Installation Instructions

Conda Environment

1. Create a conda environment with python 3.13

conda create -n pyxalign python=3.13

then activate the environment

conda activate pyxalign

Install astra-toolbox and CuPy from conda-forge:

conda install -c conda-forge astra-toolbox
conda install -c conda-forge cupy

Installing these libraries from conda-forge instead of pip ensures that required CUDA toolkit is installed.

Create an editable, developer installation of pyxalign:

git clone https://github.com/AdvancedPhotonSource/pyxalign.git
cd pyxalign
pip install -e ".[dev]"

Install the latest version of pyxalign from PyPI:

pip install pyxalign

Podman Container

1. Clone the pyxalign git repo

git clone https://github.com/AdvancedPhotonSource/pyxalign.git
cd pyxalign

2. Build the container image

podman build -t pyxalign:latest .

3. Run the container

podman run \
    -it --rm --env DISPLAY --security-opt label=type:container_runtime_t \
    --network host -v="$HOME/.Xauthority:/root/.Xauthority:rw" \
    --device nvidia.com/gpu=all \
    pyxalign:latest bash

Any directories you need access to inside the container must be explicitly mounted using the -v flag:

-v /path/on/host:/path/in/container

For example, to mount your data directory:

podman run \
    -it --rm --env DISPLAY --security-opt label=type:container_runtime_t \
    --network host -v="$HOME/.Xauthority:/root/.Xauthority:rw" \
    -v /data/my_experiment:/data/my_experiment \
    --device nvidia.com/gpu=all \
    pyxalign:latest bash