Repository navigation
Docker
Salvador Navas edited this page Jul 2, 2026
·
1 revision
The repository includes a Docker-based JupyterLab environment for running notebooks with the scientific Python stack.
- Docker Desktop on macOS/Windows or Docker Engine on Linux.
- Docker Compose v2.
- Internet access during the first build.
Check:
docker --version
docker compose versiongit clone https://github.com/navass11/pyhydra.git
cd pyhydra
docker compose -f docker/docker-compose.yml up --buildOpen:
http://localhost:8888
JupyterLab is configured without token or password for local development. Do not expose this container directly to the internet.
| Host path | Container path | Purpose |
|---|---|---|
./notebooks |
/pyhydra/notebooks |
Editable notebooks. |
./data |
/pyhydra/data |
Data workspace for downloads, model inputs and generated outputs. |
docker compose -f docker/docker-compose.yml exec jupyter pyhydra-get-data
docker compose -f docker/docker-compose.yml exec jupyter pyhydra-get-data manning_rugosidadesDownloaded files are written under /pyhydra/data/pilot_cases/..., mapped to
./data/pilot_cases/... on the host.
Stop:
docker compose -f docker/docker-compose.yml downRebuild:
docker compose -f docker/docker-compose.yml build --no-cache
docker compose -f docker/docker-compose.yml upRun in background:
docker compose -f docker/docker-compose.yml up --build -dLogs:
docker compose -f docker/docker-compose.yml logs -f jupyterThe Docker image provides the Python/Jupyter environment. It does not install commercial, Windows-specific or project-specific model runtimes such as HEC-HMS, HEC-RAS, SWAT+ or SFINCS project binaries. Those workflows may still require local model installations and mounted project folders.