Aare Oraku

I bi vou binech das momentan z viu AI-Hype ume isch aber weisch wi geil we dr Aare Guru scho paar Täg im vorus ungfähr wüsst wi warm d Aare de isch u wiviu Wasser si de het?

Namensidee: Aare Oraku
Logoidee: Wätterfrosch wo uf sire Leitere steit und id witi luegt aus wärs ä Schiffsusguck während er mit sim Glas d Aare ab tribt <- di Idee hani vorem Name gha

Interessanterwiis gits scho ä Vorhersag zum Wasserabfluss (d Wassermängi im Aare.guru) vom Bund, i nime a di isch scho sehr guet u chönnt direkt verwendet werde. Übrig blibt auso d Temperaturvorhersag, wo vermuetlech eifacher isch aus d Abflussvorhersag, drum probiere mrs eis.

Setup

Relies on uv and DVC. Uses just to simplify commands. Only tested on Linux with bash and zsh.

First install uv, then use it to install just with:

uv tool install rust-just

Then creating the Python environment is as simple as

just sync

To run the pipelines, do this. These currently just calculate the baselines.

just repro

To run the notebooks, make sure to set the kernel to the lokal venv (.venv/bin/python) in your IDE. In vscode, when it asks after first time execution, set the environment to the recommended ".venv" env.

If you want to use JupyterLab, you need to install the kernelspec first: just install-kernel
Then run jupyter lab. You'll need to change the kernel in the top right for every notebook.

To start the mlflow ui and the optuna dashboard for experiment tracking and hyperparameter tuning, use the following commands in separate terminals:

just mlflow
just optuna

Or use

just track

If you want to work with the code instead of just running it, I suggest you install nbstripout (already in venv) and pre-commit (uv tool install pre-commit). To set them up, run the following.

uv run nbstripout --install --attributes .gitattributes
pre-commit install --install-hooks

Quellen & Links

D Date chöme vom aare.guru säuber oder vom BAFU, aggregiert im wahnsinnig tolle Archiv vom Bureau für digitale Existenz.

• Aare.guru API Description
• InfluxDB Access ❤️
• SwissMetNet Dataset Description
• BAFU Hydrologie Dataset Description
• Hydrologische Vorhersagen vom Bund (nüt dopplet mache)

Potenzielle Ressourcen

• http://www.watercenter.org/physical-water-quality-parameters/water-temperature/water-temperature-ranges-in-rivers-and-streams/
• https://storymaps.arcgis.com/stories/ef8b1542a6c0411ba96779b26151e399
• https://pmc.ncbi.nlm.nih.gov/articles/PMC5994338/
• https://agupubs.onlinelibrary.wiley.com/doi/10.1029/98WR01877
• https://hess.copernicus.org/articles/25/2951/2021/
• https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1365-2427.2006.01597.x

Lizenz

AGPL-3.0

(für aare.guru müesst das nid kümmere, i mache ds hie houptsächläch füre Spass u fänds riise geil 's i aare.guru z integriere wes de ändlech fr öppis isch)