Installation

Command-line installation (pip)

Install PG-SUI from PyPI inside a fresh virtual environment so dependencies stay isolated:

python3 -m venv pg-sui-env
source pg-sui-env/bin/activate

pip install pg-sui

# sanity check
pg-sui --help

PG-SUI pulls in SNPio and all required ML/plotting dependencies. The CLI works anywhere Python 3.11+ is available.

Anaconda installation

If you prefer using Anaconda/Miniconda, install PG-SUI from the btmartin721 channel:

conda create -n pg-sui-env -c btmartin721 pg-sui
conda activate pg-sui-env

# sanity check
pg-sui --help

MacOS GUI add-on (Electron)

If you have MacOS and want a point-and-click interface, install the optional Electron wrapper. It shells out to the same CLI under the hood, so presets, YAML configs, and overrides behave identically.

# install the Python package with GUI extras (FastAPI/uvicorn helper)
pip install "pg-sui[gui]"

# one-time setup to fetch the Electron app dependencies
pgsui-gui-setup

# launch the desktop app
pgsui-gui

Node.js (with npm) is required for the Electron app. The GUI respects the active Python environment, making it easy to reuse the same configs you run via pg-sui.

Development Installation

If you want to contribute to PG-SUI, install it in development mode so changes are picked up immediately.

git clone https://github.com/btmartin721/PG-SUI.git
cd PG-SUI

python3 -m venv pg-sui-env
source pg-sui-env/bin/activate

pip3 install -e ".[dev]"

For GUI development, append [gui] to the editable install and rerun pgsui-gui-setup inside the repo.

Note

If you are using a Jupyter notebook, you will need to restart the kernel after installing PG-SUI in development mode.