Adding a new function that runs the setup of structure S

[swilliamson7]

If you're not opposed to this I added a function that sets up and returns S but doesn't do any integrating, I use it to start my Enzyme runs

[milankl]

Sounds great, yeah I mean this is where the run_model interface is falling apart because as you clearly say one does need a way to set all parameters for a model configuration, then have access to the (i.e. one single) model/simulation object and then a way to run that simulation forward. That's why we do that in SpeedyWeather. Glad you're suggesting the same!

[milankl]

What about calling run_setup instead model_setup as it creates a ModelSetup and returns it?

[swilliamson7]

Sounds great, easy to remember! I'll merge it after the checks complete