Installation

Basic install

pip install rad_point_kernel

This gives you everything for point-kernel calculations, build-up factor interpolation with Gaussian Process extrapolation (inference-tools), and plotting.

OpenMC for Monte Carlo build-up factors

Linux and macOS only

OpenMC is currently available for Linux and macOS; there are no supported Windows builds. Rad Point Kernel itself works on Windows, but the Monte Carlo build-up features require OpenMC.

To compute build-up factors from Monte Carlo simulation, you also need OpenMC. OpenMC must be installed separately:

Pre-built wheels (easiest):

python -m pip install --extra-index-url https://shimwell.github.io/wheels openmc

Or follow the official guide:

See the OpenMC installation guide.

OpenMC cross sections

OpenMC needs nuclear cross section data. If you don't already have it, one easy way to download ENDF/B-VIII.1 is:

python -m pip install openmc_data
download_endf -r b8.1

Other options and pre-packaged libraries are listed on openmc.org.

Once the data is on disk, point OpenMC at it one of two ways:

Environment variable:

Replace "/path/to/cross_sections.xml" with your actual path to where your cross_section.xml is located.

export OPENMC_CROSS_SECTIONS=/path/to/cross_sections.xml

Pass the path directly:

results = rpk.compute_buildup(
    geometries=[layers],
    source=source,
    quantities=["dose-AP"],
    cross_sections="/path/to/cross_sections.xml",
)