| File | What it shows |
|---|---|
../src/geoloupe/gallery.yaml via geoloupe gallery |
The gallery: ERA5, Copernicus DEM, Sentinel-2, ESA WorldCover and NOAA stations, all read live from public buckets, with featured views. The best place to start, and a template for your own gallery (geoloupe gallery --show-config) |
../src/geoloupe/demo.py via geoloupe demo-data DIR |
Synthetic data (1° global + 0.01° Alps "downscaled", pressure levels, a LAEA grid) and a config: the quickest way to see every feature |
fetch_public_samples.py |
Three small public datasets (ERA5 GRIB, NCEP NetCDF, ERSST NetCDF) and a config for them |
sentinel2.yaml |
A Sentinel-2 L2A scene read over HTTPS from the public AWS bucket: COG overviews as resolution levels, reflectance scale, true- and false-colour images |
public-s3.yaml |
Data from public AWS S3 buckets with s3:// paths and no setup: a CMIP6 climate model ocean (Zarr) and a Copernicus DEM tile (COG) |
arco-era5.yaml |
Hourly ERA5 since 1940 from Google's public ARCO-ERA5 Zarr store, read over HTTPS (761 160 hourly steps with data) |
fetch_ncei_stations.py |
Real station records (NOAA GHCN-Daily) as a point dataset: columns detected, units set in the config |
notebook_quickstart.py |
geoloupe.show() in a notebook: one dataset, a multi-resolution layer, and the API from Python (open as a notebook in VS Code/JupyterLab) |
api_client.py |
Using a running server from any Python with numpy + httpx: point values, time series, decoding a data tile |
hpc/build_overviews.sbatch |
Building overviews of a large store on an HPC cluster with a Slurm array (one year per task) |
Run any configuration with geoloupe serve <config.yaml> and check it first with
geoloupe info <config.yaml>.