Purpose

Create necessary input CSV files to use scripts from RegressionWithKeras repository

Create CSV files to train network to predict extrapolation weights

Script: MakeCSVfiles_NN.py

Use this script to produce CSV files from ROOT files to be used to train a DNN to predict extrapolation weights

Dependencies

• ROOT
• pandas

How to use?

This script will produce .txt files (containing mean and std. dev. of each feature, needed for inference) and .csv files (needed for training)

• Set Particles
  • One CSV file will be produced for each particle, eta range and energy
  • If more than one particle type is provided, pdgID will also be stored in the CSV files
• Set InputPATH: path to input ROOT files
• Set OutputPATHBase: location where output CSV files will be written (CSV files will be written inside a folder named accordingly the set of particles used)
• Set TTreeName: name of the TTree available in the input ROOT files
• Set ProduceTXTs to True (unless TXTs were already produced and are available in the corresponding location)

Run it:

python MakeCSVfiles_NN.py

Create CSV files with firstPCAbin values from FCS

Script: MakeCSVfiles_FCS.py

Use this script to produce CSV files from ROOT files with firstPCAbin values

Dependencies

• ROOT

How to use?

• Set InputPATH: path to input ROOT files
• Set OutputPATH: where output CSV files will be saved
• Set TTreeName: name of TTree from input ROOT files

Run it:

python MakeCSVfiles_FCS.py