def __init__(self, group, name, *paths, **kwargs):
"""Container class for CMSSW samples, e.g.:
- group: DY (used to group similar samples in final output)
- name: DYJetsToLL_M-50 (used as shorthand and jobname)
- path: /DYJetsToLL_M-50_TuneCP5_13TeV-madgraphMLM-pythia8/RunIIAutumn18NanoAODv6_Nano25Oct2019_102X_mcRun2/NANOAODSIM
- dtype: 'mc', 'data', 'embed'
"""
# PATH
LOG.insist(
len(paths) >= 1, "Need at least one path to create a sample...")
if len(paths) == 1 and isinstance(paths[0], list):
paths = paths[0]
for path in paths:
LOG.insist(
path.count('/') >= 3 and path.startswith('/'),
"DAS path %r has wrong format. Need /SAMPLE/CAMPAIGN/FORMAT." %
(path))
#sample = '/'.join(line.split('/')[-3:])
# DATA TYPE
dtype = kwargs.get('dtype', None)
dtypes = ['mc', 'data', 'embed']
if dtype == None: # automatic recognition
path = paths[0]
if 'Embed' in path:
dtype = 'embed'
elif path.endswith('SIM') or any(g in path
for g in ['pythia', 'madgraph']):
dtype = 'mc'
elif re.search(r"/Run20\d\d", path):
dtype = 'data'
dtype = 'mc' # TODO: remove
LOG.insist(
dtype in dtypes,
"Given data type '%s' is not recongized! Please choose from %s..."
% (dtype, ', '.join(dtypes)))
# ATTRIBUTES
self.group = group
self.name = name
self.paths = paths # DAS dataset path
self.dtype = dtype
self.channels = kwargs.get('channel', None)
self.channels = kwargs.get('channels', self.channels)
self.storage = None
self.storepath = kwargs.get('store',
None) # if stored elsewhere than DAS
self.url = kwargs.get('url', None) # URL if stored elsewhere
self.dasurl = kwargs.get(
'dasurl', None) or "root://cms-xrd-global.cern.ch/" # URL for DAS
self.blacklist = kwargs.get('blacklist', []) # black list file
self.instance = kwargs.get(
'instance', 'prod/phys03' if path.endswith('USER') else
'prod/global') # if None, does not exist in DAS
self.nfilesperjob = kwargs.get('nfilesperjob',
-1) # number of nanoAOD files per job
self.maxevts = kwargs.get(
'maxevtsperjob', -1) # maximum number of events processed per job
self.maxevts = kwargs.get(
'maxevts',
self.maxevts) # maximum number of events processed per job
self.extraopts = kwargs.get(
'opts', []
) # extra options for analysis module, e.g. ['doZpt=1','tes=1.1']
self.subtry = kwargs.get('subtry',
0) # to help keep track of resubmission
self.jobcfg = kwargs.get('jobcfg',
{}) # to help keep track of resubmission
self.nevents = kwargs.get(
'nevts', 0) # number of nanoAOD events that can be processed
self.nevents = kwargs.get('nevents',
self.nevents) # cache of number of events
self.files = kwargs.get(
'files', []) # list of ROOT files, OR text file with list of files
self.filenevts = {} # cache of number of events for each file
self.postfix = kwargs.get(
'postfix',
None) or "" # post-fix (before '.root') for stored ROOT files
self.era = kwargs.get('era', "") # for expansion of $ERA variable
self.dosplit = kwargs.get(
'split',
len(self.paths) >= 2) # allow splitting (if multiple DAS datasets)
self.verbosity = kwargs.get('verbosity',
0) # verbosity level for debugging
self.refreshable = not self.files # allow refresh on file list in getfiles()
# ENSURE LIST
if self.channels != None and not isinstance(self.channels, list):
self.channels = [self.channels]
if isinstance(self.extraopts, str):
if ',' in self.extraopts:
self.extraopts = self.extraopts.split(',')
self.extraopts = [self.extraopts]
# STORAGE & URL DEFAULTS
if self.storepath:
self.storepath = repkey(self.storepath,
USER=_user,
ERA=self.era,
GROUP=self.group,
SAMPLE=self.name)
self.storage = getstorage(repkey(self.storepath,
PATH=self.paths[0],
DAS=self.paths[0]),
ensure=False)
if not self.dasurl:
self.dasurl = self.url if (self.url in dasurls) else dasurls[0]
if not self.url:
if self.storepath:
if self.storage.__class__.__name__ == 'Local':
self.url = "" #root://cms-xrd-global.cern.ch/
else:
self.url = self.storage.fileurl
else:
self.url = self.dasurl
# GET FILE LIST FROM TEXT FILE
if isinstance(self.files, str):
self.loadfiles(self.files)
def loadfiles(self,listname_,**kwargs):
verbosity = LOG.getverbosity(self,kwargs)
"""Load filenames from text file for fast look up in future."""
listname = repkey(listname_,ERA=self.era,GROUP=self.group,SAMPLE=self.name)
LOG.verb("loadfiles: listname=%r -> %r, len(files)=%d, len(filenevts)=%d"%(
listname_,listname,len(self.files),len(self.filenevts)),verbosity,1)
filenevts = self.filenevts
nevents = 0
#listname = ensurefile(listname,fatal=False)
filelist = [ ]
paths = self.paths if '$PATH' in listname else [self.paths[0]]
for path in paths:
listname_ = repkey(listname,PATH=path.strip('/').replace('/','__'))
if self.verbosity>=1:
print ">>> Loading sample files from %r..."%(listname_)
self.pathfiles[path] = [ ]
if os.path.isfile(listname_):
skip = False
subpaths = [ ] # for sanity check
with open(listname_,'r') as file:
for line in file:
line = line.strip().split() # split at space to allow comments at end
if not line: continue
line = line[0].strip() # remove spaces, consider only first part of the line
if line[0]=='#': continue # do not consider comments
#if line.endswith('.root'):
if line.startswith("DASPATH="): # to keep track of multiple DAS data set paths
path = line.split('=')[-1] # DAS data set path
LOG.insist(path.count('/')>=3 and path.startswith('/'),
"DAS path %r in %s has wrong format. Need /SAMPLE/CAMPAIGN/FORMAT..."%(path,listname_))
if path in self.paths: # store file list for this path
self.pathfiles[path] = [ ]
subpaths.append(path)
skip = False
else: # do not store file list for this path
skip = True
else:
if skip: continue # only load files for this sample's DAS dataset paths
match = fevtsexp.match(line) # match $FILENAM(:NEVTS)
if not match: continue
infile = match.group(1)
if match.group(2): # found nevents in filename
nevts = int(match.group(2))
filenevts[infile] = nevts # store/cache in dictionary
nevents += nevts
filelist.append(infile)
self.pathfiles[path].append(infile)
if self.verbosity>=3:
print ">>> %7d events for %s"%(nevts,infile)
if not filelist:
LOG.warning("loadfiles: Did not find any files in %s!"%(listname_))
self.refreshable = True
else: # sanity check for empty list
for subpath in subpaths:
if not self.pathfiles[subpath]:
LOG.warning("loadfiles: Did not find any files for path %s in %s!"%(subpath,listname_))
else:
LOG.warning("loadfiles: file list %s does not exist!"%(listname_))
self.refreshable = True
for path in self.paths:
if path not in self.pathfiles: # nonexistent list
LOG.warning("loadfiles: Did not find any files for path %s in %s!"%(path,listname))
if self.nevents<=0:
self.nevents = nevents
elif self.nevents!=nevents:
LOG.warning("loadfiles: stored nevents=%d does not match the sum total of file events, %d!"%(self.nevents,nevents))
self.nevents == nevents
self.files = filelist
self.files.sort()
return self.files
