Root to HDF5 Tree Converter

Run bin/ttree2hdf5 to convert your root tree to an HDF5 file.

Branches that contain basic types (one value per entry in the tree) are stored as a 1D array of compound data type. Branches that store vector<T> are stored in a 2D array, while branches of type vector<vector<T> > are stored in a 3D array. If you want more dimensions feel free to file an issue.

Note that we only support the types listed above. The purpose of this package is not to provide a generic converter from ROOT to HDF5, but rather to convert simple ROOT files to a more widely supported data format.

Output Format

You can specify the maximum dimensions of the HDF5 array with

ttree2hdf5 <root-file> -o <output-file> -l [dims..]

where the [dims..] argument can have up to two integers.

You can also filter branches with the --branch-regex option.

For more options check ttree2hdf5 -h.

Setup

The dependencies are HDF5 (C++ interfaces required), boost (for program_options) and ROOT. If these are available on your system you can build with make.

Building with CMake

We also include a CMake file. Run

mkdir build
cd build
cmake ..
make -j 4

If you don't have access to HDF5, you can tell CMake to build it by replacing the cmake .. line with

cmake -DBUILTIN_HDF5=TRUE ..

Note that this takes about 10 minutes. If you already have HDF5 built somewhere, you can also add the executables to your path: cmake should find them.

Building in Special Environments

Here we count LHC experiments like ATLAS as "special": you typically need to figure out how to find everything on /cvmfs/. There are scripts to setup these environments in setup/, e.g.:

source setup/atlas-cvmfs.sh

Since we don't currently have HDF5 working in LCG (see this JIRA issue) I've built a version on AFS which should be compatible with the above setup script. Run

HDF5_ROOT=${HDF5_ROOT} cmake ..

And be sure to complain to LCG so they fix their HDF5 installation.

Hacking This Code

I've tried to make this code as modular and straightforward as possible. Of course if anything is unclear you should file an issue.

The main algorithm ttree2hdf5 is a thin wrapper on copy_root_tree. Reading the code in src/copy_root_tree.cxx should be a decent introduction to the higher level classes you'll need to interact with.

As for compiling: anything in src/ which starts with ttree2hdf* will compile into an executable by default, whereas anything in src/ will compile and link to the executables by default.

To Do

Other things I might do sometime:

• Allow grouping of different branches into different output datasets
• Give the option to supply a TCut string to skim the ntuple
• Merge with the other root to hdf5 converter I made for histograms.

If you want these things, or anything else, file an issue.