Are there any performance overheads if we specify Pyomo Constraints in different way?

Question

I am interested to know whether there are any performance overheads for specifying Pyomo Constraints in different ways. For example, which of the two ways is better? I am trying to speed up one of my models who uses the second option. Can someone provide more information?

Option 1:

def teaOKrule(model):
    return(model.x['butter'] + model.x['scones'] == 3)
model.TeaConst = Constraint(rule=teaOKrule)

Option 2:

@model.Constraint(model.s)
def hello_wor(model, i, j):
    return model.s[i, j] <= model.b[i, j]

Answer 1

Short answer: To within the tolerances measurable within Python, those two approaches perform equivalently.

Longer answer: there is in fact a slight overhead using the decorator notation instead of the explicit notation (the decorator effectively generates the explicit form automatically for you). This amounts to only a handful of function calls and a little bit of logic. The thing to remember is that overhead is only encountered once when declaring the component (there is NO additional overhead for each index within the Component), so is for all intents and purposes not measurable.