def deepCheckWeight(file):
""" some root files only contain the branches kept from the beginning
but not those from the filler, e.g. the weight branch
Those files are identified here, as weight==nan and thus the yield is nan
"""
from math import isnan
from RootTools.core.Sample import Sample
# convert dpm file pathes
sample = Sample.fromFiles(name="sample", treeName="Events", files=file)
# check for branch:
l = sample.chain.GetListOfBranches()
if not 'weight' in [l.At(i).GetName() for i in range(l.GetSize())]:
return 0
val = sample.getYieldFromDraw(weightString="weight")['val']
del sample
#logger.debug("Val in deepCheckWeight: %r", val)
return not isnan(val)
def __init__(self,
heppy_samples,
dpm_directories,
cache_file,
multithreading=True):
# Read cache file, if exists
if os.path.exists(cache_file) and not overwrite:
self.sample_map = pickle.load(file(cache_file))
logger.info("Loaded cache file %s" % cache_file)
else:
logger.info("Cache file %s not found. Recreate map.", cache_file)
logger.info("Check proxy.")
# Proxy certificate
from RootTools.core.helpers import renew_proxy
# Make proxy in afs to allow batch jobs to run
proxy_path = os.path.expandvars('$HOME/private/.proxy')
proxy = renew_proxy(proxy_path)
logger.info("Using proxy %s" % proxy)
# Read dpm directories
self.cmg_directories = {}
for data_path in dpm_directories:
logger.info("Walking dpm directory %s", data_path)
walker = walk_dpm(data_path)
self.cmg_directories[
data_path] = walker.walk_dpm_cmgdirectories('.', maxN=maxN)
#del walker
logger.info("Now mapping directories to heppy samples")
for heppy_sample in heppy_samples:
heppy_sample.candidate_directories = []
pd, era = heppy_sample.dataset.split('/')[1:3]
for data_path in self.cmg_directories.keys():
for dpm_directory in self.cmg_directories[data_path].keys(
):
if not ('/%s/' % pd in dpm_directory):
logger.debug("/%s/ not in dpm_directory %s", pd,
dpm_directory)
continue
if not ('/' + era in dpm_directory):
logger.debug("/%s not in dpm_directory %s", era,
dpm_directory)
continue
heppy_sample.candidate_directories.append(
[data_path, dpm_directory])
logger.debug("heppy sample %s in %s",
heppy_sample.name, dpm_directory)
logger.info("Found heppy sample %s in %i directories.",
heppy_sample.name,
len(heppy_sample.candidate_directories))
# Merge
from RootTools.core.Sample import Sample
logger.info(
"Now making new sample map from %i directories and for %i heppy samples to be stored in %s",
len(dpm_directories), len(heppy_samples), cache_file)
self.sample_map = {}
for heppy_sample in heppy_samples:
if len(heppy_sample.candidate_directories) == 0:
logger.info("No directory found for %s", heppy_sample.name)
else:
normalization, files = walker.combine_cmg_directories(\
cmg_directories = {dpm_directory:self.cmg_directories[data_path][dpm_directory] for data_path, dpm_directory in heppy_sample.candidate_directories },
multithreading = multithreading,
)
logger.info(
"Sample %s: Found a total of %i files with normalization %3.2f",
heppy_sample.name, len(files), normalization)
tmp_files = []
for f in files:
isGoodFile = False
try:
isGoodFile = checkRootFile(
"root://hephyse.oeaw.ac.at/" + os.path.join(f))
logger.debug("File %s got added", f)
except IOError:
logger.info("File %s is corrupted, skipping", f)
if isGoodFile: tmp_files.append(f)
self.sample_map[heppy_sample] = Sample.fromFiles(
heppy_sample.name,
files=[
'root://hephyse.oeaw.ac.at/' + f for f in tmp_files
],
normalization=normalization,
treeName='tree',
isData=heppy_sample.isData,
maxN=maxN)
logger.info(
"Combined %i directories for sample %s to a total of %i files with normalization %3.2f",
len(heppy_sample.candidate_directories),
heppy_sample.name, len(files), normalization)
# Store cache file
dir_name = os.path.dirname(cache_file)
if len(self.sample_map.keys()) > 0:
if not os.path.exists(dir_name): os.makedirs(dir_name)
pickle.dump(self.sample_map, file(cache_file, 'w'))
logger.info("Created MC sample cache %s", cache_file)
else:
logger.info("Skipping to write %s because map is empty.",
cache_file)