-
Notifications
You must be signed in to change notification settings - Fork 3
Description
Previously when simulating with a ModelList we could retreive the model from the models argument to the run function by directly indexing inside it, e.g.: models.energy_balance, and get access to the instantiated model (with its parameters if any). But this is different in a multiscale setup, where we index first by scale, e.g. models["Leaf"].energy_balance. This means that models have different ways of accessing this information depending on if it's applied in a single or multiscale setup. This is bad design.
So what we should do instead is tell users to enforce the multiscale setup, and indexing with the default scale value (not sure which one, maybe "Default", or "single-scale") defined when we take the single-scale pathway. Then to put this parameter as a model parameter so users can change it.