Skip to content

Latest commit

 

History

History
157 lines (119 loc) · 7.73 KB

README.md

File metadata and controls

157 lines (119 loc) · 7.73 KB

MacrosNtuples/plotting

This folder contains several scripts to draw plots generated using the tools in MacrosNtuples/l1macros. You can draw plots individually by calling drawplots.py, use one of the make_"OBJ"_plots.sh to draw a bunch of plots for a given type of object (Muon, EGamma, Jets or MET) at once, or use the draw_all.sh to make all plots for all 4 object types.

You can find plots generated with these tools for all of 2022 here: https://test-l1studies-plots.web.cern.ch/L1studies/

Drawing plots individually

This is done using the drawplots.py program. If saved, the plots will be stored in a directory called plotL1Run3. You need to create this directory yourself beforehand.

Types of plots

The drawplots.py can make different types of plots, specified with the --type command line argument.

These types are:

  • --type efficiency computes and draws one/several efficiencies given numerator(s) and denominator(s).
  • --type resolvsx compute the response and resolution of a 2D histogram along its x axis. Makes two separate plots.
  • --type distribution draws the distribution from a 1D histogram.
  • --type profilex_fromh2 draws the profile of a 2D histogram along its x axis.

Command line arguments

The different type need different command line arguments. These are:

  • Common, required arguments:

    • --type The type of plot (see above)
    • --input The input file to read the histograms from

  • Required for efficiency:

    • --den Name of the denominator histogram(s). If only one, the same is used for every numerator(s). If several, pairs each denominator with one numerator, the numbers need to be the same.
    • --num Name of the numerator histogram(s).

  • Optionnal for efficiency:

    • --addnumtoden Add numerator histogram to denominator

  • Required for resolvsx and profilex_fromh2:

    • --h2d The 2D histogram to use.

  • Required for distribution:

    • --h1d The 1D histogram to use.

  • Commun, optionnal arguments:

    • --xtitle, --ytitle, --ztitle The X/Y/Z axis title
    • --extralabel A label to put in the top left corner of the plot.
    • --toplabel A label to put above the plot (for sqrt(s) and lumi values)
    • --setlogx Whether to set the x axis to log scale or not.
    • --axisranges Axis ranges [xmin, xmax, ymin, ymax, zmin, zmax]
    • --legend Legend labels. For efficiency, one per numerator.
    • --legendpos Position of the legend. (Either top or bottom, default is bottom)
    • --saveplot Whether to save the plot(s) or not.
    • --plotname Name of the plot to save. Saves both a .png and a .pdf
    • --interactive Run in interactive mode.
    • --suffix_files Input files suffix.
    • --nvtx_suffix Suffix to append to dirname and histogram names, to make plots in bins of nvtx.

Example usage

To draw the effiency, given a file input.root with two histograms numerator and denominator:

python3 drawplots.py --type efficiency --input input.root --num numerator --den denominator --saveplot True --plotname myplot

This will save to files, myplot.png and myplot.pdf.

Drawing all the plots for one kind of object

2022 setup

This is done using one of the for script, make_mu_plots.sh, make_eg_plots.sh, make_jets_plots.sh or make_etsum_plots.sh. These are for Muons, EGamma, Jets and MET respectively.

Every script is called the same way:

./make_"OBJ"_script dirname lumi_value nvtx_suffix

The dirname argument is required. The script will search for input files and store plots inside dirname. Depending of the type of object, you will require an input files named dirname/all_zmumu.root, dirname/all_zee.root, dirname/all_mujet.root and/or dirname/all_photonjet.root. You will also need to create a dirname/plotL1Run3 directory to save the plots.

The lumi_value is the value of the integrated luminosity to be added in the label above the plot.

The nvtx_suffix can be passed if you want to make the plots for a certain bin of nvtx. It is of the form _nvtxXXtoYY where XX and YY are the minimum and maximum nvtx values respectively. Make sure the correponding histograms exist in your input file. To generate histograms in bins of nvtx, see --plot_nvtx and --nvtx_bins arguments in MacrosNtuples/l1macros/performances.py or MacrosNtuples/l1macros/performances_nano.py. If nvtx_suffix is passed, then the corresponding directory to save the plots should be dirname/plotL1Run3_nvtxXXtoYY (where _nvtxXXtoYY is the argument that was passed).

Example usage

To draw all plots for muons, for a directory MyData containing the input file MyData/all_zmumu.root and directory MyData/plotsL1Run3 to store the plot, corresponding to X inverse femtobarns of inverse luminosity, run:

./make_mu_plots.sh MyData X

If you now want to make the same plots, with nvtx between 30 and 40, you will need the directory MyData/plotsL1Run3_nvtx30to40, and run:

./make_mu_plots.sh MyData X _nvtx30to40

2023 setup

The bash script from 2022 should be somewhat compatible with histograms generated from NANOAOD using performances_nano.py, but some figures might be broken.

The recommanded scripts for the 2023 setup are the make_ZToMuMu_plots.py make_ZToEE_plots.py make_MuonJet_plots.py make_PhotonJet_plots.py python scripts.

Each script draws all the plots for the corresponding channel. All the scripts take the same following inputs:

  • --dir The directory to read the inputs files from and draw the plots to.
  • --config The YAML configuration file to read from. The default ones are the templates ../config_cards/full_<Channel>.yaml
  • --lumi The integrated luminosity to display in the top right corner of the plot.

make_<Channel>_plots.py will look in the input directory for an input file called all_<Channel>.root, and will try to save the plots in a subdirectory called plotsL1Run3 inside the input directory. You need to create this subdirectory before running the scripts.

Furthermore, the scripts will only draw a plot if the field for the corresponding kind is set in the configuration file. For instance, by setting TurnOns: false in the configuration file, the script will not produce any of the turn on plots. The scripts also rely on the values of the different fields in the configuration files to look for histograms in the input file.

You should make sure the configuration file used for producing the plots is compatible with the one given to perfomances_nano.py to fill the histograms

Otherwise, the script might look for histogram names that are not present in the input file. The default cards contain comments on how to use them.

Example usage

To draw all plots for muons, for a directory MyData containing the input file MyData/all_ZToMuMu.root, produced with the configuration card my_ZToMuMu_config.yaml, and a directory MyData/plotsL1Run3 to store the plot, corresponding to X inverse femtobarns of inverse luminosity, run:

python3 make_ZToMuMu_plots.py --dir MyData --config my_ZToMuMu_config.yaml --lumi "X fb^{-1}"

Making plots in bins of nvtx is not supported yet.

Drawing all the plots for all objects

2022 setup

The script draw_all.sh loops on a (hard-coded) list of directories and calls the for make_"OBJ"_plots.sh scripts on each of them, with all possible nvtx_suffix. It takes no arguments, and can be run by simply calling ./draw_all.sh (provided that all needed directories and input files exist).

2022 setup

The script draw_nano.sh works in a similar way to draw_all.sh. It internaly calls the make_<Channel>_plots.py script, and runs on a (hard-coded) loop. You can run it by simply calling ./draw_nano.sh.