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])
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']),
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))
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,
units_per_task=4
),
category=Category(
name='aod',
cores=2,
disk=1000,
"{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])
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,
gridpack_inputs = [
os.path.join(base, madgraph),
os.path.join(base, np_model), '{}/gridpack.py'.format(base), cards
]
workflows = []
for process in processes:
for coefficient_group in itertools.combinations(coefficients, dimension):
tag = '_'.join(coefficient_group)
gridpacks = Workflow(
label='{}_gridpacks_{}'.format(process, tag),
dataset=MultiGridpackDataset(events_per_gridpack=26000,
events_per_lumi=13000),
category=gridpack_resources,
sandbox=cmssw.Sandbox(release=release),
# Use the command and extra_inputs below to constrain coefficient values with an input scan and
# scale value rather than an interval. You can obtain the input scan by running 'cross_sections.py'
# and then (example assuming scan='final_pass.total.npz' and cross_sections_version='ttV/cross_sections/1')
# merge_scans final_pass.total.npz /hadoop/store/user/$USER/ttV/cross_sections/1/final_pass_*/*npz
# mv final_pass.total.npz /hadoop/store/user/$USER/ttV/cross_sections/1/
# command='python gridpack.py {np} {cores} {coefficients} {events} {mg} {model} {pp} {cards} {pcard} --constraints {constraints} --scale {scale} --scan {scan}'.format(
# np=numvalues,
# cores=cores,
# coefficients=','.join(coefficient_group),
# events=events,
# mg=madgraph,
# model='HEL_UFO.third_gen.tar.gz',
# pp=np_param_path,
# cards=os.path.split(cards)[-1],
# pcard='{}.dat'.format(process),
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],
name = gp_info['name']
gp_loc = os.path.join(gp_info['location'],gp_info['tarball'])
print "[{0}/{1}] Gridpack: {gp} (nevts {events})".format(idx+1,len(gridpacks),gp=name,events=nevents)
print "\tGridpack: {path}".format(path=gp_loc)
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']
wf_fragments = {}
for step in wf_steps:
if fragment_map.has_key(p) and fragment_map[p].has_key(step):
wf_fragments[step] = fragment_map[p][step]
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,
len(maod_dirs),
dir=maod_dir)
head, tail = os.path.split(maod_dir)
arr = tail.split('_')
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,
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)
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']),
],
'L1Stub_SingleElectron_D49_SW6p0': [
'rgoldouz/FullProduction/L1tracker_DAS_CMSSW112pre5/L1Stub_SingleElectron_D49_SW6p0',
'100000', 'ele'
]
}
wf = []
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"]
# ),