About Pytorch super()

  • 作者：miaulab
  • 时间：2020-01-02 20:50:23
  • 99次访问
Pytorch basics

Ref: https://blog.csdn.net/shiheyingzhe/article/details/83051471

super() has 2 major pros:

  • Avoids using the base class explicitly
  • Working with Multiple Inheritance

reminder:

  • During super instantiation, it firstly tries to find the base class of the same level, if no such case, executes its own base class.
  • No repeat call

Example1:

1
class A():
2
    def __init__(self):
3
        print('Call class A')
4
        super(A, self).__init__()
5
        print('Leave class A')
6
class B(A):
7
    def __init__(self):
8
        print('Call class B')
9
        super(B,self).__init__()
10
        print('Leave class B')
11
class C(A):
12
    def __init__(self):
13
        print('Call class C')
14
        super(C,self).__init__()
15
        print('Leave class C')
16
class D(A):
17
    def __init__(self):
18
        print('Call class D')
19
        super(D, self).__init__()
20
        print('Leave class D')
21
class E(B,C,D):
22
    def __init__(self):
23
        print('Call class E')
24
        super(E, self).__init__()
25
        print('Leave class E')
26
sample=E() 
27
print(E.__mro__)

the output:

1
Call class E
2
Call class B
3
Call class C
4
Call class D
5
Call class A
6
Leave class A
7
Leave class D
8
Leave class C
9
Leave class B
10
Leave class E
11
(<class '__main__.E'>, <class '__main__.B'>, <class '__main__.C'>, <class '__main__.D'>, <class '__main__.A'>, <class 'object'>)

Example2:

1
class A():
2
    def __init__(self):
3
        print('Call class A')
4
        print('Leave class A')
5
class B(A):
6
    def __init__(self):
7
        print('Call class B')
8
        print('Leave class B')
9
class C(A):
10
    def __init__(self):
11
        print('Call class C')
12
        super(C,self).__init__()
13
        print('Leave class C')
14
class D(A):
15
    def __init__(self):
16
        print('Call class D')
17
        super(D, self).__init__()
18
        print('Leave class D')
19
class E(B,C,D):
20
    def __init__(self):
21
        print('Call class E')
22
        B.__init__(self)
23
        C.__init__(self)
24
        D.__init__(self)
25
        print('Leave class E')
26
sample=E()
27
print(E.__mro__)

the output:

1
Call class E
2
Call class B
3
Leave class B
4
Call class C
5
Call class D
6
Call class A
7
Leave class A
8
Leave class D
9
Leave class C
10
Call class D
11
Call class A
12
Leave class A
13
Leave class D
14
Leave class E
15
(<class '__main__.E'>, <class '__main__.B'>, <class '__main__.C'>, <class '__main__.D'>, <class '__main__.A'>, <class 'object'>)

上一篇：python basics

下一篇：Setup Jupyter Notebook with Python 3 on Ubuntu18.04