In the case of a CallableNumericalModel with mixed analytical and non-analytical components, the connectivity mapping corresponding to the analytical components can just be infered automatically. This is a lot easier on the eyes:
x, y, z = variables('x, y, z')
a, b = parameters('a, b')
model_dict = {z: lambda y, a, b: a * y + b,
y: x ** a}
mixed_model = CallableNumericalModel(
model_dict, connectivity_mapping={z: {y, a, b}}
)
Giving the full connectivity mapping however, is also allowed.
In the case of a CallableNumericalModel with mixed analytical and non-analytical components, the connectivity mapping corresponding to the analytical components can just be infered automatically. This is a lot easier on the eyes:
Giving the full connectivity mapping however, is also allowed.