Config Files

Config Files

There are two main input files needed to run the code which I will call the Run Config File and Style File. The Run Config file is what is an immediate input of the Plotter: it needs a Run Config File to run the program (hence the name). The Style File is specified in the Run Config File and it gives the global styling of the graphs through the TStyle object in ROOT.

This page is used to learn the format and uses of these files.

Run Config File

This is the file that is given to run the program, ie ./Plotter <CONFIG FILE>. The config file generally will look something like this:

luminosity 15900
output     output.root
style      cms

files/Data.root Data.root

files/DYJetsToLL_M-10to50_amcatnloFXFX.root  DY+Jets.root  18610   0.999226 bg
...
...
file/Signal.root   Signal.root #### ### sig

The first three lines are variables used to set up the subsequent run: without these, the program will not run. The luminosity and output are fairly obvious, but the style is the name of the style file you are using. These are stored in the style directory, namely Plotter/style. More on the style classes in the next section.

The next lines are for the actual files to be normalized. There are three types of files used in the program: Data, Background, Signal. Background is what is actually stacked while the Data and Signal are plotted alone.

For Signal and Background there are several pieces of information needed:

<Infile location>  <Outfile location>  <Cross-section>  <Skim efficiency>  <tag> 

• Infile location is just the location of the specific file
• Outfile location is the name for the normalized file. NOTE: multiple files can be associated with a single outfile. This allows multiple samples to be added together easily (eg ZZ, WZ, and WW can all be added to a file VV.root)
• Cross-section is just the cross-section for the input file
• Skim Efficiency is important for the Ntuples, for privately generated samples, just set this to 1
• tag tells the program what type of sample you are using the tags are:
  • bg -- background
  • sig -- signal

For Data, since it doesn't need to be normalized, the only information needed is the infile location and outfile location. The tag is also excluded because a sample without a cross-section is assumed to be data.

As with all BSM3G code, these config files support C++ and Python commenting (// and # respectively).

Style File

Style files are simply holders for the information that will go into the overall TStyle class for each graph. Benefits are that the style for the graphs are easily configured and allow for different styles for different uses (eg different styles for different journals). The drawback is these styles are global so minor tweaks still require changing in ROOT or making changes to the Plotter code. These files actually interface with a helper class called Style. To read more in depth about this class, please go here

Regardless, if you look in the file style/default, you will see a long list of variables, some with numbers after them, some without. This list is all of the variables that can be changed in the TStyle class. The style file only reads in variables that have a input value after them, so all of the variables without numbers are skipped.

If you have any questions about what any of the variables do, visit the TStyle ROOT page and search for the variable; each variable VAR correspond to a ROOT function setVAR().

There are a few Plotter specific style variables though:

• PadRatio: This is the ratio of the height of the top graph and bottom graph. If only the top graph is requested (see Options), this variable is ignored
• TopWSRatio: This variable stands for "Top Whitespace Ratio." When making the stack plots, the ratio of the height of the graph to the height of the whitespace above the graph can be set here. example: the graph has a maximum of 1200 and the TopWSRatio is set to 15. The Y Axis will be set to the range 1280.
• RebinLimit: The plotter automatically combines bins together if they have a large error (if Error is just sqrt(Number of Events) then it effectively adds bins together that have low number of events.) The percent error of a bin must be less than (up to a constant factor) of the RebinLimit, else it will be added to adjacent bins till the percent error is lower than the RebinLimit. To remove rebinning, just make the RebinLimit an arbitrarily large number.
• DivideBins: This represents a boolean value, so set it only to 0 or 1 (ie false or true). To get a smoother distribution after variable rebinning, you can set the option to divide each bin by its width. This also updates the Y Axis label to [Events/GeV].
• BinLimit: There are certain graphs we don't want rebinned (eg Multiplicity graphs). In general these graphs have a small number of bins, so to insure these graphs aren't rebinned, the BinLimit is the minimum number of bins needed after selecting the bins to rebin the graphs in the final root file.

The style file is read in by the Style Class, so if any features want to be added, be sure to have them read in properly by the Style Class