Getting started¶
SPM-Kit separates numerical work from presentation. Use the Python API or CLI for scripts, pipelines, and reproducible runs; use Fathom when visual inspection and interactive region selection are part of the work.
Choose a path¶
| Goal | Start here | First successful result |
|---|---|---|
| Install a specific release or development state | Installation | spmkit --version and spmkit --help succeed |
| Analyze a file or an in-memory surface | First analysis | roughness values with declared units and preprocessing |
| Explore data interactively | Fathom quick start | a channel is loaded into the Imagen perspective |
| Connect discovery, synthetic truth, and validation | Example workflows | a provenance-preserving artifact chain |
Requirements¶
- Python 3.11 or newer;
piporuv;- the
guiextra only for Fathom; - format-specific extras only where the format matrix says they are required.
Thirty-second source install¶
The site documents the 0.1.5.dev0 source tree. Install that state explicitly:
python -m pip install "spmkit[gui] @ git+https://github.com/kegouro/spmkit@main"
spmkit --version
spmkit --help
PyPI currently serves 0.1.2; the latest GitHub release is 0.1.4. The
installation matrix explains the difference.
Scientific checkpoint¶
Loading a file is not validation. Before using a result in a scientific claim, record the input checksum, reader path, channel, units, preprocessing, parameters, software version, and output checksum. Then match the claim to the scientific status matrix.