Getting started
The website runs on the data already in the repository, so you need no NASA account to try it. Rebuilding the data from NISAR is a separate step, covered in Pipeline.
What you need#
| For | Install |
|---|---|
| The website | Node.js 20.19 or later (22 LTS recommended), with npm |
| The pipeline | Miniconda or Anaconda, for the sartale environment in environment.yml (Python 3.12, NumPy, SciPy, h5py, rasterio, GDAL, pyproj, Pillow, requests, PyYAML, pytest) |
| Rebuilding data from NISAR | A free NASA Earthdata login, saved in ~/.netrc |
Run the website#
git clone https://github.com/ML-Dev-Hub/SARtale.git
cd SARtale
npm --prefix app install
npm --prefix app run dev # opens on http://localhost:5173
The site opens on the story page. Use the header to switch to Lake Watch and How we know.
Run the tests#
npm --prefix app test -- --run # app tests, including the wording check
npm --prefix app run build # type-check, then build into app/dist/
For the Python side, create the environment once, then run pytest from processing/:
conda env create -f environment.yml
cd processing
conda run -n sartale pytest tests -q
Both suites run offline. The Python tests use small synthetic NISAR files built in tests/gcov_fixture.py, not real downloads.
Try the "Add a lake" form#
The form is in the story page's Add a lake section, and its button is Request a lake. It needs no server. It checks your input in the browser, then opens GitHub's "Add a lake" issue form with the fields filled in, so you can only submit it from a copy with access to the repository. Nothing is saved locally. See Lake requests.
Port already in use? If 5173 is taken, another copy is probably running. Stop it, or pass -- --port 5174 to npm run dev.
SARtale Guide