def submit_jobs(output_directory, tag, samples, events_total,
events_per_output, tarfile):
# Long loop to monitor jobs and resubmit failed jobs
for i in range(10000):
total_summary = {}
# Loop through samples
for sample in samples:
config_info = samples_config[sample]
executable = create_executable(tag, sample, config_info)
task = CondorTask(sample=DummySample(
N=int(float(events_total) / float(events_per_output)),
nevents=events_total,
dataset="/" + sample + "_NANO"),
tag=tag,
special_dir=output_directory,
events_per_output=events_per_output,
total_nevents=events_total,
split_within_files=True,
executable=executable,
open_dataset=False,
tarfile=tarfile,
condor_submit_params=CONDOR_SUBMIT_PARAMS)
task.process()
total_summary[
task.get_sample().get_datasetname()] = task.get_task_summary()
StatsParser(data=total_summary,
webdir="~/public_html/dump/Hgg-MC-Production/").do()
time.sleep(300) # power nap
def submit(which):
total_summary = {}
extra_requirements = "true"
tag = "v0_noFilter"
pdname = "HVNoFilter"
events_per_point = int(1E4)
events_per_job = int(100)
cfgsDir = "psets_gensim_noFilter/"
modelsFile = cfgsDir + "/models.txt"
df = pd.read_csv(modelsFile)
for year in ["2016", "2017", "2018"]:
for iterr, row in df.iterrows():
# fmass = float(mass)
# mass = str(mass).replace(".","p")
epp = int(events_per_point)
reqname = "noFilter_m{}{}_ctau{}_xi_{}_{}_{}".format(
row.portal, row.mass, row.ctau, row.xi, tag, year)
njobs = epp // events_per_job
sample = DummySample(
dataset="/{}/params_{}_m_{}_ctau_{}mm_xi_{}_{}/LLPNTUPLE".
format(pdname, row.portal, row.mass, row.ctau * 10, row.xi,
year),
N=njobs,
nevents=epp)
task = CondorTask(
sample=sample,
output_name="output.root",
executable="executables/condor_executable_{}.sh".format(which),
tarfile="package_{}.tar.xz".format(year),
open_dataset=True,
files_per_output=1,
condor_submit_params={
"classads": [
["param_mass", row.mass],
["param_ctau", row.ctau],
["param_xi", str(row.xi).replace(".", "p")],
["param_portal", row.portal],
["param_year", year],
["param_nevents", events_per_job],
["metis_extraargs", ""],
["JobBatchName", reqname],
],
"requirements_line":
'Requirements = ((HAS_SINGULARITY=?=True) && (HAS_CVMFS_cms_cern_ch =?= true) && {extra_requirements})'
.format(extra_requirements=extra_requirements),
},
tag=tag,
recopy_inputs=True)
task.process()
total_summary[
task.get_sample().get_datasetname()] = task.get_task_summary()
StatsParser(data=total_summary,
webdir="~/public_html/dump/metis_test/").do()
def submit():
requests = {
#'TTWJetsToLNuEWK_5f_NLO': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_NLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball_retired.tar.xz', # that's the SM point, but using the SMEFT model. No lepton filtering, so name is actually confusing
#'TTWJetsToLNuEWK_5f_NLO_v2': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_NLO_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # that's the actual SM
#'TTWplusJetsToLNuEWK_5f_NLO_v2': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWplusJetsToLNuEWK_5f_NLO_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # that's the actual SM
#'TTWminusJetsToLNuEWK_5f_NLO_v2': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWminusJetsToLNuEWK_5f_NLO_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # that's the actual SM
#'TTWJetsToLNuEWK_5f_EFT_myNLO_full': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_cpt8_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz', # one of the BSM points
#'TTWJetsToLNuEWK_5f_EFT_mix_myNLO_full': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz', # EFT mix
'TTWJetsToLNuEWK_5f_EFT_cpq3_4_myNLO_full': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks//TTWJetsToLNuEWK_5f_EFT_myNLO_cpq3_4_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # C_pq3 = 4
}
total_summary = {}
extra_requirements = "true"
tag = "v4"
events_per_point = 500000 # produced 500k events before
events_per_job = 2000 # up to 2000 works
#events_per_point = 50
#events_per_job = 10
njobs = int(events_per_point)//events_per_job
for reqname in requests:
gridpack = requests[reqname]
#reqname = "TTWJetsToLNuEWK_5f_EFT_myNLO"
#gridpack = '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz'
task = CondorTask(
sample = DummySample(dataset="/%s/RunIIAutumn18/NANO"%reqname,N=njobs,nevents=int(events_per_point)),
output_name = "nanoAOD.root",
executable = "executables/condor_executable_Autumn18.sh",
tarfile = "package.tar.gz",
#scram_arch = "slc7_amd64_gcc630",
open_dataset = False,
files_per_output = 1,
arguments = gridpack,
condor_submit_params = {
"sites":"T2_US_UCSD", #
"classads": [
["param_nevents",events_per_job],
["metis_extraargs",""],
["JobBatchName",reqname],
#["SingularityImage", "/cvmfs/singularity.opensciencegrid.org/cmssw/cms:rhel6-m202006"],
],
"requirements_line": 'Requirements = (HAS_SINGULARITY=?=True)' # && (HAS_CVMFS_cms_cern_ch =?= true) && {extra_requirements})'.format(extra_requirements=extra_requirements),
},
tag = tag,
min_completion_fraction = 0.90,
)
task.process()
total_summary[task.get_sample().get_datasetname()] = task.get_task_summary()
StatsParser(data=total_summary, webdir="~/public_html/dump/tW_gen/").do()
def submit():
requests = {
'TTWJetsToLNuEWK_5f_EFT_myNLO': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz', # that's EFT
'TTWJetsToLNuEWK_5f_EFT_myNLO_cpt8': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_cpt8_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz', # that's EFT
'TTWJetsToLNuEWK_5f_NLO': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_NLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz', # that's the SM
}
total_summary = {}
extra_requirements = "true"
tag = "v1"
events_per_point = 500000
events_per_job = 10000
#events_per_point = 500
#events_per_job = 100
njobs = int(events_per_point)//events_per_job
for reqname in requests:
gridpack = requests[reqname]
#reqname = "TTWJetsToLNuEWK_5f_EFT_myNLO"
#gridpack = '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz'
task = CondorTask(
sample = DummySample(dataset="/%s/RunIIAutumn18/NANOGEN"%reqname,N=njobs,nevents=int(events_per_point)),
output_name = "output.root",
executable = "executables/condor_executable.sh",
tarfile = "package.tar.gz",
open_dataset = False,
files_per_output = 1,
arguments = gridpack,
condor_submit_params = {
"sites":"T2_US_UCSD", #
"classads": [
["param_nevents",events_per_job],
["metis_extraargs",""],
["JobBatchName",reqname],
],
"requirements_line": 'Requirements = ((HAS_SINGULARITY=?=True) && (HAS_CVMFS_cms_cern_ch =?= true) && {extra_requirements})'.format(extra_requirements=extra_requirements),
},
tag = tag,
)
task.process()
total_summary[task.get_sample().get_datasetname()] = task.get_task_summary()
StatsParser(data=total_summary, webdir="~/public_html/dump/tW_gen/").do()
def main():
main_dir = os.path.dirname(os.path.abspath(__file__))
metis_path = os.path.dirname(os.path.dirname(metis.__file__))
exec_path = main_dir + "/metis.sh"
hadoop_path = "metis/"
metis_dashboard_path = os.path.join(metis_path, "dashboard")
job_tag = ""
total_summary = {}
os.chdir(metis_path)
while True:
masspoints = [125]
tasks = []
for mass in masspoints:
miniaod = CondorTask(
## Dummy sample as no input is needed in generating the events
#sample = DummySample(
# N=3000,
# dataset="/VHToNonbb_M125_13TeV_amcatnloFXFX_madspin_pythia8/PRIVATE-RunIISummer16MiniAODv2-PUMoriond17_80X_mcRun2_asymptotic_2016_TrancheIV_v6_ext1-v1/MINIAODSIM"
# ),
# Dummy sample as no input is needed in generating the events
sample = DummySample(
N=3500,
dataset="/WWW_4F_TuneCP5_13TeV-amcatnlo-pythia8/PRIVATE-RunIIFall17MiniAODv2-PU2017_12Apr2018_94X_mc2017_realistic_v14-v1/MINIAODSIM"
),
tag = job_tag,
executable = exec_path,
special_dir = hadoop_path + "/private_miniaod",
output_name = "output.root",
files_per_output = 1,
condor_submit_params = {"sites" : "T2_US_UCSD"},
open_dataset = True,
flush = True,
#no_load_from_backup = True,
)
tasks.extend([miniaod])
all_tasks_complete = True
for task in tasks:
task.process()
summary = task.get_task_summary()
total_summary[task.get_sample().get_datasetname()] = summary
all_tasks_complete = all_tasks_complete and task.complete()
# parse the total summary and write out the dashboard
StatsParser(data=total_summary, webdir=metis_dashboard_path).do()
# Print msummary table so I don't have to load up website
os.system("msummary -v0 -p ext2 | tee summary.txt")
os.system("chmod -R 755 {}".format(metis_dashboard_path))
# If all done exit the loop
if all_tasks_complete:
print ""
print "Job={} finished".format(job_tag)
print ""
break
# Neat trick to not exit the script for force updating
print 'Press Ctrl-C to force update, otherwise will sleep for 300 seconds'
try:
for i in range(0,300):
sleep(1) # could use a backward counter to be preeety :)
except KeyboardInterrupt:
raw_input("Press Enter to force update, or Ctrl-C to quit.")
print "Force updating..."
from metis.CMSSWTask import CMSSWTask
from metis.Sample import DirectorySample, DummySample
from metis.Path import Path
from metis.StatsParser import StatsParser
import time
for _ in range(1000):
proc_tag = "v1"
#special_dir = "workflowtest/ProjectMetis"
special_dir = "miniaod_runII"
step1 = CMSSWTask(
# Change dataset to something more meaningful (but keep STEP1, as we use this
# for string replacement later); keep N=1
sample=DummySample(N=1,
dataset="/JHUSample_tHq_cpodd_2017_20200520/STEP1"),
# A unique identifier
tag=proc_tag,
special_dir=special_dir,
# Probably want to beef up the below two numbers to control splitting,
# but note that step2 is the bottleneck, so don't put too many events
# in one output file here
events_per_output=200,
total_nevents=500000,
# We have one input dummy file, so this must be True
split_within_files=True,
pset="psets/HIG-RunIIFall17wmLHEGS-01343_1_cfg_jhu.py",
#cmssw_version = "CMSSW_9_3_6_patch2",
cmssw_version="CMSSW_9_4_6_patch1",
condor_submit_params={"sites": "T2_US_UCSD"},
scram_arch="slc6_amd64_gcc630",
def runall(special_dir, tag, total_nevents, events_per_output, config):
for _ in range(25):
proc_tag = "v1"
#special_dir = "workflowtest/ProjectMetis"
#special_dir = "miniaod_runII/JHUSample_ttH"
cmssw_v_gensim = config["cmssw_v_gensim"]
pset_gensim = config["pset_gensim"]
scram_arch_gensim = config["scram_arch_gensim"]
cmssw_v_aodsim = config["cmssw_v_aodsim"]
pset_aodsim = config["pset_aodsim"]
scram_arch_aodsim = config["scram_arch_aodsim"]
pset_aodsim2 = config["pset_aodsim2"]
cmssw_v_aodsim2 = cmssw_v_aodsim #config["pset_aodsim2"]
scram_arch_aodsim2 = scram_arch_aodsim #config["scram_arch_aodsim2"]
cmssw_v_miniaodsim = config["cmssw_v_miniaodsim"]
pset_miniaodsim = config["pset_miniaodsim"]
scram_arch_miniaodsim = config["scram_arch_miniaodsim"]
step1 = CMSSWTask(
# Change dataset to something more meaningful (but keep STEP1, as we use this
# for string replacement later); keep N=1
sample=DummySample(N=1, dataset="/" + tag + "_STEP1"),
# A unique identifier
tag=proc_tag,
special_dir=special_dir,
# Probably want to beef up the below two numbers to control splitting,
# but note that step2 is the bottleneck, so don't put too many events
# in one output file here
events_per_output=events_per_output,
total_nevents=total_nevents,
#events_per_output = 50,
#total_nevents = 1000,
# We have one input dummy file, so this must be True
split_within_files=True,
pset="psets/" + pset_gensim,
cmssw_version=cmssw_v_gensim,
scram_arch=scram_arch_gensim,
)
step2 = CMSSWTask(
sample=DirectorySample(
location=step1.get_outputdir(),
dataset=step1.get_sample().get_datasetname().replace(
"STEP1", "STEP2"),
),
tag=proc_tag,
special_dir=special_dir,
open_dataset=True,
files_per_output=1,
pset="psets/" + pset_aodsim,
cmssw_version=cmssw_v_aodsim,
scram_arch=scram_arch_aodsim,
)
step3 = CMSSWTask(
sample=DirectorySample(
location=step2.get_outputdir(),
dataset=step2.get_sample().get_datasetname().replace(
"STEP2", "STEP3"),
),
tag=proc_tag,
special_dir=special_dir,
open_dataset=True,
files_per_output=1,
pset="psets/" + pset_aodsim2,
cmssw_version=cmssw_v_aodsim2,
scram_arch=scram_arch_aodsim2,
)
step4 = CMSSWTask(
sample=DirectorySample(
location=step3.get_outputdir(),
dataset=step3.get_sample().get_datasetname().replace(
"STEP3", "STEP4"),
),
tag=proc_tag,
special_dir=special_dir,
open_dataset=True,
files_per_output=1,
output_name="step4.root",
pset="psets/" + pset_miniaodsim,
cmssw_version=cmssw_v_miniaodsim,
scram_arch=scram_arch_miniaodsim,
# condor_submit_params = {"sites":"UAF,UCSD"},
)
'''
step5 = CMSSWTask(
sample = DirectorySample(
location = step4.get_outputdir(),
dataset = step4.get_sample().get_datasetname().replace("STEP4","STEP5"),
),
tag = proc_tag,
special_dir = special_dir,
open_dataset = True,
files_per_output = 1,
pset = "psets/TOP-RunIIFall17NanoAODv7-00001_1_cfg.py",
# The below two lines should match output file names in the pset
output_name = "step5.root",
#other_outputs = ["step3_inMINIAODSIM.root","step3_inDQM.root"],
cmssw_version = "CMSSW_10_2_22",
scram_arch = "slc6_amd64_gcc700",
# condor_submit_params = {"sites":"UAF,UCSD"},
)
'''
#for _ in range(25):
total_summary = {}
for task in [step1, step2, step3, step4]:
task.process()
summary = task.get_task_summary()
total_summary[task.get_sample().get_datasetname()] = summary
StatsParser(data=total_summary,
webdir="~/public_html/dump/metis/").do()
time.sleep(600)
from metis.CMSSWTask import CMSSWTask
from metis.Sample import DirectorySample, DummySample
from metis.StatsParser import StatsParser
import time
import numpy as np
tag = "v1"
total_summary = {}
for _ in range(10000):
# first tast has no input files, just make GENSIM from a fragment with pythia commands
gen = CMSSWTask(
sample=DummySample(
N=1, dataset="/WH_HtoRhoGammaPhiGamma/privateMC_102x/GENSIM"),
events_per_output=1000,
total_nevents=1000000,
pset="gensim_cfg.py",
cmssw_version="CMSSW_10_2_5",
scram_arch="slc6_amd64_gcc700",
tag=tag,
split_within_files=True,
)
raw = CMSSWTask(
sample=DirectorySample(
location=gen.get_outputdir(),
dataset=gen.get_sample().get_datasetname().replace(
"GENSIM", "RAWSIM"),
),
open_dataset=True,
#lhe = CMSSWTask(
# sample = DirectorySample(
# location="/hadoop/cms/store/user/dspitzba/tW_scattering/test/",
# #globber="*seed6*.lhe",
# #dataset="/stop-stop/procv2/LHE",
# ),
# events_per_output = 20,
# total_nevents = 100,
# pset = "cfgs/pset_gensim.py",
# cmssw_version = "CMSSW_10_2_7",
# scram_arch = "slc6_amd64_gcc700",
# #split_within_files = True,
# )
gen = CMSSWTask(
sample=DummySample(N=1, dataset="/ttWq/privateMC_102x/GENSIM"),
events_per_output=10,
total_nevents=100,
pset="cfgs/pset_gensim.py",
cmssw_version="CMSSW_10_2_7",
scram_arch="slc6_amd64_gcc700",
tag='v0',
split_within_files=True,
)
for task in tasks:
task.process()
summary = task.get_task_summary()
total_summary[task.get_sample().get_datasetname()] = summary
StatsParser(data=total_summary, webdir="~/public_html/dump/metis/").do()
def submit():
requests = {
#'TTWJetsToLNuEWK_5f_NLO': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_NLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball_retired.tar.xz', # that's the SM point, but using the SMEFT model. No lepton filtering, so name is actually confusing
#'TTWJetsToLNuEWK_5f_NLO_v2': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_NLO_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # that's the actual SM
#'TTWplusJetsToLNuEWK_5f_NLO_v2': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWplusJetsToLNuEWK_5f_NLO_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # that's the actual SM
#'TTWminusJetsToLNuEWK_5f_NLO_v2': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWminusJetsToLNuEWK_5f_NLO_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # that's the actual SM
#'TTWJetsToLNuEWK_5f_EFT_myNLO_full': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_cpt8_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz', # one of the BSM points
#'TTWJetsToLNuEWK_5f_EFT_mix_myNLO_full': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_slc6_amd64_gcc630_CMSSW_9_3_16_tarball.tar.xz', # EFT mix
#'TTWJetsToLNuEWK_5f_EFT_cpq3_4_myNLO_full': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks//TTWJetsToLNuEWK_5f_EFT_myNLO_cpq3_4_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz', # C_pq3 = 4
#'TTWJetsToLNuEWK_5f_NLO': '/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_NLO_slc7_amd64_gcc730_CMSSW_9_3_16_tarball.tar.xz',
'TTWJetsToLNuEWK_5f_SMEFTatNLO_weight':
'/hadoop/cms/store/user/dspitzba/tW_scattering/gridpacks/TTWJetsToLNuEWK_5f_EFT_myNLO_slc7_amd64_gcc700_CMSSW_10_6_19_tarball.tar.xz',
}
total_summary = {}
extra_requirements = "true"
# v6+ is UL
tag = "UL18_v7"
#events_per_point = 250000
#events_per_job = 250
#events_per_point = 2000000
events_per_point = 4000000
events_per_job = 2000
njobs = int(events_per_point) // events_per_job
for reqname in requests:
gridpack = requests[reqname]
task = CondorTask(
sample=DummySample(dataset="/%s/RunIIAutumn18/NANO" % reqname,
N=njobs,
nevents=int(events_per_point)),
output_name="nanoAOD.root",
executable="executables/condor_executable_UL18.sh",
#executable = "executables/condor_executable_UL17.sh",
#executable = "executables/condor_executable_UL16_postVFP.sh",
tarfile="package.tar.gz",
additional_input_files=[gridpack],
open_dataset=False,
files_per_output=1,
arguments=gridpack.split('/')[-1],
condor_submit_params={
"sites":
"T2_US_UCSD", #
#"memory": 1950,
#"cpus": 1,
"memory":
15600,
"cpus":
8,
"classads": [
["param_nevents", events_per_job],
["metis_extraargs", ""],
["JobBatchName", reqname],
["IS_CLOUD_JOB", "yes"],
],
"requirements_line":
'Requirements = (HAS_SINGULARITY=?=True)'
},
tag=tag,
min_completion_fraction=0.90,
)
task.process()
total_summary[
task.get_sample().get_datasetname()] = task.get_task_summary()
StatsParser(data=total_summary,
webdir="~/public_html/dump/tW_gen/").do()
def test_all(self):
dsname = "/blah/blah/BLAH/"
nfiles = 15
s1 = DummySample(N=nfiles, dataset=dsname)
self.assertEqual(len(s1.get_files()), nfiles)
from metis.CMSSWTask import CMSSWTask
from metis.Sample import DirectorySample, DummySample
from metis.Path import Path
from metis.StatsParser import StatsParser
import time
proc_tag = "v1"
special_dir = "workflowtest/ProjectMetis"
step1 = CMSSWTask(
# Change dataset to something more meaningful (but keep STEP1, as we use this
# for string replacement later); keep N=1
sample=DummySample(N=1, dataset="/test/testv1/STEP1"),
# A unique identifier
tag=proc_tag,
special_dir=special_dir,
# Probably want to beef up the below two numbers to control splitting,
# but note that step2 is the bottleneck, so don't put too many events
# in one output file here
events_per_output=30,
total_nevents=120,
# We have one input dummy file, so this must be True
split_within_files=True,
pset="psets/step1.py",
cmssw_version="CMSSW_10_0_0_pre1",
scram_arch="slc6_amd64_gcc630",
)
step2 = CMSSWTask(
sample=DirectorySample(
location=step1.get_outputdir(),
dataset=step1.get_sample().get_datasetname().replace("STEP1", "STEP2"),
from metis.CMSSWTask import CMSSWTask
from metis.Sample import DirectorySample, DummySample
from metis.Path import Path
from metis.StatsParser import StatsParser
import time
for _ in range(25):
proc_tag = "v3"
special_dir = "workflowtest/ProjectMetis"
step1 = CMSSWTask(
# Change dataset to something more meaningful (but keep STEP1, as we use this
# for string replacement later); keep N=1
sample=DummySample(N=1, dataset="/test/ttHH_v2/STEP1"),
# A unique identifier
tag=proc_tag,
special_dir=special_dir,
# Probably want to beef up the below two numbers to control splitting,
# but note that step2 is the bottleneck, so don't put too many events
# in one output file here
events_per_output=50,
total_nevents=1000,
# We have one input dummy file, so this must be True
split_within_files=True,
pset="psets/TOP-RunIIFall17wmLHEGS-00072_1_cfg.py",
cmssw_version="CMSSW_9_3_4",
scram_arch="slc6_amd64_gcc630",
)
step2 = CMSSWTask(
sample=DirectorySample(