nbscuid-adf-sandbox 🦑☁️

How to run

1. Install an environment that contains nbscuid, e.g. with the instructions here
2. Activate this environment (e.g. conda activate nbscuid-dev)
3. Clone this repo, which must be done recursively in order to clone the associated submodules:

git clone --recurse-submodules https://github.com/rmshkv/nbscuid-adf-sandbox.git

4. Navigate to the top-level folder, nbscuid-adf-sandbox

5. Install the environment specified by mom6-environment.yml: mamba env create -f mom6-environment.yml, or if you don't have mamba installed, conda env create -f mom6-environment.yml. (It'll probably be a lot faster with mamba!)

   NOTE: This will create an environment called mom6-tools-nbscuid. There's no need to activate it--it'll just be used when nbscuid runs the surface notebook, since it has mom6-tools-nbscuid specified as its kernel_name in config.yml.

6. Run nbscuid-run config.yml, making sure you're still in an environment with nbscuid installed

   NOTE: You can actually run this command from anywhere, just change config.yml to the full path relative to your location.

7. Run nbscuid-build config.yml

   NOTE: The built Jupyter Book files will appear in computed_notebooks/adf-quick-run/_build, try opening index.html!

How to add stuff

Check out the config.yml file. This page has some instructions for how to add new notebooks and their parameters, and this is an example of a config.yml file that uses some more options, including adding scripts.