An example of a charged current (CC) inclusive cross section analysis using the MINERvA Analysis Toolkit (MAT). Starts from the official MINERvA data preservation "anaTuple" .root files, produced by MasterAnaDev (MAD), and produces all histograms needed to extract a cross section. The basic tutorial does not attempt to constrain backgrounds with sidebands. Based on Ben Messerly's examples of using the MINERvA Analysis Toolkit and work by Mehreen Sultana on her thesis analysis.
First try the wiki instructions for a comprehensive guide to the dependencies as well as this package. If that file no longer exists, you'll have to install the dependencies yourself and follow the instrutions below.
In case the wiki instructions aren't available:
- Install dependencies
- Make a working directory. Mine is called
MINERvA101_2021/ - Download the complete source from MinervaExpt:
git clone https://github.com/MinervaExpt/MAT_IncPions.git #Makes a src subdirectory for you - Make a build directory:
mkdir opt && cd opt && mkdir build && cd build #opt for optimized build as opposed to debug build - Run cmake to generate a build system:
cmake ../../src -DCMAKE_INSTALL_PREFIX=`pwd`/.. -DCMAKE_BUILD_TYPE=ReleaseIfcmakefails talking about dependencies, you may have to manually point to them with command line arguments like-DPlotUtils_DIR=/path/to/PlotUtilsInstallPrefix/lib - Compile and install:
make install
runEventLoop <dataPlaylist.txt> <mcPlaylist.txt>
*** Explanation ***
Reduce MasterAnaDev AnaTuples to event selection histograms to extract a
single-differential inclusive cross section for the 2021 MINERvA 101 tutorial.
*** The Input Files ***
Playlist files are plaintext files with 1 file name per line. Filenames may be
xrootd URLs or refer to the local filesystem. The first playlist file's
entries will be treated like data, and the second playlist's entries must
have the "Truth" tree to use for calculating the efficiency denominator.
*** Output ***
Produces a single runEventLoop.root file with all histograms needed for the
ExtractCrossSection program also built by this package. You'll need a
.rootlogon.C that loads ROOT object definitions from PlotUtils to access
systematics information from these files.
*** Return Codes ***
0 indicates success. All histograms are valid only in this case. Any other
return code indicates that histograms should not be used. Error messages
about what went wrong will be printed to stderr. So, they'll end up in your
terminal, but you can separate them from everything else with something like:
"runEventLoop data.txt mc.txt 2> errors.txt"
- git: version control system. You probably already have this
- CMake 2.8: build system generator for lots of operating systems. You probably already have this
- ROOT: object-oriented toolkit for high energy physics analysis. Make sure to enable at least xrootd, kerberos, and Minuit for this tutorial. Already installed on MINERvA GPVMs
- CVS: legacy version control system for MParamFiles. Install it with your package manager.
- MAT: the MINERvA Analysis Toolkit
- MAT-MINERvA: MINERvA-specific systematics and other plugins to the MAT
- UnfoldUtils: MINERvA's fork of RooUnfold with a compatibility layer for embedding systematics in histograms and MINERvA-specific systematics tweaks (TODO: via github)
- GENIEXSecExtract: MINERvA's closure test procedure
- MParamFiles: additional reweighting and calibration parameters for MINERvA systematics
- MATFluxAndReweightFiles: the flux incident on the MINERvA detector during different run periods and some more reweight configurations for systematic uncertainties