sarpyx is a specialized Python toolkit for Synthetic Aperture Radar (SAR) processing with tight integration to ESA SNAP. It focuses on reproducible pipelines, fast tiling workflows, and advanced research features like sub-aperture decomposition.

• SNAP GPT integration with configurable graphs and operator chaining.
• Sub-aperture decomposition for squint-angle diversity and motion sensitivity.
• Parallel tiling and batch processing for large product volumes.
• Geocoded outputs ready for GIS and downstream ML.
• Extensible architecture compatible with rasterio, geopandas, and pyproj.

For container workflows, use the Docker Compose CLI plugin (docker compose) with full commands:

docker compose version
make recreate

uv sync --extra copernicus

uv sync --extra copernicus --extra dev --extra test --extra docs

python -m pip install -e .

See docs/user_manual.md for full CLI usage and end-to-end workflows.

At startup the container checks for grid files in this order:

1. GRID_PATH (or grid_path) if it points to an existing *.geojson
2. First *.geojson found in /workspace/grid
3. Generate a grid on startup only if no *.geojson is available

To use a mounted grid:

mkdir -p ./grid
# put any grid GeoJSON here, e.g. ./grid/my_region.geojson
docker compose up

For direct docker run, pass an explicit in-container path when needed:

export GRID_PATH=/workspace/grid/my_region.geojson
docker run --rm -e GRID_PATH=$GRID_PATH ...

You can also pass --grid-path to the worldsar CLI command.