jacanchaplais / heparchy

Hierarchical database storage and access for high energy physics event data.

heparchy

Hierarchical database storage and access for high energy physics event data.

Docstrings are available, markdown documentation coming soon.

Installation

pip install heparchy

Features

• Fast and efficient storage
• Writes and reads from HDF5 files
• Data stored hierarchically
  • Files contain processes
  • Processes contain events
  • Events contain datasets for the final state particles
• Process level metadata can be attached
• Context managers provide access to these containers

Data writing interface

Process

• Metadata writing methods:
  • string(): MadGraph formatted string, ie. p p > t t~
  • decay(): pdgids of incoming and outgoing particles for the hard event
  • com_energy()
  • signal_id(): pdgid of the particle of interest in the hard event
  • custom(): extend with your own key / value metadata pair for the process

Events

• Data writing methods for final state particles:
  • pmu(): 2d numpy array of 4-momenta, each row [px, py, pz, e]
  • pdg(): pdgids
  • is_signal(): boolean tags identifying if particle constituent of signal
  • custom(): extend with your own key / value dataset pair for the event

Coming soon

• Direct interface from HepMC files to HDF5 format
• Jupyter notebook examples
• Pip installation script

Warning: before the first release, the read interface may change to improve consistency with the write interface.

Breaking changes will be avoided following the iminent release of 1.0.0.