Instructions on running zero finding tools

• (1) Make CSV files

  • update run_runBG.sh and put in your root files and output directory
  • do ./run_runBG.sh

• (2) Analyze CSV files (you will need pandas)

  • update run_zero_finder.sh, put in csv directory from the last step and output text file name
  • do ./run_zero_Finder.sh

Overview

Code consists of two stages

(1) C based tool that generates CSV files of yields from BFI
(2) python pandas script that operates on dataframe created from CSV files

CSV creation

There are two scripts macroBG.C and macroS.C. Both take input of number of leptons and create a directory which have a csv file for every sub region for the defined lepton multiplicity. Each csv file contains the yields and errors for every mperp and risr bin. The signal csv files (macroS) contain all of the mass grid of the specified process. The BG csv files (macroBG) contain the contributions from all background.

The BFI input files have both "signal like" contributions and fakes muf0 elf0 ... etc The macro automatically adds the TH1Ds for both signal-like and fakes.

The inputs files that are automatically read in are the BFI root files.

• BG input file is hardcoded on line 268
• Sig input files are passed in on call (They need to follow a naming convention)

shell scripts to run

• runBG.sh will automatically run over all lepton channels
• runS.sh will run a specific signal over all lepton channels
• run_runS.sh will run over all known signal processes

Pandas scripts

These scripts pull together BG and a chosen signal and stitch together all lepton channels into a single dataframe for analysis

much of the scripts are many little analyses that i comment in and out
to extract reasonable output to look at, there are two approaches
with both I simply pipe the python std out into a file
python script.py > output.x
(1) printing dataframe.to_csv(index=False) -- this method i reanalyze sliced down frames
(2) printing dataframe -- this method is very readable

Old

The main directory hosts code that operates on the latest version of BFI The old directory can be used on the old version with the old naming convention specifically "Ch1L-" (old) versus "CH1L_" (new)

Old2

This holds versions that operates on BFI versions that are after "old" but before the 2L & 3L flavor consoldation