Hpluscharm

Higgs plus charm analysis framework coffea based

Setup

! This enviroment should under bash shell!

Coffea installation with Miniconda

only first time

git clone git@github.com:Ming-Yan/Hpluscharm.git For installing Miniconda, see also https://hackmd.io/GkiNxag0TUmHnnCiqdND1Q#Local-or-remote

wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
# Run and follow instructions on screen
bash Miniconda3-latest-Linux-x86_64.sh

NOTE: always make sure that conda, python, and pip point to local Miniconda installation (which conda etc.).

You can either use the default environmentbase or create a new one:

# create new environment named coffea with env.yml
conda coffea create -f env.yml
# activate environment coffea
conda activate coffea

Setup things with x509 proxy

# There are some dependencies which are currently not available as packages, so we'll have to copy them from `/cvmfs/grid.cern.ch/etc/grid-security/`
   
    mkdir ~/.grid-security
    scp -r lxplus:/cvmfs/grid.cern.ch/etc/grid-security/* ~/.grid-security
   
    
#All that's left is to point `voms-proxy-init` to the right vomses directory
   
    voms-proxy-init --voms cms --vomses ~/.grid-security/vomses

Other installation options for coffea

See https://coffeateam.github.io/coffea/installation.html

Running jupyter remotely

See also https://hackmd.io/GkiNxag0TUmHnnCiqdND1Q#Remote-jupyter

1. On your local machine, edit .ssh/config:

Host lxplus*
  HostName lxplus7.cern.ch
  User <your-user-name>
  ForwardX11 yes
  ForwardAgent yes
  ForwardX11Trusted yes
Host *_f
  LocalForward localhost:8800 localhost:8800
  ExitOnForwardFailure yes

2. Connect to remote with ssh lxplus_f
3. Start a jupyter notebook:

jupyter notebook --ip=127.0.0.1 --port 8800 --no-browser

activate enviroment once you have coffea framework

1. activate coffea conda activate coffea
2. setup the proxy voms-proxy-init --voms cms --vomses ~/.grid-security/vomses
3. Run the framework

Structure

Each workflow can be a separate "processor" file, creating the mapping from NanoAOD to the histograms we need. Workflow processors can be passed to the runner.py script along with the fileset these should run over. Multiple executors can be chosen (for now iterative - one by one, uproot/futures - multiprocessing and dask-slurm).

For now only the selections and histogram producer are available.

To test a small set of files to see whether the workflows run smoothly, run:

python runner.py  --json metadata/higgs.json --limit 1 --wf HWW2l2nu ( --export_array)

• Samples are in metadata
• Corrections in utils/correction.py

Scale out @lxplus (not recommended)

python runner.py  --json metadata/higgs.json --limit 1 --wf HWW2l2nu -j ${njobs} --executor dask/lxplus

Scale out @DESY/lxcluster Aachen

python runner.py  --json metadata/higgs.json --limit 1 --wf HWW2l2nu  --year ${year}  -j ${cores} -s ${scaleout jobs} --executor parsl/condor --memory ${memory_per_jobs} --export_array

Scale out @HPC Aachen

python runner.py  --json metadata/higgs.json --limit 1 --wf HWW2l2nu -j ${njobs} --executor parsl/slurm

Make the json files

Use the fetch.py in filefetcher, the $input_DAS_list is the info extract from DAS, and output json files in metadata/

python fetch.py --input ${input_DAS_list} --output ${output_json_name} --site ${site}

Create compiled corretions file(pkl.gz)

Use the utils/compile_jec.py, editted the path in the dict of jet_factory

python -m utils.compile_jec data/jec_compiled.pkl.gz