python class attribute editing

Question

i try to write a template class in python. my aim is importing this class for other projects. class have a class attribute M. it's default value is zero. when i import class in other project i change M with a new value. after that, i sampling class but instance attributes dont change. they consider zero value for M. how can i solve this problem? i dont want to use instance attribute for 'M'.

import numpy as np
class Example():
    M=0
    matrix=np.ones((M,M))
    def __init__(self,a,b):
        self.a=a
        self.b=b
        self.operate()
    def operate(self):
        Example.matrix=Example.matrix*self.a*self.b
        print(Example.matrix)
Example.M=5
m1=Example(3,5)

when i run the code, [] is the result. but i expect this matrix:

[[15. 15. 15. 15. 15.] [15. 15. 15. 15. 15.] [15. 15. 15. 15. 15.] [15. 15. 15. 15. 15.] [15. 15. 15. 15. 15.]]

Answer 1

This is because matrix is decrlared once, if you try to modify M, it doesn't change the matrix.

By changing the code a bit, this gives:

import numpy as np
class Example():
    def __init__(self,a,b):
        self.M=5
        self.matrix=np.ones((self.M,self.M))
        self.a=a
        self.b=b
        self.operate()
    def operate(self):
        self.matrix=self.matrix*self.a*self.b
        print(self.matrix)
Example.M=5
m1=Example(3,5)

Basically, I moved M and matrix to __init__, changed M to self.M and matrix to self.matrix.

Answer 2

You need to move the definition of your matrix inside the __init__ function - right now, the class is parsed, M is set to 0 and matrix is set to np.ones(0,0) which is an empty list []. Then, you assign 5 to M, but matrix remains unchanged.

So you need to either redefine matrix manually:

Example.M = 5
Example.matrix = np.ones(Example.M, Example.M)

or you define a class method and use that set M:

class Example():
    M=0
    matrix=np.ones((M,M))
    @classmethod
    def setM(m):
        Example.M = m

or you move the definition of matrix inside the __init__ method.