Skip to content

Latest commit

 

History

History
107 lines (86 loc) · 4.49 KB

README.md

File metadata and controls

107 lines (86 loc) · 4.49 KB
Raw

Tutorials

Here is a collection of tutorial macros. If you are learning this software, it is recommended to go through these tutorials in the numerical order given below.

  • Note: execute macros from the sidis-eic top directory, not from this tutorial directory, e.g., root -b -q tutorial/analysis_template.C

Generate or Obtain Simulation Output Trees

To run tutorials, you need to generate or obtain ROOT files, from fast or full simulation

  • the datarec/ directory is provided for storage of these files, but is not required to use
  • the following sub-sections describe how to obtain these files from the common storage area
    • this requires access to S3, the common storage area
    • environment variables $S3_ACCESS_KEY and $S3_SECRET_KEY must contain the login and password; follow s3tools documentation for guidance

Fast Simulation

  • run tutorial/makeSampleDelphesFiles.sh to download HEPMC files from S3, and run them through Delphes
    • edit the settings in this script, such as beam energy
    • change the maximum number of files to process, to limit disk space consumption

Full Simulation

  • full simulation files are streamed from S3 using tutorial/s3files.config

Introductory Notes

Switching between Fast and Full Simulations

  • many of these examples focus on fast simulations; to switch between fast and full simulations, change the Analysis-derived class in the macro:
    • AnalysisDelphes is used for fast simulation
    • AnalysisDD4hep is used for full simulation from DD4hep+Juggler
  • note: some extra settings and features differ between these

Input File Lists (Config Files)

  • in the analysis macros, the input files are specified by a list, a "config file", with the additional columns
    • see documentation here for the formatting of these files, as well as scripts to help generate them
    • for example, this file allows one to combine different Q2 regions together using relative weights

Tutorials:

Each of these examples has two macros:

  • analysis macro, to run an Analysis-derived class, which will analyze simulated data in an event loop, and generate a variety of output data structures
  • postprocessor macro, to process the output from the analysis macro, such as drawing plots
  • the analysis macro will take some time to run, since it runs the event loop; the postprocessor macro is typically fast, since it analyzes the resulting data structures
  1. Template
  • analysis_template.C: minimal analysis macro to demonstrate how to run Analysis; no bins are specified
  • there is no postprocessor macro (see other examples); instead, inspect the output root file from the analysis macro, to learn what objects are stored
  1. (x,Q2) Binning
  • analysis_xqbins.C: bin the analysis in a few 2D bins of x and Q2
    • there is a switch statement to allow the choice of various example binning schemes
    • this example also describes how cuts are defined
  • postprocess_xqbins_draw.C: draws a couple sample histograms for the given binning scheme
  1. Full Simulations (all other tutorials are for fast simulations)
  • analysis_dd4hep.C: basically a copy of analysis_xqbins.C, but shows how to analyze full simulation data; the main difference is using AnalysisDD4hep instead of AnalysisDelphes
  • postprocess_dd4hep_draw.C: clone of postprocess_xqbins_draw.C, specific for this example
  1. Average kinematics table
  • analysis_qbins.C: bin the analysis in several Q2 bins, for a couple pT bins
  • postprocess_qbins_tables.C: produce a text file containing tables of averages of kinematics variables, where each row is for a Q2 bin; one table is produced for each pT bin
  1. Draw ratio of histograms for different y-minima
  • analysis_yRatio.C: bins in 3 y-minima, along with a full-y bin
  • postprocess_yRatio.C: produces ratios of all histograms, with a y-minimum divided by no y-minimum
  1. Test DAG lambda staging
  • analysis_testDAG.C: define multi-dimensional binning
  • postprocess_testDAG.C: contains various lambdas examples and describes how to stage them
  1. Conditional Subloops
  • analysis_coverage.C: define 4-D binning scheme, including extra "full-range" bins
  • postprocess_coverage.C: draw certain plots, while restricting certain bins to be "full-range", i.e., "integrated over"; in order to restrict the execution of certain subloops, conditional control functions are used