python - 初始化派生自和抽象类的类时出现python错误

Question

我有这个简单的代码，我得到一个奇怪的错误：

from abc import ABCMeta, abstractmethod

class CVIterator(ABCMeta):

    def __init__(self):

        self.n = None # the value of n is obtained in the fit method
        return


class KFold_new_version(CVIterator): # new version of KFold

    def __init__(self, k):
        assert k > 0, ValueError('cannot have k below 1')
        self.k = k
        return 


cv = KFold_new_version(10)

In [4]: ---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-4-ec56652b1fdc> in <module>()
----> 1 __pyfile = open('''/tmp/py13196IBS''');exec(compile(__pyfile.read(), '''/home/donbeo/Desktop/prova.py''', 'exec'));__pyfile.close()

/home/donbeo/Desktop/prova.py in <module>()
     19 
     20 
---> 21 cv = KFold_new_version(10)

TypeError: __new__() missing 2 required positional arguments: 'bases' and 'namespace'

我究竟做错了什么？理论上的解释将不胜感激。

Answer 1

您ABCMeta错误地使用了元类。它是一个元类，而不是一个基类。就这样使用它。

对于 Python 2，这意味着将其分配给__metaclass__类的属性：

class CVIterator(object):
    __metaclass__ = ABCMeta

    def __init__(self):
        self.n = None # the value of n is obtained in the fit method

在 Python 3 中，您将metaclass=...在定义类时使用以下语法：

class CVIterator(metaclass=ABCMeta):
    def __init__(self):
        self.n = None # the value of n is obtained in the fit method

从 Python 3.4 开始，您可以使用abc.ABC辅助类作为基类：

from abc import ABC

class CVIterator(ABC):
    def __init__(self):
        self.n = None # the value of n is obtained in the fit method