Code for the GT tau performance enhancer

for usage in the cern swan infrastructure.

1. Setup:

Update the python awkward library:

1. Inside swan, open a terminal and run:

pip install --user awkward --upgrade
pip install --user pyarrow --upgrade
pip install --user vector

2. Restart swan afterwards with the following configuration:

It is recommended to request access to a GPU via the cern service portal as 32 GB of ram are required for the minimum bias evaluation.

3. Clone this repository in swan via the git > Clone a repository option in swan.
4. Enter the tau-pe directory and
5. run the setup.ipynb notebook, which clones the Menu Tools and generates required input data (might take a few hours, but has to be done only once)

The notebooks are intended to be run in the following order:

1. smallnetdata.ipynb
2. smallnet_modeltraining.ipynb
3. rte.ipynb To get the rate plots this file has to be run with sample = "MinBias"

to evaluate the performance and produce the plots, run:

cd tau-pe
unset PYTHONHOME && unset PYTHONPATH && cd Phase2-L1MenuTools/
. menuenv/bin/activate
cp ../taus.yaml configs/V37nano/objects/taus.yaml
cp ../taus_rate.yaml configs/V37nano/rate_plots/
rate_plots configs/V37nano/rate_plots/taus_rate.yaml 
cp ../gttau_matching.yaml configs/V37nano/object_performance/
object_performance configs/V37nano/object_performance/gttau_matching.yaml 

the Plots are in:

tau-pe/Phase2-L1MenuTools/outputs/V37nano/object_performance/