Python相当于自定义类的C ++ begin（）和end（） [英] Python equivalent of C++ begin() and end() for custom classes

问题描述

假设您有一个字符，其键是整数。值也是字典，其键是字符串，其值是numpy数组。
类似于：

Say you have a dictionary whose keys are integers. The values are also dictionaries whose keys are strings and whose values are numpy arrays. Something like:

custom = {1: {'a': np.zeros(10), 'b': np.zeros(100)}, 2:{'c': np.zeros(20), 'd': np.zeros(200)}}

我在代码中一直使用这个自定义数据结构，每次我需要迭代这个结构的numpy数组中的每一行，我必须这样做：

I've been using this custom data structure quite a lot in the code, and every time I need to iterate over each of the rows in the numpy arrays of this structure, I have to do:

for d, delem in custom.items():
    for k, v in delem.items():
        for row in v:
            print(row)

是否可以在函数àla C ++中封装此行为，您可以在其中实际实现自定义 begin（）和端（）？此外，迭代器还应该包含有关其相应字典中的键的信息。我设想类似于：

Is it possible to encapsulate this behavior in functions à la C++ where you can actually implement custom begin() and end()? Also, the iterator should also have information about the keys in their corresponding dictionaries. I envision something like:

for it in custom:
    d, e, row = *it
    # then do something with these

推荐答案

import numpy as np

custom = {
    1: {'a': np.zeros(10), 'b': np.zeros(100)}, 
    2:{'c': np.zeros(20), 'd': np.zeros(200)}
}

my_gen = (
    (key, subkey, np_array) 
    for (key, a_dict) in custom.items() 
    for subkey, np_array in a_dict.items() 
)

for key, subkey, np_array in my_gen:
    print(key, subkey, np_array)

--output:--
1 b [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
1 a [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
2 d [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.]
2 c [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.  0.
  0.  0.]

或者，您可以重建数据结构为你的目的更有用的东西：

Or, you could reconstitute your data structure into something that is more useful for your purposes:

import numpy as np

custom = {
    1: {'a': np.zeros(10), 'b': np.zeros(100)}, 
    2:{'c': np.zeros(20), 'd': np.zeros(200)}
}

#Create a *list* of tuples:
converted_data = [
    (np_array, subkey, key)
    for (key, a_dict) in custom.items() 
    for subkey, np_array in a_dict.items() 
]

for np_array, subkey, key in converted_data:
    print(key, subkey, np_array)

创建自定义迭代器：

class Dog:
    def __init__(self, data):
        self.data = data
        self.max = len(data)
        self.index_pointer = 0

    def __next__(self):
        index = self.index_pointer

        if index < self.max:
            current_val = self.data[index]
            self.index_pointer += 1
            return current_val
        else:
            raise StopIteration


class MyIter:
    def __iter__(self):
        return Dog([1, 2, 3])


for i in MyIter():
    print(i)

--output:--
1
2
3

__ iter __（）只需要返回一个实现的对象__next __（）方法，所以你可以将这两个类组合起来：

__iter__() just needs to return an object that implements a __next__() method, so you can combine those two classes like this:

class MyIter:
    def __init__(self, data):
        self.data = data
        self.max = len(data)
        self.index_pointer = 0

    def __iter__(self):
        return self  #I have a __next__() method, so let's return me!

    def __next__(self):
        index = self.index_pointer

        if index < self.max:
            current_val = self.data[index]
            self.index_pointer += 1
            return current_val
        else:
            raise StopIteration

for i in MyIter([1, 2, 3]):
    print(i)

--output:--
1
2
3

更复杂 __ next __（）方法：

import numpy as np

class CustomIter:
    def __init__(self, data):
        self.data = data
        self.count = 0


    def __iter__(self):
        return self

    def __next__(self):
        count = self.count
        self.count += 1

        if count == 0:  #On first iteration, retun a sum of the keys
            return sum(self.data.keys())

        elif count == 1: #On second iteration, return the subkeys in tuples
            subkeys =  [ 
                a_dict.keys()
                for a_dict in self.data.values()
            ]

            return subkeys

        elif count == 2: #On third iteration, return the count of np arrays
            np_arrays = [
                np_array
                for a_dict in self.data.values()
                for np_array in a_dict.values()
            ]

            return len(np_arrays)

        else:  #Quit after three iterations
            raise StopIteration


custom = {
    1: {'a': np.zeros(10), 'b': np.zeros(100)}, 
    2:{'c': np.zeros(20), 'd': np.zeros(200)}
}

for i in CustomIter(custom):
    print(i)


--output:--
3
[dict_keys(['b', 'a']), dict_keys(['d', 'c'])]
4

这篇关于Python相当于自定义类的C ++ begin（）和end（）的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！