nanogen

Setup

Set environment variables before sourcing setup.sh and consider creating an alias.

# required
export NG_CERN_USER="your_cern_username"
# optional (storage location of software and local files; defaults to "/data" in the repo)
export NG_DATA_BASE="/data/dust/user/$( whoami )/nanogen_data"
# setup
source setup.sh ""

Tasks

flowchart TD
  FetchLumiMask --> PrepareForConfig
  GetDatasetLFNs == many ==> PrepareForConfig["PrepareForConfig ❗️"]
  GetDatasetLFNs --> CreateNano
  CreateCMSRunConfig --> CreateNano
  BundleRepo -- htcondor<br />workflow --> CreateNano
  FetchLFN -. optional .-> CreateNano
  FetchLumiMask -. data only .-> CreateNano
  CreateNano ----> GenerateNanoDocs
  CreateNano --> CollectNanoSizes
  CollectNanoSizes --> MergeNano
  CreateNano -- many --> MergeNano
  MergeNano --> ValidateNano
  MergeNano --> CreateDBEntry
  ExportCentralNanoKey -----> CreateDBEntry
  CreateDBEntry == mane ==> ListDBEntries

Loading

Almost all tasks have an accompanying *Wrapper task that adds functionality like --dataset-names PATTERNS, --skip-dataset-names PATTERNS to trigger multiple wrapped tasks at once.

❗️ Note: PrepareForConfig --config-name ... should be run manually before CreateNano or any other task downstream, mainly to have the list of LFNs available a-priori, since dynamic dependency generation can be costly in this case.

Yaml configs

General config

The name of general configs must be in the format config_<ERA>_<VNANO>.yaml.

# example

# cmssw setup
cmssw_version: CMSSW_X_Y_Z
scram_arch: el9_amd64_gccXYZ
cmssw_env_name: nanogen
sandbox_script: cmssw_nanogen.sh

# arguments for cmsDriver.py that creates the cmsRun config
global_tag:
  mc: ...
  data: ...
era:
  # use anchor if same era
  mc: &era Run3
  data: *era

# lumi mask for data
lumi_mask: /path/or/url/to/lumi_mask.json

# postfix for produced nano datasets in the cms-store-like path
campaign_postfix: NanoAODv<VNANO>UHH

# references to dataset and nano configs
dataset_config: datasets_<ERA>.yaml
nano_config: nano_run3_<VNANO>.yaml

Dataset entries in datasets_*.yaml files

Central samples

# example

# standard dataset entry
tt_dl:
  # the miniaod key
  key: /TTTo2L2Nu_TuneCP5_13TeV-powheg-pythia8/RunIISummer20UL16MiniAODAPVv2-106X_mcRun2_asymptotic_preVFP_v11-v1/MINIAODSIM

  # list of lfns to skip (optional)
  skip_lfns:
    - ...

  # custom era to use for this dataset, defaults to the era of the config (optional)
  era: optional_custom_era

  # custom gt to use for this dataset, defaults to the gt of the config (optional)
  global_tag: optional_custom_gt

  # custom campaign postfix, defaults to the postfix of the config (optional)
  campaign_postfix: optional_custom_postfix


# extension of tt_dl
# ❗️ name must start with tt_dl and end with _ext{N}
tt_dl_ext1:
  # all other fields as shown above
  key: ...


# variation of tt_dl
# ❗️ name must start with tt_dl and end with _{up,down}
# (to standardize the naming of the cmsdb entry which requires a specific format)
tt_dl_tune_up:
  # all other fields as shown above
  key: ...

# data entry
data_e_c:
  # mini aod key
  key: ...

  # mandatory prompt flag
  prompt: True

  # optional key of the corresponding central nano aod
  nano_key: ...

  # optional jec era
  jec_era: RunCD

Privately produced samples

hh_vbf_hbb_htt_kv1_k2v0_kl1_prv_madgraph:
  # a fake miniaod key (note the UHH suffix behind the miniaod campaign)
  key: /VBFHHto2B2Tau_CV_1_C2V_0_C3_1_TuneCP5_13p6TeV_madgraph-pythia8/Run3Summer22MiniAODv4UHH-130X_mcRun3_2022_realistic_v5-v2/MINIAODSIM

  # a private mapping entry qualifies the dataset as being private
  private:
    # local path where mini files are stored
    path: /data/dust/user/riegerma/mcgen_vbf_data/outputs/vbf_22pre_cv1.0_c2v0.0_kl1.0
    # regex pattern to match the mini files
    regex: mini_.+_.+_500\.root
    # a dataset id that will be used in the cmsdb entry export
    id: 22761002

References

• General nano docs: https://gitlab.cern.ch/cms-nanoAOD/nanoaod-doc
• Private productions: https://gitlab.cern.ch/cms-nanoAOD/nanoaod-doc/-/wikis/Instructions/Private-production
• CMS DAS: https://cmsweb.cern.ch/das
• HLepRare: https://cms-higgs-leprare.docs.cern.ch/#common-hleprare-nanoaod-skims