arr = tail.split('_')
p, c, r = arr[2], arr[3], arr[4]
cms_cmd = ['cmsRun', 'EFTLHEReader_cfg.py']
cms_cmd.extend(['datatier=MINIAODSIM'])
print "\t[{n}/{tot}] mAOD Input: {dir}".format(n=idx + 1,
tot=len(maod_dirs),
dir=maod_dir)
print "\tCommand: {cmd}".format(cmd=' '.join(cms_cmd))
output = Workflow(
label='output_{p}_{c}_{r}'.format(p=p, c=c, r=r),
command=' '.join(cms_cmd),
sandbox=cmssw.Sandbox(
release='../../../../../CMSSW_10_6_8'
), # This file should be in CMSSW_10_6_8/src/EFTGenReader/LHEReader/test/lobster. TODO: Specify path in a better way.
merge_size='1.0G',
cleanup_input=False,
dataset=Dataset(files=maod_dir, files_per_task=5, patterns=["*.root"]),
category=processing)
wf.extend([output])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(
dashboard=False,
bad_exit_codes=[127, 160],
log_level=1,
continue
elif len(runs_whitelist) > 0 and not r in runs_whitelist:
continue
lhe_dirs.append(f)
wf = []
print "Generating workflows:"
for idx, lhe_dir in enumerate(lhe_dirs):
arr = lhe_dir.split('_')
p, c, r = arr[2], arr[3], arr[4]
print "\t[%d/%d] LHE Input: %s" % (idx + 1, len(lhe_dirs), lhe_dir)
output = Workflow(label='output_%s_%s_%s' % (p, c, r),
command='cmsRun EFTLHEReader_cfg.py',
merge_size='1.0G',
cleanup_input=False,
dataset=Dataset(files=lhe_dir,
files_per_task=10,
patterns=["*.root"]),
category=processing)
wf.extend([output])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(
dashboard=False,
bad_exit_codes=[127, 160],
log_level=1,
))
else:
raise ValueError("can't find dataset associated with {}".format(path))
part = counter.get(dset, 1)
counter[dset] = part + 1
aod = Workflow(
label='{}_aod_p{}'.format(dset, part),
pset='configs/' + dset + '_aod.py',
dataset=cmssw.Dataset(
dataset=path,
dbs_instance='phys03',
lumis_per_task=4
),
category=Category(
name='aod',
cores=2,
disk=1000,
memory=3000,
runtime=120 * 60
),
sandbox=[
cmssw.Sandbox(release='/afs/crc.nd.edu/user/m/mwolf3/work/ttH/mcgen/moriond17_part1/CMSSW_8_0_21'),
cmssw.Sandbox(release='/afs/crc.nd.edu/user/m/mwolf3/work/ttH/mcgen/moriond17_part1_rh7/CMSSW_8_0_21')
]
)
maod = Workflow(
label='{}_maod_p{}'.format(dset, part),
pset='configs/' + dset + '_maod.py',
merge_size='2000M',
cleanup_input=True,
'minPtJet=30.0', 'maxEtaJet=2.5', 'maxEtaLep=2.5'
])
if not is_eft:
cms_cmd.extend(['iseft=False'])
print "\tCommand: {cmd}".format(cmd=' '.join(cms_cmd))
# The workflow label can't have any dashes (-) in it, so remove them
safe_label_name = sample_name.replace('-', '')
output = Workflow(
label='output_{label}'.format(label=safe_label_name),
command=' '.join(cms_cmd),
sandbox=cmssw.Sandbox(
release='../../../../../CMSSW_10_6_8/'
), # This file should be in CMSSW_10_6_8/src/EFTGenReader/GenReader/test/lobster. TODO: Specify path in a better way.
cleanup_input=False,
outputs=['output_tree.root'],
merge_size=
merge_size, # Note: Lobster takes a very long time trying to merge large numbers of small files for some reason
dataset=ds,
merge_command='hadd @outputfiles @inputfiles',
category=processing)
wf.extend([output])
config = Config(
label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
name='processing',
cores=1,
runtime=900,
memory=1000
)
workflows = []
ttH = Workflow(
label='ttH',
dataset=cmssw.Dataset(
dataset='/ttHToNonbb_M125_13TeV_powheg_pythia8/RunIIFall15MiniAODv2-PU25nsData2015v1_76X_mcRun2_asymptotic_v12-v1/MINIAODSIM',
lumis_per_task=20,
file_based=True
),
category=processing,
command='root -b -q -l script_macro.C @outputfiles @inputfiles',
extra_inputs=['script_macro.C'],
publish_label='test',
merge_command='hadd @outputfiles @inputfiles',
merge_size='3.5G',
outputs=['output.root']
)
workflows.append(ttH)
config = Config(
workdir='/tmpscratch/users/$USER/lobster_test_' + version,
plotdir='~/www/lobster/test_' + version,
storage=storage,
workflows=workflows,
for key, value in samples.items():
if 'DisplacedMu' not in key:
continue
print key
Analysis = Workflow(
label='FE_L1Analysis_%s' % (key),
sandbox=cmssw.Sandbox(
release='/afs/crc.nd.edu/user/r/rgoldouz/CMSSW_10_4_0'),
dataset=Dataset(files=value[0], files_per_task=50),
globaltag=False,
command='python Lobster_check.py ' + value[1] + ' ' + value[2] +
' @inputfiles',
extra_inputs=[
'Lobster_check.py',
'../lib/main.so',
'../include/MyAnalysis.h',
],
outputs=['ANoutput.root'],
# dataset=Dataset(
# files=value[0],
# files_per_task=50,
# patterns=["*.root"]
# ),
# merge_command='hadd @outputfiles @inputfiles',
# merge_size='3.5G',
category=gs_resources)
wf.append(Analysis)
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
processing = Category(
name='processing',
cores=1,
runtime=900,
memory=1000
)
workflows = []
ttH = Workflow(
label='ttH',
dataset=cmssw.Dataset(
dataset='/ttHToNonbb_M125_13TeV_powheg_pythia8/RunIIFall15MiniAODv2-PU25nsData2015v1_76X_mcRun2_asymptotic_v12-v1/MINIAODSIM',
events_per_task=50000
),
category=processing,
command='cmsRun simple_pset.py',
publish_label='test',
merge_size='3.5G',
outputs=['output.root']
)
workflows.append(ttH)
config = Config(
workdir='/tmpscratch/users/$USER/lobster_test_' + version,
plotdir='~/www/lobster/test_' + version,
storage=storage,
workflows=workflows,
advanced=AdvancedOptions(
bad_exit_codes=[127, 160],
shutil.copy(template_loc, mod_loc)
for sed_str in sed_str_list:
if sed_str:
run_process(['sed', '-i', '-e', sed_str, mod_loc])
else:
mod_loc = template_loc
wf_fragments[step] = mod_loc
if mod_tag == 'base': mod_tag = ''
label_tag = "{p}_{c}{mod}_{r}".format(p=p, c=c, r=r, mod=mod_tag)
gen = Workflow(
label='gen_step_{tag}'.format(tag=label_tag),
command='cmsRun {cfg}'.format(cfg=wf_fragments['gen']),
sandbox=cmssw.Sandbox(release=rel_map[UL_YEAR]['gen']),
merge_size=-1, # Don't merge files we don't plan to keep
cleanup_input=False, # Do not accidently clean up the LHE files!!!
globaltag=False,
outputs=['GEN-00000.root'],
dataset=Dataset(files=lhe_dir,
files_per_task=1,
patterns=["*.root"]),
category=gen_resources)
sim = Workflow(
label='sim_step_{tag}'.format(tag=label_tag),
command='cmsRun {cfg}'.format(cfg=wf_fragments['sim']),
sandbox=cmssw.Sandbox(release=rel_map[UL_YEAR]['sim']),
merge_size=-1, # Don't merge files we don't plan to keep
cleanup_input=True,
#cleanup_input=False,
globaltag=False,
outputs=['SIM-00000.root'],
wf = []
print "Generating workflows:"
for idx,gen_dir in enumerate(gen_dirs):
#arr = gen_dir.split('_')
head,tail = os.path.split(gen_dir)
arr = tail.split('_')
p,c,r = arr[2],arr[3],arr[4]
print "\t[{n}/{tot}] GEN Input: {dir}".format(n=idx+1,tot=len(gen_dirs),dir=gen_dir)
output = Workflow(
label='output_{p}_{c}_{r}'.format(p=p,c=c,r=r),
command='cmsRun EFTLHEReader_cfg.py',
sandbox=cmssw.Sandbox(release='../../../../../CMSSW_10_6_8'), # This file should be in CMSSW_10_6_8/src/EFTGenReader/LHEReader/test/lobster. TODO: Specify path in a better way.
merge_size='1.0G',
cleanup_input=False,
dataset=Dataset(
files=gen_dir,
files_per_task=5, # Remember that the GEN step already does 5-10 files per task
patterns=["*.root"]
),
category=processing
)
wf.extend([output])
config = Config(
label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(
'minPtJet=30.0',
'maxEtaJet=2.5',
'maxEtaLep=2.5'
])
if not is_eft:
cms_cmd.extend(['iseft=False'])
print "\tCommand: {cmd}".format(cmd=' '.join(cms_cmd))
# The workflow label can't have any dashes (-) in it, so remove them
safe_label_name = sample_name.replace('-','')
output = Workflow(
label='output_{label}'.format(label=safe_label_name),
command=' '.join(cms_cmd),
cleanup_input=False,
outputs=['output_tree.root'],
merge_size=merge_size, # Note: Lobster takes a very long time trying to merge large numbers of small files for some reason
dataset=ds,
merge_command='hadd @outputfiles @inputfiles',
category=processing
)
wf.extend([output])
config = Config(
label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(
dashboard=False,
p, c, r = arr[2], arr[3], arr[4]
#print("p c r:",p,c,r)
wf_fragments = {}
for step in wf_steps:
template_loc = fragment_map["all_procs"][step]
wf_fragments[step] = template_loc
label_tag = "{p}_{c}_{r}".format(p=p, c=c, r=r)
naod = Workflow(
label='nAOD_step_{tag}'.format(tag=label_tag),
command='cmsRun {cfg}'.format(cfg=wf_fragments['naod']),
sandbox=cmssw.Sandbox(release=PATH_TO_NAOD_CMSSW),
#merge_size='256M',
merge_size='1000M',
merge_command='python haddnano.py @outputfiles @inputfiles',
extra_inputs=[
os.path.join(PATH_TO_NAOD_CMSSW,
'src/PhysicsTools/NanoAODTools/scripts/haddnano.py')
],
cleanup_input=False, # Leave the MAOD files
outputs=['NAOD-00000.root'],
dataset=Dataset(files=maod_dir, files_per_task=1, patterns=["*.root"]),
category=naod_resources)
wf.extend([naod])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
else:
mod_loc = template_loc
wf_fragments[step] = mod_loc
if mod_tag == 'base': mod_tag = ''
label_tag = "{p}_{c}{mod}_{r}".format(p=p, c=c, r=r, mod=mod_tag)
print "\t\tLabel: {label}".format(label=label_tag)
print "\nThis is the wf_fragments:", wf_fragments, "\n"
gen = Workflow(
label='gen_step_{tag}'.format(tag=label_tag),
command='cmsRun {cfg}'.format(cfg=wf_fragments['gen']),
sandbox=cmssw.Sandbox(release='CMSSW_9_3_6'),
merge_size=-1, # Don't merge files we don't plan to keep
cleanup_input=False,
globaltag=False,
outputs=['GEN-00000.root'],
dataset=Dataset(
files=lhe_dir,
#files_per_task=2,
files_per_task=1,
patterns=["*.root"]),
category=gen_resources)
sim = Workflow(
label='sim_step_{tag}'.format(tag=label_tag),
command='cmsRun {cfg}'.format(cfg=wf_fragments['sim']),
sandbox=cmssw.Sandbox(release=rel_map[UL_YEAR]['sim']),
merge_size=-1, # Don't merge files we don't plan to keep
cleanup_input=False,
globaltag=False,
outputs=['SIM-00000.root'],
tail = tail.replace("cfg.py", "{tag}_cfg.py".format(tag=cfg_tag))
mod_loc = os.path.join(MODIFIED_CFG_DIR, tail)
shutil.copy(template_loc, mod_loc)
for sed_str in sed_str_list:
if sed_str:
run_process(['sed', '-i', '-e', sed_str, mod_loc])
wf_fragments[step] = mod_loc
if mod_tag == 'base': mod_tag = ''
gen = Workflow(
label='gen_step_{p}_{c}{mod}_{r}'.format(p=p,
c=c,
mod=mod_tag,
r=r),
command='cmsRun {cfg}'.format(cfg=wf_fragments['gen']),
#sandbox=cmssw.Sandbox(release='CMSSW_9_3_1'),
sandbox=cmssw.Sandbox(release='CMSSW_9_3_6'),
merge_size=-1, # Don't merge files we don't plan to keep
cleanup_input=False,
globaltag=False,
outputs=['GEN-00000.root'],
dataset=Dataset(files=lhe_dir,
files_per_task=5,
patterns=["*.root"]),
category=gen_resources)
wf.extend([gen])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(dashboard=False,
head, tail = os.path.split(maod_dir)
arr = tail.split('_')
p, c, r = arr[2], arr[3], arr[4]
cms_cmd = ['cmsRun', 'EFTLHEReader_cfg.py']
cms_cmd.extend(['datatier=MINIAODSIM'])
print "\t[{n}/{tot}] mAOD Input: {dir}".format(n=idx + 1,
tot=len(maod_dirs),
dir=maod_dir)
print "\tCommand: {cmd}".format(cmd=' '.join(cms_cmd))
output = Workflow(label='output_{p}_{c}_{r}'.format(p=p, c=c, r=r),
command=' '.join(cms_cmd),
merge_size='1.0G',
cleanup_input=False,
dataset=Dataset(files=maod_dir,
files_per_task=5,
patterns=["*.root"]),
category=processing)
wf.extend([output])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(
dashboard=False,
bad_exit_codes=[127, 160],
log_level=1,
))
else:
wf_fragments[step] = fragment_map['default'][step]
multiplier = event_multiplier['default']
if event_multiplier.has_key(p):
multiplier = event_multiplier[p]
nevents = int(multiplier * events_per_gridpack)
print "\t[{0}/{1}] Gridpack: {gp} (nevts {events})".format(
idx + 1, len(gridpack_list), gp=gridpack, events=nevents)
lhe = Workflow(
label=label,
command='cmsRun {cfg}'.format(cfg=wf_fragments['lhe']),
sandbox=cmssw.Sandbox(release='CMSSW_9_3_1'),
merge_size=
-1, # Don't merge the output files, to keep individuals as small as possible
cleanup_input=False,
globaltag=False,
outputs=['HIG-RunIIFall17wmLHE-00000ND.root'],
dataset=MultiProductionDataset(gridpacks=gridpack,
events_per_gridpack=nevents,
events_per_lumi=events_per_lumi,
lumis_per_task=1,
randomize_seeds=True),
category=cat)
wf.extend([lhe])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(dashboard=False,
workflows = []
for dset, tasksize, events in zip(datasets, tasksizes, events):
tasks = int(events / tasksize)
lhe = Workflow(
label=dset + '_lhe',
pset='configs/' + dset + '_lhe.py',
merge_size='2000M',
dataset=ProductionDataset(
events_per_task=tasksize,
events_per_lumi=200,
number_of_tasks=tasks
),
category=Category(
name='lhe',
cores=2,
disk=2000,
memory=2000
),
sandbox=[
cmssw.Sandbox(release='/afs/crc.nd.edu/user/m/mwolf3/work/ttH/mcgen/moriond17_part1/CMSSW_8_0_21'),
cmssw.Sandbox(release='/afs/crc.nd.edu/user/m/mwolf3/work/ttH/mcgen/moriond17_part1_rh7/CMSSW_8_0_21')
]
)
aod = Workflow(
label=dset + '_aod',
pset='configs/' + dset + '_aod.py',
dataset=ParentDataset(
parent=lhe,
cmd = ['cmsRun']
cmd.append(gp_info['lhe_cfg'])
label = 'lhe_step_{tag}'.format(tag=name)
print "\tLHE Step: {label}".format(label=label)
print "\tLHE cfg: {cfg}".format(cfg=gp_info['lhe_cfg'])
lhe = Workflow(
label=label,
command=' '.join(cmd),
sandbox=cmssw.Sandbox(release=gp_info['lhe_release']),
merge_size=-1,
cleanup_input=False,
globaltag=False,
outputs=['LHE-00000.root'],
dataset=MultiProductionDataset(
gridpacks=gp_loc,
events_per_gridpack=nevents,
events_per_lumi=events_per_lumi,
lumis_per_task=1,
randomize_seeds=True
),
category=lhe_resources
)
cmd = ['cmsRun']
cmd.append(gp_info['gen_cfg'])
label = 'gen_step_{tag}'.format(tag=name)
print "\tGEN Step: {label}".format(label=label)
print "\tGEN cfg: {cfg}".format(cfg=gp_info['gen_cfg'])
if sed_str:
run_process(['sed','-i','-e',sed_str,mod_loc])
else:
mod_loc = template_loc
wf_fragments[step] = mod_loc
if mod_tag == 'base': mod_tag = ''
label_tag = "{p}_{c}{mod}_{r}".format(p=p,c=c,r=r,mod=mod_tag)
print "\t\tLabel: {label}".format(label=label_tag)
gs = Workflow(
label='gs_step_{tag}'.format(tag=label_tag),
command='cmsRun {cfg}'.format(cfg=wf_fragments['gs']),
sandbox=cmssw.Sandbox(release='CMSSW_9_3_6'),
merge_size=-1, # Don't merge files we don't plan to keep
cleanup_input=False,
globaltag=False,
outputs=['HIG-RunIIFall17wmLHEGS-00040ND.root'],
dataset=Dataset(
files=lhe_dir,
files_per_task=1,
patterns=["*.root"]
),
category=gs_resources
)
digi = Workflow(
label='digi_step_{tag}'.format(tag=label_tag),
command='cmsRun {cfg}'.format(cfg=wf_fragments['digi']),
sandbox=cmssw.Sandbox(release='CMSSW_9_4_0_patch1'),
merge_size=-1, # Don't merge files we don't plan to keep
cleanup_input=True, # Save the GEN-SIM step
outputs=['HIG-RunIIFall17DRPremix-00823ND_step1.root'],
print "Generating workflows:"
for idx, gen_dir in enumerate(gen_dirs):
#arr = gen_dir.split('_')
head, tail = os.path.split(gen_dir)
arr = tail.split('_')
p, c, r = arr[2], arr[3], arr[4]
print "\t[{n}/{tot}] GEN Input: {dir}".format(n=idx + 1,
tot=len(gen_dirs),
dir=gen_dir)
output = Workflow(
label='output_{p}_{c}_{r}'.format(p=p, c=c, r=r),
command='cmsRun EFTLHEReader_cfg.py',
merge_size='1.0G',
cleanup_input=False,
dataset=Dataset(
files=gen_dir,
files_per_task=
5, # Remember that the GEN step already does 5-10 files per task
patterns=["*.root"]),
category=processing)
wf.extend([output])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(
dashboard=False,
bad_exit_codes=[127, 160],
version = "v5"
storage = StorageConfiguration(output=[
"hdfs://eddie.crc.nd.edu:19000/store/user/khurtado/lobster_mc_" + version,
# "file:///hadoop/store/user/khurtado/lobster_mc_" + version,
# "root://deepthought.crc.nd.edu//store/user/khurtado/lobster_mc_" + version,
# "chirp://eddie.crc.nd.edu:9094/store/user/khurtado/lobster_test_" + version,
"gsiftp://T3_US_NotreDame/store/user/khurtado/lobster_mc_" + version,
# "srm://T3_US_NotreDame/store/user/khurtado/lobster_mc_" + version,
])
workflows = []
lhe = Workflow(label='lhe_step',
pset='mc_gen/HIG-RunIIWinter15wmLHE-00196_1_cfg.py',
sandbox=cmssw.Sandbox(release='mc_gen/CMSSW_7_1_16_patch1'),
merge_size='10M',
dataset=ProductionDataset(events_per_task=50,
events_per_lumi=5,
number_of_tasks=10),
category=Category(name='lhe', cores=1, memory=1000))
gs = Workflow(label='gs_step',
pset='mc_gen/HIG-RunIISummer15GS-00177_1_cfg.py',
sandbox=cmssw.Sandbox(release='mc_gen/CMSSW_7_1_18'),
merge_size='100M',
dataset=ParentDataset(parent=lhe, units_per_task=1),
category=Category(name='gs',
cores=1,
memory=2000,
runtime=45 * 60))
digi = Workflow(label='digi_step',
continue
sample_loc = ds_helper.getData(sample_name, 'loc')
ds = cmssw.Dataset(dataset=sample_loc, events_per_task=30000)
cms_cmd = ['cmsRun', 'EFTLHEReader_cfg.py']
cms_cmd.extend(['datatier=MINIAODSIM'])
print "\t[{n}/{tot}] mAOD Input: {dir}".format(n=idx + 1,
tot=len(samples),
dir=sample_name)
print "\tCommand: {cmd}".format(cmd=' '.join(cms_cmd))
safe_label_name = sample_name.replace('-', '')
output = Workflow(label='output_{label}'.format(label=safe_label_name),
command=' '.join(cms_cmd),
cleanup_input=False,
merge_size=-1,
dataset=ds,
category=processing)
wf.extend([output])
config = Config(label=master_label,
workdir=workdir_path,
plotdir=plotdir_path,
storage=storage,
workflows=wf,
advanced=AdvancedOptions(dashboard=False,
bad_exit_codes=[127, 160],
log_level=1,
xrootd_servers=[
'ndcms.crc.nd.edu',