RootTuple is a lightweight C++ library, providing FORTRAN subroutines which enable the output of ROOT n-tuples. It was written specifically for high energy physics applications, since each element in the n-tuple can represent an event.
This is in no way supposed to replace the Les Houches Event File (LHEF) format. Rather, it was created to allow FORTRAN programs to output fully differential cross section information as an n-tuple. Selection cuts can then be applied to, and histograms filled from, this n-tuple.
A description of the subroutines provided is found in the documentation, together with a guide on how they may be used. Note: This PDF was created for RootTuple 1.0.0 and makes outdated references to files hosted on HepForge. The latest version of RootTuple is available in GitHub.
Firstly, we require ROOT to be installed on your system (you can download it here).
Then, during installation, RootTuple needs root-config to be in your PATH environment variable.
The easiest way to achieve this is to source the ROOT setup script (e.g. thisroot.sh).
There are two possible installation methods for RootTuple, both of which are further described in the documentation.
The recommended installation method uses the CMake build system v2.8.6 or higher (which can be downloaded here).
This has the advantage that it assesses your FORTRAN compiler for its name mangling conventions, allows the installation to be separated from the source code and also produces a RootTuple-config executable.
tar -xzvf RootTuple-1.0.0.tar.gz
mkdir build; cd build
cmake -DCMAKE_INSTALL_PREFIX=<prefix> ../RootTuple-1.0.0
make
make installThe alternative installation uses a premade Makefile and assumes a particular name mangling convention (should work for g77, gfortran and ifort compilers):
tar -xzvf RootTuple-1.0.0.tar.gz
cd RootTuple-1.0.0
cp external/FC_namemangling.h src/
make