Skip to content

Package to parse decay files, describe and convert particle decays between digital representations.

License

Notifications You must be signed in to change notification settings

scikit-hep/decaylanguage

Repository files navigation

DecayLanguage logo

DecayLanguage: describe, manipulate and convert particle decays

Scikit-HEP PyPI Package latest release Conda latest release Supported versions DOI

GitHub Actions Status: CI Code Coverage Documentation Status

Binder demo

DecayLanguage implements a language to describe and convert particle decays between digital representations, effectively making it possible to interoperate several fitting programs. Particular interest is given to programs dedicated to amplitude analyses.

DecayLanguage provides tools to parse so-called .dec decay files, and describe, manipulate and visualize decay chains.

Installation

Just run the following:

pip install decaylanguage

You can use a virtual environment through pipenv or with --user if you know what those are. Python 3.8+ supported (see version 0.12 for Python 2 & 3.5 support, 0.14 for Python 3.6 support, 0.18 for Python 3.7 support).

Dependencies: (click to expand)

Required and compatibility dependencies will be automatically installed by pip.

Required dependencies:

  • particle: PDG particle data and identification codes
  • NumPy: The numerical library for Python
  • pandas: Tabular data in Python
  • attrs: DataClasses for Python
  • plumbum: Command line tools
  • lark: A modern parsing library for Python
  • graphviz to render (DOT language) graph descriptions and visualizations of decay chains.

Python compatibility:

Getting started

The Binder demo is an excellent way to get started with DecayLanguage.

This is a quick user guide. For a full API docs, go here (note that it is presently work-in-progress).

What is DecayLanguage?

DecayLanguage is a set of tools for building and transforming particle decays:

  1. It provides tools to parse so-called .dec decay files, and describe, manipulate and visualize the resulting decay chains.

  2. It implements a language to describe and convert particle decays between digital representations, effectively making it possible to interoperate several fitting programs. Particular interest is given to programs dedicated to amplitude analyses.

Particles

Particles are a key component when dealing with decays. Refer to the particle package for how to deal with particles and Monte Carlo particle identification codes.

Parse decay files

Decay .dec files can be parsed simply with

from decaylanguage import DecFileParser

parser = DecFileParser('my-decay-file.dec')
parser.parse()

# Inspect what decays are defined
parser.list_decay_mother_names()

# Print decay modes, etc. ...

A copy of the main DECAY.DEC file used by the LHCb experiment is provided here for convenience.

The DecFileParser class implements a series of methods giving access to all information stored in decay files: the decays themselves, particle name aliases, definitions of charge-conjugate particles, variable and Pythia-specific definitions, etc.

It can be handy to parse from a multi-line string rather than a file:

s = """Decay pi0
0.988228297   gamma   gamma                   PHSP;
0.011738247   e+      e-      gamma           PI0_DALITZ;
0.000033392   e+      e+      e-      e-      PHSP;
0.000000065   e+      e-                      PHSP;
Enddecay
"""

dfp = DecFileParser.from_string(s)
dfp.parse()

Advanced usage

The list of .dec file decay models known to the package can be inspected via

from decaylanguage.dec import known_decay_models

Say you have to deal with a decay file containing a new model not yet on the list above. Running the parser as usual will result in an error in the model parsing. It is nevertheless easy to deal with this issue; no need to wait for a new release: Just call load_additional_decay_models with the models you'd like to add as arguments before parsing the file:

dfp = DecFileParser('my_decfile.dec')
dfp.load_additional_decay_models("MyModel1", "MyModel2")
dfp.parse()
...

This being said, please do submit a pull request to add new models, if you spot missing ones ...

Visualize decay files

The class DecayChainViewer allows the visualization of parsed decay chains:

from decaylanguage import DecayChainViewer

# Build the (dictionary-like) D*+ decay chain representation setting the
# D+ and D0 mesons to stable, to avoid too cluttered an image
d = dfp.build_decay_chains('D*+', stable_particles=('D+', 'D0'))
DecayChainViewer(d)  # works in a notebook

DecayChain D*

The actual graph is available as

# ...
dcv = DecayChainViewer(d)
dcv.graph

making all graphviz.Digraph class properties and methods available, such as

dcv.graph.render(filename='mygraph', format='pdf', view=True, cleanup=True)

In the same way, all graphviz.Digraph class attributes are settable upon instantiation of DecayChainViewer:

dcv = DecayChainViewer(chain, name='TEST', format='pdf')

Universal representation of decay chains

A series of classes and methods have been designed to provide universal representations of particle decay chains of any complexity, and to provide the ability to convert between these representations. Specifically, class- and dictionary-based representations have been implemented.

An example of a class-based representation of a decay chain is the following:

>>> from decaylanguage import DaughtersDict, DecayMode, DecayChain
>>>
>>> dm1 = DecayMode(0.0124, 'K_S0 pi0', model='PHSP')
>>> dm2 = DecayMode(0.692, 'pi+ pi-')
>>> dm3 = DecayMode(0.98823, 'gamma gamma')
>>> dc = DecayChain('D0', {'D0':dm1, 'K_S0':dm2, 'pi0':dm3})
>>> dc
<DecayChain: D0 -> K_S0 pi0 (2 sub-decays), BF=0.0124>

Decay chains can be visualised with the DecayChainViewer class making use of the dictionary representation dc.to_dict(), which is the simple representation understood by DecayChainViewer, as see above:

DecayChainViewer(dc.to_dict())

The fact that 2 representations of particle decay chains are provided ensures the following:

  1. Human-readable (class) and computer-efficient (dictionary) alternatives.
  2. Flexibility for parsing, manipulation and storage of decay chain information.

Decay modeling

The most common way to create a decay chain is to read in an AmpGen style syntax from a file or a string. You can use:

from decaylanguage.modeling import AmplitudeChain
lines, parameters, constants, states = AmplitudeChain.read_ampgen(text='''
EventType D0 K- pi+ pi+ pi-

D0[D]{K*(892)bar0{K-,pi+},rho(770)0{pi+,pi-}}                            0 1 0.1 0 1 0.1

K(1460)bar-_mass  0 1460 1
K(1460)bar-_width 0  250 1

a(1)(1260)+::Spline::Min 0.18412
a(1)(1260)+::Spline::Max 1.86869
a(1)(1260)+::Spline::N 34
''')

Here, lines will be a list of AmplitudeChain lines (pretty print supported in Jupyter notebooks), parameters will be a table of parameters (ranged parameters not yet supported), constants will be a table of constants, and states will be the list of known states (EventType).

Converters

You can output to a format (currently only GooFit supported, feel free to make a PR to add more). Use a subclass of DecayChain, in this case, GooFitChain. To use the GooFit output, type from the shell:

python -m decaylanguage -G goofit myinput.opts

Contributors

We hereby acknowledge the contributors that made this project possible (emoji key):

Eduardo Rodrigues
Eduardo Rodrigues

🚧 πŸ’» πŸ“–
Henry Schreiner
Henry Schreiner

🚧 πŸ’» πŸ“–
Yipeng Sun
Yipeng Sun

πŸ’»
Chris Burr
Chris Burr

πŸ“–
Kilian Lieret
Kilian Lieret

πŸ“–
Moritz Bauer
Moritz Bauer

πŸ’»
FlorianReiss
FlorianReiss

πŸ’» πŸ“–
Adam Morris
Adam Morris

πŸ’»
Dazhi Wang
Dazhi Wang

πŸ’»
Manuel Franco Sevilla
Manuel Franco Sevilla

πŸ’»

This project follows the all-contributors specification.

Acknowledgements

The UK Science and Technology Facilities Council (STFC) and the University of Liverpool provide funding for Eduardo Rodrigues (2020-) to work on this project part-time.

Support for this work was provided by the National Science Foundation cooperative agreement OAC-1450377 (DIANA/HEP) in 2016-2019 and has been provided by OAC-1836650 (IRIS-HEP) since 2019. Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.