GitHub - shedprog/GeomTauId: Geometric Learning for tau identification

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 7 Commits
Backup		Backup
History		History
LR		LR
LRP		LRP
Loss		Loss
Models		Models
__pycache__		__pycache__
DrawCorrelPlot.py		DrawCorrelPlot.py
DrawGraphs.py		DrawGraphs.py
DrawROC.py		DrawROC.py
EvalModel.py		EvalModel.py
EvalModel.pyc		EvalModel.pyc
EvalModelMCC.py		EvalModelMCC.py
EventFiller.py		EventFiller.py
EventFiller.pyc		EventFiller.pyc
ModelLoader.py		ModelLoader.py
ModelLoader.pyc		ModelLoader.pyc
PrepareValidation.py		PrepareValidation.py
Readme.txt		Readme.txt
TrainMultyClass.py		TrainMultyClass.py
Trainer.py		Trainer.py
Trainer_CNNInvers.py		Trainer_CNNInvers.py
Trainer_DeepSet.py		Trainer_DeepSet.py
Utilities.py		Utilities.py
Utilities.pyc		Utilities.pyc
argparse		argparse
bss		bss
check_res.py		check_res.py
config.ini		config.ini
config2.ini		config2.ini
config3.ini		config3.ini
config_deepsets.ini		config_deepsets.ini
config_maxwell.ini		config_maxwell.ini
requirements.txt		requirements.txt
subm_eval_CNN.sh		subm_eval_CNN.sh
test_inspect.py		test_inspect.py
validation.txt		validation.txt

Repository files navigation

In order to Train DPF CNN:
python Trainer.py --config=config_maxwell.ini
DeepSets:
python Trainer.py --config=config_deepsets.ini

#Evaluate:
python Evaluate.py --config=config.ini --file=DYInc_2016_25_tree.root
#For the BinaryClassification Evaluation

python EvaluateMCC.py --config=config.ini --file=DYInc_2016_25_tree.root
#For the MultyClassification Evaluation

config any config with the same structure as config_maxwell.ini
file root file from validation directory
Its is possible to run it using batch system
mkdir jobs
cp HTS_submit.sh bss jobs
./submit_eval_CNN.sh validation.txt config_maxwell.ini