What’s the "@" in python for?

Question

Several times I came across snippets of python code that used "@" as in flask:

import flask
App = flask.Flask(__name__)
@App.route("/")

Why not just App.route?

Answer 1

The "@" prefix of an expression that is a line above a function or class statement is a "decorator" (Decorator) . What they are and how decorators work is explained in the question that was marked as a suggestion that this would be a duplicate: How Python Decorators Work?

@ as a decorator:

@modificar_funcao
def minha_funcao():
    ...

However, I reopened the question, why is there another use of @that is not addressed in that question: it can be used with an operator that uses two operands (as well as + - * /) - and in this context, it indicates a "multiplication of matrices". The matrix multiplication operation is not defined in any type that accompanies a standard Python instance, so it is little known - but the "numpy", the main library with Python matrix operations uses the "@" in that sense. Also, any class created in your program can define an operation with "@" as an operator if you implement the method __matmul__.

@ as a matrix multiplication operator:

In [7]: import numpy as np                                                                                  

In [8]: A = np.ones((2,2))                                                                                  

In [9]: B = np.ones((2,1))                                                                                  

In [12]: A[0,0] = 3                                                                                         

In [13]: B[0, 0] = 5                                                                                        

In [14]: A                                                                                                  
Out[14]: 
array([[3., 1.],
       [1., 1.]])

In [15]: B                                                                                                  
Out[15]: 
array([[5.],
       [1.]])

In [16]: A @ B                                                                                              
Out[16]: 
array([[16.],
       [ 6.]])