kakwok / LLP_coffea

coffea analysis of LLP

LLP_coffea

coffea analysis of LLP

Getting coffea

Interactive usage on LPC

First time setup:

tar -xvf coffeaenv.tar.gz
sed -i "3a export PYTHONPATH=${PWD}/coffeaenv/lib/python3.7/site-packages:\$PYTHONPATH" coffeaenv/bin/activate
source coffeaenv/bin/activate

After first time set-up, enable coffea env with

source coffeaenv/bin/activate

Jupyter notebook access

You can launch Jupyter notebook server on LPC node and forward the port to your computer.

To do so, you need to connect to LPC with

ssh -Y -L localhost:8887:localhost:8887 [e-mail]

After enabling coffea env, you can start the server with

jupyter notebook --no-browser --port=8887 --ip 127.0.0.1

Preparing filepaths

MC samples are located in the store/group/lpclonglived/HNL area, which is accessible via xrootd

To make the input file for the HNLprocessor, create the json file with

python filepaths.py

Making corrections

Correction files includes WpT reweighting, pileup corrections, and MC cross section.

python compile_corrections.py

Running the processor

Edit runHNL.py with the output filename and the fileset json (made with filepaths.py) For local testing with a small fileset:

python runHNL.py --test

For local testing with a fraction of the full fileset:

python runHNL.py --local

Running with condor

To run with lpc condor, first follow instructions at lpcjobqueue!!

voms-proxy-init
./shell                          ## enter virtual environment
[python makezip.py]              ## if you have updated anything in the `HNLprocessor` package
Singularity> python runHNL.py --condor   ## switch to runLPC in the script 

Processing all samples takes about ~10 mins.