine] = dataset_values[tmpLine]
else:
print("cannot find %s" % tmpLine)
# close file
txt.close()
dataset_values = complete_values
logging.info(
"Obtaining %s for selected datasets at timestamp=%s... (please be patient)"
% (args.fields, args.timestamp))
#do as one query, to be efficient
if (args.timestamp == current_time):
res = client.execute([
'GetPhysicsParamsForDataset',
"--logicalDatasetName=%s" % datasetsToQuery,
"--timestamp='%s'" % args.timestamp
],
format='dom_object')
else:
res = client.execute([
'GetPhysicsParamsForDataset',
"--logicalDatasetName=%s" % datasetsToQuery,
"--timestamp='%s'" % args.timestamp, "--history=true"
],
format='dom_object')
#organize results by dataset
parameterQueryResults = dict()
for r in res.get_rows():
if r[u'logicalDatasetName'] not in parameterQueryResults.keys():
parameterQueryResults[r[u'logicalDatasetName']] = []
parameterQueryResults[r[u'logicalDatasetName']] += [
r
] #puts row in the list for this dataset
if args.oldTimestamp != "":
logging.info(
"Obtaining %s for selected datasets at timestamp=%s... (please be patient)"
% (args.fields, args.oldTimestamp))
res2 = client.execute([
'GetPhysicsParamsForDataset',
"--logicalDatasetName=%s" % datasetsToQuery,
"--timestamp='%s'" % args.oldTimestamp, "--history=true"
],
format='dom_object')
old_parameterQueryResults = dict()
for r in res2.get_rows():
if r[u'logicalDatasetName'] not in old_parameterQueryResults.keys(
):
old_parameterQueryResults[r[u'logicalDatasetName']] = []
old_parameterQueryResults[r[u'logicalDatasetName']] += [
r
] #puts row in the list for this dataset
headerString = ""
doneHeader = False
commentCache = ""
commentCount = 0
#result is a list of lists (each list is 1 row)
outputTable = []
tableHeaders = []
for ds in dataset_values.keys():
if ds.startswith('comment'):
if commentCount > 0: commentCache += "\n"
commentCache += dataset_values[ds]
commentCount = commentCount + 1
continue
#obtain list of parameters for this dataset
#if(args.timestamp==current_time):
# res = client.execute(['GetPhysicsParamsForDataset',"--logicalDatasetName=%s"% ds,"--timestamp='%s'"%args.timestamp], format='dom_object')
#else:
# res = client.execute(['GetPhysicsParamsForDataset',"--logicalDatasetName=%s"% ds,"--timestamp='%s'"%args.timestamp,"--history=true"], format='dom_object')
res = parameterQueryResults.get(ds, [])
if args.oldTimestamp != "":
res2 = old_parameterQueryResults.get(ds, [])
#first we have to determine how many subprocesses this ds has
dsSubprocesses = [0] #always have the 0 subprocess
for r in res:
sp = int(r[u'subprocessID'])
if sp not in dsSubprocesses: dsSubprocesses += [sp]
#now for each subprocess we have to locate each required field value (in paramFields)
#rank by physicsGroup
for sp in dsSubprocesses:
paramVals = dict()
paramVals2 = dict()
groupsWithVals = dict() #held for helpful output
#need to keep explanations for requested fields
explainInfo = dict()
for i in args.explainFields:
explainInfo[i] = dict()
for param in paramFields:
groupsWithVals[param] = []
bestGroupIndex = len(args.physicsGroups)
import copy
paramVals[param] = copy.copy(fieldDefaults[param])
for r in res:
if int(r[u'subprocessID']) != sp: continue
if str(r[u'paramName']) != param and not (
param == "crossSection_pb"
and str(r[u'paramName']) == "crossSection"):
continue
if str(r[u'physicsGroup']) not in args.physicsGroups:
groupsWithVals[param] += [(str(r[u'physicsGroup']),
str(r[u'paramValue']))]
continue
if args.physicsGroups.index(str(
r[u'physicsGroup'])) > bestGroupIndex:
continue
if args.physicsGroups.index(str(
r[u'physicsGroup'])) == bestGroupIndex:
logging.warning(
"Duplicate parameter %s for group %s in dataset %s (subprocess %d). Please report this!"
% (param, str(r[u'physicsGroup']), ds, sp))
paramVals[param] = str(r[u'paramValue'])
if param == "crossSection_pb":
paramVals[param] = str(
float(paramVals[param]) * 1000.0)
bestGroupIndex = args.physicsGroups.index(
str(r[u'physicsGroup']))
#keep the explanation info for the requested fields
if param in explainInfo.keys():
for e in args.explainInfo:
if unicode(e) not in r:
logging.error(
"Unrecognised explainInfo field: %s" % e)
return -1
explainInfo[param][e] = str(r[unicode(e)])
if args.oldTimestamp != "":
bestGroupIndex = len(args.physicsGroups)
paramVals2[param] = copy.copy(fieldDefaults[param])
for r in res2:
if int(r[u'subprocessID']) != sp: continue
if str(r[u'paramName']) != param and not (
param == "crossSection_pb"
and str(r[u'paramName']) == "crossSection"):
continue
if str(r[u'physicsGroup']) not in args.physicsGroups:
continue
if args.physicsGroups.index(str(
r[u'physicsGroup'])) > bestGroupIndex:
continue
if args.physicsGroups.index(str(
r[u'physicsGroup'])) == bestGroupIndex:
logging.warning(
"Duplicate parameter %s for group %s in dataset %s (subprocess %d). Please report this!"
% (param, str(r[u'physicsGroup']), ds, sp))
paramVals2[param] = str(r[u'paramValue'])
if param == "crossSection_pb":
paramVals2[param] = str(
float(paramVals2[param]) * 1000.0)
bestGroupIndex = args.physicsGroups.index(
str(r[u'physicsGroup']))
#at this stage, parameters reside in paramVals dict or dataset_values[ds] dict
#print them in the requested order .. if any is "None" then stop, because it doesn't have a default value and didn't find a value for it either
rowString = ""
rowList = []
firstPrint = False
for param in args.fields:
val = None
if param == "ldn": val = ds
elif param == "subprocessID": val = sp
elif param in dataset_values[ds].keys():
val = dataset_values[ds][param]
else:
val = paramVals.get(param, None)
if val == None:
if args.outFile != sys.stdout:
logging.warning(
"dataset %s (subprocess %d) does not have parameter %s, which has no default."
% (ds, sp, param))
if len(groupsWithVals.get(param, [])) > 0:
logging.warning(
"The follow physicsGroups have defined that parameter though:"
)
logging.warning(groupsWithVals[param])
val = "#UNKNOWN#"
#return -1
#if isfloat(str(val)): val = "%.6g" % float(val)
if args.oldTimestamp != "":
#diff val to old val
val2 = None
if param == "ldn": val2 = ds
elif param == "subprocessID": val2 = sp
elif param in dataset_values[ds].keys():
val2 = dataset_values[ds][param]
else:
val2 = paramVals2.get(param, None)
if val2 == None: val2 = "#UNKNOWN#"
#if isfloat(str(val2)): val2 = "%.6g" % float(val)
if (str(val) != str(val2)):
if not firstPrint: print("%s:" % ds)
firstPrint = True
print(" %s : %s ---> %s" %
(param, str(val2), str(val)))
print(" insert_time : %s" %
explainInfo[param]['insert_time'])
print(" explanation : %s" %
explainInfo[param]['explanation'])
print(" createdby : %s" %
explainInfo[param]['createdby'])
print(" physicsGroup : %s" %
explainInfo[param]['physicsGroup'])
continue
rowList += [str(val)]
if rowString != "" and args.delim != "":
rowString += args.delim
rowString += str(val)
#inspect the type of str(val) to build up the header
if not doneHeader:
headerString += param
if args.outFile != sys.stdout:
if type(fieldDefaults[param]) == bool:
headerString += "/O:"
elif type(fieldDefaults[param]) == int:
headerString += "/I:"
elif type(fieldDefaults[param]) == float:
headerString += "/D:"
elif isfloat(str(val)):
headerString += "/D:"
#elif isint(str(val)): headerString += "/I:" TO BE SAFE WE MAKE ALL NUMERIC FIELDS FLOATS, EXCEPT if the default value is type int
else:
headerString += "/C:"
else:
v = param
if param in paramUnits:
headerString += " [%s]" % paramUnits[param]
v += " [%s]" % paramUnits[param]
tableHeaders += [v]
headerString += " "
if args.oldTimestamp != "":
continue #print nothing more for diff mode
if not doneHeader:
doneHeader = True
if args.outFile != sys.stdout:
print(headerString[:-1], file=args.outFile)
if commentCount > 0:
if args.outFile != sys.stdout and args.delim != "":
print(commentCache, file=args.outFile)
outputTable += [["COMMENT", commentCache]]
commentCache = ''
commentCount = 0
if args.outFile != sys.stdout and args.delim != "":
print(rowString, file=args.outFile)
outputTable += [rowList]
#also print the required explanations
for (field, expl) in explainInfo.items():
outString = "#%s: { " % field
doneFirst = False
for eField in args.explainInfo:
if doneFirst: outString += " , "
if not eField in expl.keys():
outString += " %s: <NONE .. value is default>" % eField
else:
outString += "%s: %s" % (eField, expl[eField])
doneFirst = True
outString += " }"
#print(outString,file=args.outFile)
outputTable += [["COMMENT", outString]]
if args.oldTimestamp != "":
args.outFile.close()
return 0
#print the table in nicely formatted state
if args.outFile == sys.stdout or args.delim == "":
#determine column widths
columnWidths = [0] * len(args.fields)
for i in range(0, len(tableHeaders)):
columnWidths[i] = len(tableHeaders[i])
for r in outputTable:
if len(r) > 0 and r[0] == "COMMENT": continue
for i in range(0, len(r)):
if len(r[i]) > columnWidths[i]: columnWidths[i] = len(r[i])
lineout = ""
for i in range(0, len(tableHeaders)):
lineout += tableHeaders[i].ljust(columnWidths[i]) + " "
print(lineout)
for r in outputTable:
lineout = ""
if len(r) > 0 and r[0] == "COMMENT": lineout = r[1]
else:
for i in range(0, len(r)):
lineout += r[i].ljust(columnWidths[i]) + " "
print(lineout, file=args.outFile)
#print the footer, which is the command to reproduce this output
import os
if args.outFile != sys.stdout:
#remove comment from dataset_values
datasetss = [
x for x in dataset_values.keys() if not x.startswith("comment")
]
print("", file=args.outFile)
print("#lsetup \"asetup %s,%s\" pyAMI" %
(os.environ.get('AtlasProject', 'UNKNOWN!'),
os.environ.get('AtlasVersion', 'UNKNOWN!')),
file=args.outFile)
print(
"#getMetadata.py --timestamp=\"%s\" --physicsGroups=\"%s\" --fields=\"%s\" --inDS=\"%s\""
% (args.timestamp, ",".join(args.physicsGroups), ",".join(
args.fields), ",".join(datasetss)),
file=args.outFile)
logging.info("Results written to: %s" % args.outFile.name)
args.outFile.close()
import os
import pyAMI.atlas.api as ami
import pyAMI.client
from dotenv import load_dotenv
load_dotenv()
CERT_PATH = os.getenv('CERT_FILE')
KEY_PATH = os.getenv('KEY_FILE')
client = pyAMI.client.Client('atlas',
cert_file=CERT_PATH,
key_file=KEY_PATH,
ignore_proxy=True,
verbose=True)
# extract data
# df = uproot.concatenate('../data/wminmunu_MC.root:sumWeights')
# result = client.execute('list datasets --dataset-number 301170 -f cross_section,nfiles,physics_short,events,total_size', format='dict_object')
kwargs = {'dataset-number': 301170}
result = ami.get_dataset_info(client, **kwargs)
print(result)
def main():
# configurable options
config = parseCmdLine(sys.argv[1:])
if (config.baseline or config.official ) and config.sample:
print "--baseline, --official and --sample are mutually exclusive"
sys.exit(1)
# AMI client connection
client = pyAMI.client.Client('atlas')
pyAMI.client.endpoint = config.server
pyAMI.atlas.api.init()
# data type is NTUP_SUSY for 2011/2012 and AOD for 2014 on
datatype = config.datatype
if 'mc11_' in config.prefix or 'mc12_' in config.prefix or 'data11_' in config.prefix or 'data12_' in config.prefix : datatype = '%.merge.NTUP_SUSY%'
# make list of official datasets (baseline+alt)
officialids = []
if config.official or config.baseline or config.sample:
if 'mc12_8TeV' in config.prefix or 'mc14_8TeV' in config.prefix:
import mc12_8TeV_MCSampleList as mcsl
elif 'mc14_13TeV' in config.prefix:
import mc14_13TeV_MCSampleList as mcsl
else:
print '--official is only supported for mc12_8TeV, mc14_8TeV and mc14_13TeV'
sys.exit(1)
if config.sample:
officialids = mcsl.__dict__[str(config.sample)]
else:
officialids = mcsl.__dict__["lbaseline"]
if config.official:
officialids += mcsl.__dict__["lalt"]
# get all datasets matching prefix & tag and then filter them
from pyAMI.atlas.api import get_dataset_info, list_datasets
dskey = config.prefix+datatype+config.tag
print 'Querying AMI for datasets matching pattern',dskey
alldatasets = list_datasets(client,dskey)
acceptedDS = []
for DSlist in alldatasets:
dsname = DSlist['ldn']
cut = False
for filter in filters:
if filter in dsname.split('.')[2]: cut = True
if (config.official or config.baseline or config.sample) and not int(dsname.split('.')[1]) in officialids: cut = True
if config.signal :
cut = True
for pattern in lsignals:
if pattern in dsname: cut = False
if cut: continue
acceptedDS.append(dsname)
pass
acceptedDS.sort()
# get informations for all accepted datasets
dsinfos = []
for dsname in acceptedDS:
dsinfos.append(get_dataset_info(client,dsname)[0])
pass
# write file
fout = open('datasets.txt','w')
for info in dsinfos:
try:
dsname = info['logicalDatasetName']
generatorString = info['generatorName']
version = info['version']
if badDataset(dsname,generatorString,version): continue
availability = info['prodsysStatus']
nFiles = int(info['nFiles'])
if nFiles>0:
period = 'MC'
xsec = 0.
effic = 1.
if info.has_key('period'):
period = info['period']
else:
#(xsec, effic) = get_dataset_xsec_effic(client,info.info['logicalDatasetName'])
# confirmed with AMI team that this should be enought, no need
# to re-implement get_dataset_xsec_effic for PyAMI5
xsec = info[u'crossSection']
effic = info[u'approx_GenFiltEff']
nevts = info['totalEvents']
nfiles = info['nFiles']
if not dsname.endswith('/'): dsname += '/'
fout.write("%s %s %s %s %s %s\n" % (dsname,nevts,nfiles,period,xsec,effic))
except KeyError as prop:
print 'Missing property',prop,'for dataset ',dsname,'in AMI, skip'
fout.close()
pass
