instructions for slow analysis example

ewkcoffea_setup.txt

# in a fresh conda environment, >= py3.8
conda install xrootd -c conda-forge
pip install coffea xgboost mt2

git clone https://github.com/TopEFT/topcoffea.git -b coffea2023
pushd topcoffea
pip install -e .
popd

git clone https://github.com/cmstas/ewkcoffea.git -b coffea2023
cd ewkcoffea
pip install -e .

cd analysis/wwz
vi run_wwz4l.py
# comment out line 20
# comment out line 343
# on line 327 of run_wwz4l.py edit it to look like this:
# #dataset_runnable, dataset_updated = preprocess(
# #    fileset, 
# #    step_size=50_000, 
# #    align_clusters=False, 
# #    files_per_batch=1,
# #    save_form=True,
# #)
#
# with gzip.open("dataset_runnable_feb07_2024_ewkcoffea.json.gz") as fin:
#     dataset_runnable = json.load(fin)
# dataset_runnable = filter_files(dataset_runnable)
# first_key, first_value = dataset_runnable.popitem()
# dataset_runnable = {first_key: first_value}
wget https://github.com/lgray/ewkcoffea/raw/main/analysis/wwz/dataset_runnable_feb07_2024_ewkcoffea.json.gz
python run_wwz4l.py ../../input_samples/cfgs/wwz_analysis/mc_sig_bkg_samples.cfg,../../input_samples/cfgs/wwz_analysis/data_samples.cfg -x iterative --do-systs