FUDGE (For Updating Data and Generating Evaluations)

FUDGE is a nuclear data package built around Python that supports reading, viewing, modifying, writing and processing nuclear data in the GNDS format.

The current release of FUDGE focuses on supporting version 2.0 of the Generalized Nuclear Database Structure (GNDS). The FUDGE package includes tools to translate other nuclear data formats to and from GNDS, plus tools for testing, visualizing, and processing GNDS data.

Contents of this README:

• Getting Started
• Basic Use (translating ENDF files)
• Getting Help

Getting Started:

Installing FUDGE only requires Python (version 3.6 or higher) and NumPy (version 1.15 or higher). Optional packages matplotlib and PyQT5 are also recommended to support plotting.

The following installation options are currently supported:

• Installation via pip install: This is ideal for use in virtual environments. The following steps are recommended to pip install FUDGE:

  • Create a new virtual environment if desired. For example, using Anaconda:

    conda create --name fudge python=3.8 numpy matplotlib PyQT
    conda activate fudge
    

    Some users have reported problems with installing PyQT with conda on Windows. It may be ommitted from the above conda create command and, if required, be installed via pip install.

  • Ensure that NumPy (version 1.15 or higher) and wheels are present in the Python environment.

    python
    import numpy
    print( numpy.__version__ )
    import wheel
    exit()
    

  • Install FUDGE:

    pip install git+https://github.com/LLNL/[email protected]
    

• Installation by cloning the git repository and building with the unix make command: This is the typical mode for active FUDGE maintenance and development. The following steps are recommended:

  • Ensure that NumPy (version 1.15 or higher) is installed

  • Clone FUDGE in the current directory:

    git clone https://github.com/LLNL/fudge.git

    or using SSH (requires creating a github account and registering an ssh key):

    git clone [email protected]:LLNL/fudge.git

  • Build FUDGE:

    cd fudge; make -s
    

  • [Optional] Set environment variables. For general use of the FUDGE package, some changes should be made to your computer's environment. The following lines make the required change. Note that <path_to_FUDGE> indicates the path to the directory containing this README.md file.

    • on unix with the bash, ksh, or similar shell, put the following line in the .bashrc or equivalent file:

      export PYTHONPATH=$PYTHONPATH:<path_to_FUDGE>
      

    • on csh, tcsh or similar shell into the appropriate shell file:

      setenv PYTHONPATH $PYTHONPATH:<path_to_FUDGE>
      

    • on Windows, the environment variable should be added to the registry (see for example http://www.support.tabs3.com/main/R10463.htm)

Differences between FUDGE installed via pip install vs. via make:

When FUDGE is installed via pip install, several executable scripts are installed into the virtual environment 'bin' directory and can be used directly. The same behavior can be achieved using the make installation method by adding <path_to_FUDGE>/bin and <path_to_FUDGE>/brownies/bin to the PATH.

Basic Use:

After installing FUDGE, a user may be interested in translating an ENDF-6 file into a GNDS file and vice versa. Two tools for translating GNDS files to and from ENDF-6 files are included in the brownies/bin directory.

The Python script endf2gnds.py translations an ENDF-6 file to a GNDS file. This script writes a GNDS file containing the reactionSuite node and, if covariance data are present in the ENDF-6 file, its also writes a GNDS file containing the covarianceSuite node.

For example, the following command translating the ENDF/B-VIII.0 file n-001_H_001.endf and generates n-001_H_001.xml and n-001_H_001-covar.xml:

python3 -m brownies.bin.endf2gnds.py n-001_H_001.endf n-001_H_001.xml

# or, if FUDGE was installed via pip:
endf2gnds.py n-001_H_001.endf n-001_H_001.xml

Some ENDF-6 files fail to translate unless the --continuumSpectraFix option is included when calling endf2gnds. This option addresses a problem with unnormalizeable continuum photon spectra. It has no effect unless an evaluation contains an unnormalizeable spectrum so can be used safely for all translations. Sample use:

endf2gnds.py <originalFile.endf> <newFile.xml> --continuumSpectraFix

The Python script gnds2endf.py translates a GNDS file to an ENDF-6 file. This script will look for a corresponding covariance file and, if present, add its data to the ENDF-6 result. For example, the following translates the n-001_H_001.xml file generated in the endf2gnds.py comannd above back into an ENDF-6 file:

python3 -m brownies.bin.gnds2endf.py n-001_H_001.xml

# or,
gnds2endf.py n-001_H_001.xml

FUDGE scripts use the argparse.py module to parse input arguments. To get help on a script, run the script with the -h option. For example, to get help on the usage of the bin/processProtare.py script execute:

<path_to_FUDGE>/bin/processProtare.py -h

There are more scripts in the bin directory that a user may find useful. One can also run FUDGE interactively. The following Python commands show how to read in the n-001_H_001.xml generated by the example above, print a list of its reactions and plot each reaction's cross section on a single plot:

from fudge import GNDS_file     # FUDGE must be in the PYTHONPATH.
n_H1 = GNDS_file.read( 'n-001_H_001.xml' )
for reaction in n_H1.reactions: print( reaction )

crossSections = []
for reaction in n_H1.reactions:
    crossSection = reaction.crossSection.toPointwise_withLinearXYs( lowerEps = 1e-7 )
    crossSection.plotLegendKey = str( reaction )
    crossSections.append( crossSection )

crossSection.multiPlot( crossSections, rangeMin = 1e-5, xylog = 3,
        title = 'n + U230', xLabel = 'Neutron energy [eV]', yLabel = 'Cross section [b]' )

Note regarding HDF5 support:

FUDGE imports the h5py module in some places to read and write HDF5 files. In some circumstances h5py fails to import if MPI-related environment variables are set. This issue can cause the Python interpreter to crash. The best fix at the moment is to ensure that environment variable PMI_FD is not set before running anything requiring h5py.

Getting Help:

FUDGE documentation is currently undergoing an update but some documentation can be found by executing the following make command in the FUDGE directory (note this only works after cloning FUDGE, not with pip install):

make -s docs

To view the documentation, open the file doc/sphinx/_build/html/index.html inside the FUDGE directory.

For other questions about the FUDGE, please feel free to contact the maintainers at [email protected], [email protected], or [email protected].

License

FUDGE is distributed under the terms of the 3-clause BSD license.

All new contributions must be made under the 3-clause BSD license.

See LICENSE, COPYRIGHT and NOTICE for details.

SPDX-Licence-Identifier: BSD-3-Clause

This package includes several components:
LLNL-CODE-683960 (FUDGE)

LLNL-CODE-770134 (numericalFunctions
LLNL-CODE-771182 (statusMessageReporting)
LLNL-CODE-725546 (Merced)

FUDGE is a product of the Nuclear Data and Theory Group at Lawrence Livermore National Laboratory (LLNL).

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.