Difference between setting current through parameters versus through experiment

[SimMLPhys]

Between using Current Function in parameters and defining it through pybamm.experiment, is there a preferred way to set the current if my simulation's input is the current profile?

I'm also trying to define a current profile which uses both the string-based instructions (Pybamm Tutorial 5) and appends a custom current function (like at the end of Tutorial 4) at the end of the string-based instructions. Is that possible or am I mixing two different things?

[valentinsulzer]

Using the experiment is slightly better as you get access to summary variables. You can mix a custom current function and string based instructions

[SimMLPhys]

I'm having trouble figuring out how to combine the custom current function with the string based instructions. For instance, this code and some other variations of it I tried keep giving me one or the other kind of a compatibility error:

def create_custom_experiment():
    def my_fun(A, constant):
        def current(t):
            sine_terms = A * pybamm.sin(2 * np.pi * t) 
            return sine_terms + constant
        return current

    experiment_starting = pybamm.Experiment(
        [
            ("Discharge at C/10 for 3 hours or until 3.3 V", "Rest for 1 hour"),
        ] * 3
    )
    experiment_ending = my_fun(5, 0.1)
    return experiment_starting + pybamm.Experiment(experiment_ending)