Installation

End user installation

MCIO can be installed from source using the below commands. The cad and jupyter extras install the dependencies for CAD-to-CSG decomposition and Jupyter; they are not needed for building CSG models directly.

git clone https://github.com/ukaea/mcio.git

cd mcio

python -m pip install .[cad,jupyter]

Dev installation

For those wishing to contribute to development, install the dependencies for testing, linting and building these docs.

git clone -b dev https://github.com/ukaea/mcio.git

cd mcio

python -m pip install .[dev,docs,cad,jupyter]

The docs can then be built locally with:

python -m sphinx -b html docs/source docs/build/html

OpenMC

Writing OpenMC xml fires requires the instllation of the OpenMC Python API, instructions fo rthe install of which can be found at OpenMC installation guide. For developers, the test suite requires it and the CI installs it with:

python -m pip install "openmc @ git+https://github.com/openmc-dev/openmc.git"

Conda installation for CadQuery-OCP

Some operating systems don’t have pre-compiled wheels for the CadQuery-OCP binaries, such as on clusters with older versions of Linux. Conda can be installed without the need for super user access, as below:

# create a directory
mkdir -p ~/software/conda && cd ~/software/conda

# pull the mini-conda installer
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh

# run the installer - provide path when requested (/lustre/home/<USERNAME>/software/conda/miniconda3)
bash Miniconda3-latest-Linux-x86_64.sh

Add the below to your .bashrc file, replacing your user name with the string.

# to add to your .bashrc, swapping out <USERNAME> for your user name

function load_conda {
    echo "loading conda"
    declare -a arr=(
        "GCC/13.3.0"
    )
    deactivate
    module purge
    for i in "${arr[@]}"; do
        module add "${i}" -s
    done
    # >>> conda initialize >>>
    # !! Contents within this block are managed by 'conda init' !!
    __conda_setup="$('/lustre/home/<USERNAME>/software/conda/miniconda3/bin/conda' 'shell.bash' 'hook' 2> /dev/null)"
    if [ $? -eq 0 ]; then
        eval "$__conda_setup"
    else
        if [ -f "/lustre/home/<USERNAME>/software/conda/miniconda3/etc/profile.d/conda.sh" ]; then
            . "/lustre/home/<USERNAME>/software/conda/miniconda3/etc/profile.d/conda.sh"
        else
            export PATH="/lustre/home/<USERNAME>/software/conda/miniconda3/bin:$PATH"
        fi
    fi
    unset __conda_setup
    # <<< conda initialize <<<
    }
export -f load_conda

Now install the cadquery-ocp package in conda

# reload your .bashrc file
. ~/.bashrc 

# load the base conda environment
load_conda

# install cadquery and ocp
conda create --name ocp -c conda-forge cadquery ocp python=3.13 pip -y

# activate the ocp conda environment
conda activate ocp

Once in the Conda environment, you can install MCIO as described above in the installation section.