PyHEP -- Python tools for High Energy Physics

Making plots is fun again!

Download

See the download page to download the PyHEP source or i386 binary RPMs, as well as docmentation (including browseable HTML).

Introduction

PyHEP is a set of Python packages and extension modules to provide support for high energy physics data analysis in Python. PyHEP is written by Alex Samuel, alex@alexsamuel.net.

Why Python? I have written down some of the reasons I have given up PAW and Root, and written PyHEP (and not SchemeHep, TclHep, PerlHep, or yet another C++ HEP library) here.

Currenly, the major components of PyHEP include are:

one- and multidimensional histograms
flat-format data tables, similar to "ntuples" in other HEP packages
the ability to read and write HBOOK and Root files

In addition, minor components provide support for Lorentz four-vectors, some miscellaneous numerical and functional-programming constructs, some XML support, etc.

Goals

My main goals in writing PyHEP are:

To produce Python software components necessary for HEP data handling and analysis.
To establish Python interfaces for common HEP functions and objects. The interfaces should be fully "Pythonesque," not simply wrappers around existing C++ or Fortran APIs currently in use.
To develop a system that allows very fast and convenient interactive use, and also provides a well-designed API suitable for writing self-documenting physics analysis scripts and software.

In addition, PyHEP is a test bed for some software experiments I've been thinking about for a while, including:

Using a functional programming style for implementing ntuple analysis.
Using relational database terminology, concepts, and software for ntuple analysis and HEP event data handling.
Designing visual styles for plots according to Edward R Tufte's notions of information design.

To get a sense for PyHEP's capabilities and design, you may browse its documentation on line; see the download section below.

Reporting Bugs

Please send bug reports, requests, comments, complaints, etc. to the author, alex@alexsamuel.net.

Any of these is considered a bug:

Variance among documentation, doc strings, and implementation.
Missing documentation.
An exception raised inside the PyHEP library for valid input.
An exception or stack trace that does not contain useful information about fixing the problem.
Any segv, abort, failed assertion, etc. that causes the Python process to terminate unceremoniously.
Leaked memory or Python objects.
Any unsolicited output to stdout or stderr.
Behaviour that does not correspond to obvious expectations.