Usage Guides
Common Workflows
L0 to Focused Product
sarpyx unzip --input /data/raw.zip --output /data/extracted
sarpyx decode --input /data/extracted/S1A_*.SAFE --output /data/decoded
sarpyx focus --input /data/decoded/product.zarr --output /data/focused
Mission Pipeline and Tiling
sarpyx worldsar --input /data/product.SAFE --output /data/preprocessed --cuts-outdir /data/tiles --grid-path /data/grid/grid_10km.geojson
Advanced Usage
- Use
--gpt-memory, --gpt-parallelism, and --gpt-timeout to tune SNAP runtime.
- Use
ProductHandler and ZarrManager APIs for selective slicing and export.
- Use mission-specific worldsar pipelines for Sentinel, TerraSAR-X, COSMO, BIOMASS, and NISAR paths (inferred from implementation).
Integration Patterns
- Pipeline orchestration through shell scripts or workflow managers invoking CLI commands.
- Python-first integration importing processing modules directly for notebook or service use.
- Containerized integration via
docker compose for reproducible environments.
- Enable slicing for large products to reduce memory pressure during focusing.
- Prefer Zarr chunking defaults unless profiling shows bottlenecks.
- Adjust GPT heap and worker count based on available host memory.
- Use parallel tile extraction settings conservatively to avoid I/O saturation.