The super() function in Python is a built-in that allows you to call methods from a parent class. It’s particularly useful in scenarios involving inheritance, especially multiple inheritance, to avoid explicitly naming the parent class [[1]].

How super() Works

In single inheritance, super() simplifies calling a parent’s method. However, its power truly shines with multiple inheritance. When dealing with multiple inheritance, Python uses Method Resolution Order (MRO) to determine the order in which parent classes are searched for a method [[1]].

“Now call to super in init defined in First will continue searching MRO and find init defined in Second, and any call to super will hit the default object init.” [[1]]

Consider this example:

            
class First:
    def __init__(self):
        print("First init")
        super().__init__()

class Second:
    def __init__(self):
        print("Second init")
        super().__init__()

class Combined(First,Second):
    def __init__(self):
        print("Combined init")
        super().__init__()

instance = Combined()
            
        

In this case, the MRO would be [combined, First, Second, object]. When super().__init__() is called in Combined, it calls First.__init__(). Then, super().__init__() in First calls Second.__init__(). super().__init__() in Second calls the object class’s init method [[1]].

Python 2 vs. Python 3

In Python 2, using super() required specifying the class and instance explicitly, which could be confusing. Python 3 simplifies this by allowing super() to be called without arguments [[2]].

            
# Python 2
super(CurrentClass, self).__init__()

# Python 3
super().__init__()
            
        

Common Issues

One common issue is encountering the error 'super' object has no attribute '__sklearn_tags__'. This can occur when using libraries like Scikit-learn and XGBoost, possibly due to compatibility issues or Python version. Ensuring you have the latest versions of these libraries might resolve the problem [[3]].