The main end-users of the Annotated Atlas are both the operational and the research meteorologists. The operational meteorologists will use the atlas for analyzing the weather warning(CAP) proposal by reviewing historical weather warnings, but also to review earlier situations and give the earlier warnings hit rate grades. For the research meteorologist, the atlas will be a convenient tool for identifying good training data for machine learning algorithms by giving easy access to source observation datasets.
Hans Christian Nenseth ([email protected]) is responsible for this repository.
There is a room in google chat called WeaMyL-chatroom where we discuss development. If you need access to the group, contact [email protected]
We recommend using the latest stable version of node and npm in this project. If you want to test it without development tools, see description on how to run with docker below.
If you need to run against another backend/database you have set up on your own system you have to edit
the .env
-file to point to the right address.
Installs all dependencies (based on package-lock.json).
Runs the app in the development mode.
Open http://localhost:8080 to view it in the browser.
The page will reload if you make edits.
You will also see any lint errors in the console.
If you have docker you can build and run the app without any other dependencies this way.
docker build -t annotated-atlas .
docker run --name annotaded-atlas -p 8080:8080 annotated-atlas
The docker image is suitable for production in k8s.met.no.
You will need to fork this repo and make your own. Further we reccomend:
- Add appropriate license.
If you have browser support issues, consider adding Babel
If you find bugs or other issues, please make merge requests.
- Test format with
npm run format:check
- Make sure test pass with
npm run test
- Ensure that CI pipeline does not fail
- Additional jest tests are very welcome