How is numpy and pandas library able to change what print function does

Every full-featured data type has conversion functions for conversion to string, __str__, and output, __repr__. Both PANDAS and NumPy have implemented those methods as part of each data type's class definition -- those control how the data types appear.

You can look up details in any mid-level tutorial on classes.

Prune's answer is correct, but I want to make it more concrete with an example. The print function actually outputs whatever is returned by the instance's class's __repr__ magic method example: let's say you have a Person class:

class Person:
    pass

p = Person()

print(p)

The result will be something like:

<__main__.Person object at 0x000000D9D5034508>

Now let's implement the __repr__ magic method:

class Person:
    def __init__(self, name):
        self.name = name

    def __repr__(self):
        return f"< Person Name: {self.name} >"

p = Person("Robert")

print(p)

I also added the constructor to demonstrate how one may use it. The result will be:

< Person Name: Robert >

And that's how Numpy and Pandas are able to display the array and Dataframes like that.