The term “super” appears in various programming contexts, each with its own specific meaning. It’s crucial to understand the context to grasp its function accurately.This article will explore the use of “super” in Java generics and general object-oriented programming, as well as a potential issue encountered in Python when using Scikit-learn and XGBoost.

‘super’ in Java

In Java, “super” plays a vital role in inheritance, particularly when dealing with constructors and generics.

When a subclass constructor needs to call its parent class’s constructor, it uses the super() keyword [3].If you don’t explicitly call a super constructor, Java will automatically invoke the no-argument super constructor. Though, using super() becomes essential when the parent class’s constructor requires arguments from the subclass [3].

In Java generics, super is used to define lower bounds for type parameters. Such as, List<? super Suit> indicates a list that can hold objects of type Suit or any of its superclasses [1]. This is particularly useful when you intend to write into the list, ensuring that the objects you add are compatible with the list’s type [1].

“When you put an Object to the List, all you care about is that the object is of a type that is compatible with type held by the list.”

‘super’ in Python (Scikit-learn/XGBoost)

In Python, specifically within the context of scikit-learn and XGBoost, a “'super' object has no attribute '__sklearn_tags__'” error can arise [2]. This issue often surfaces when invoking the fit method on a RandomizedSearchCV object [2].

While the exact cause can vary, it’s frequently enough attributed to compatibility problems between Scikit-learn and XGBoost, or potentially Python version discrepancies. one user reported this using Python 3.12 with the latest versions of both Scikit-learn and XGBoost [2]. Resolving this may involve checking version compatibility and ensuring all libraries are correctly installed.