Refactor `OCPBalance` to align with `DataFit`

[brosaplanella]

At the moment OCPBalance is very different from DataFit which makes it hard to maintain. It should be reformatted so it is more similar (e.g. in the use of the parameters_optimise, cost_function_parameters, etc.).

[PipGrylls]

Hi, Ahead of our meeting on Wednesday I've started looking at this as it looked like a good way to get to know the codebase and use the examples.

[brosaplanella]

Just realised that many things will change once #32 is merged, so might be better to do that first. If you want, you can review that PR first. I have now added you to the team so you should be able to leave a review.

[PipGrylls]

Ah thanks for the heads up I will have a look

[PipGrylls]

Changes pulled from main (post main updated with #32) https://github.com/paramm-team/pybamm-param/commit/e3e20ff5388a8e000baaa0b8669930d1f47e1264

[PipGrylls]

In the base class __init__ we define the following variables:

def __init__(self):

        """
        Initialize the class properties

        x0 : array-like
            Initial guess for the optimization problem
        bounds : tuple
            bounds of the optimization problem
        scalings : array-like
            scalings for the optimization problem
        """
        self.x0 = None
        self.bounds = None
        self.scalings = None

OCPBalance does use super().init(), data fit does not.

We then have some disconnect between the names of input data and reference data.

I will try and pull these things together, to fulfil the brief of this issue but additionally I think there needs to be some strict templating applied to how further optimisation problems are defined.

We can choose the direction the templating goes in but making it permissive will make long term upkeep harder due to edge cases.

[brosaplanella]

I agree. I am away this week, but will think about this and we discuss when I am back.

[brosaplanella]

I have revisited the two classes (plus the new GITT one we are starting to develop) and they all have the following common features. The first column are the abstract concepts needed to define the class, while the value in each cell is how these are passed/defined for the optimisation problem of the respective column.

	DataFit	OCPBalance	GITT
model	simulation	data_fit	choose from given
data	data	data_ref	data
parameters	model_parameters	fixed	fixed
variables fit	variables_optimise	fixed	fixed
cost function	cost_function	cost_function	cost_function

The various classes might have inputs that are specific to them (e.g. GITT will require some parameters).

Then, the __init__ would have to:

• Process/assemble model
• Process/clean data
• Assemble parameters (including those from cost function)

The specific methods will be different for each class, but we could just call process_model, etc. from the base class and define what the function does in each subclass.

[PipGrylls]

Thats really useful thanks.

I think there is some argument to include some sphinx docs (new issue needed for this), pybamm uses them so it would be consistent.

[brosaplanella]

There are some docs here (https://pybamm-param.readthedocs.io/en/latest/?badge=latest) but there is a lot that can be improved.

[PipGrylls]

Gitt pulled from branch 26-implement-gitt, then updated to make naming self similar. This commit, 2335211, supersedes PR #45