我希望通过将迭代器从列表中删除，我不是简单地复制列表的一部分。生成器可以被切片，它们将自动仍然是一个生成器，而列表将被切片成1000个条目的大块，这是较低的效率。

def iter_group(iterable, batch_size:int):
    length = len(iterable)
    start = batch_size*-1
    end = 0
    while(end < length):
        start += batch_size
        end += batch_size
        if type(iterable) == list:
            yield (iterable[i] for i in range(start,min(length-1,end)))
        else:
            yield iterable[start:end]

用法:

items = list(range(1,1251))

for item_group in iter_group(items, 1000):
    for item in item_group:
        print(item)

这个问题的理想解决方案是使用迭代器(而不仅仅是序列)。它还应该是快速的。

这是itertools文档提供的解决方案:

def grouper(n, iterable, fillvalue=None):
    #"grouper(3, 'ABCDEFG', 'x') --> ABC DEF Gxx"
    args = [iter(iterable)] * n
    return itertools.izip_longest(fillvalue=fillvalue, *args)

在我的mac book air上使用ipython的%timeit，我每次循环得到47.5 us。

然而，这真的不适合我，因为结果被填充为偶数大小的组。没有填充的解决方案稍微复杂一些。最天真的解决方案可能是:

def grouper(size, iterable):
    i = iter(iterable)
    while True:
        out = []
        try:
            for _ in range(size):
                out.append(i.next())
        except StopIteration:
            yield out
            break
        
        yield out

简单，但相当慢:每循环693个

我能想到的最好的解决方案是使用islice进行内循环:

def grouper(size, iterable):
    it = iter(iterable)
    while True:
        group = tuple(itertools.islice(it, None, size))
        if not group:
            break
        yield group

对于同样的数据集，我每循环得到305 us。

由于无法更快地得到一个纯粹的解决方案，我提供了以下解决方案，但有一个重要的警告:如果您的输入数据中有filldata的实例，则可能会得到错误的答案。

def grouper(n, iterable, fillvalue=None):
    #"grouper(3, 'ABCDEFG', 'x') --> ABC DEF Gxx"
    args = [iter(iterable)] * n
    # itertools.zip_longest on Python 3
    for x in itertools.izip_longest(*args, fillvalue=fillvalue):
        if x[-1] is fillvalue:
            yield tuple(v for v in x if v is not fillvalue)
        else:
            yield x

我真的不喜欢这个答案，但它明显更快。每回路124 us

如果你不介意使用外部包，你可以使用iteration_utilities。Grouper from iteration_utilities它支持所有可迭代对象(不仅仅是序列):

from iteration_utilities import grouper
seq = list(range(20))
for group in grouper(seq, 4):
    print(group)

打印:

(0, 1, 2, 3)
(4, 5, 6, 7)
(8, 9, 10, 11)
(12, 13, 14, 15)
(16, 17, 18, 19)

如果长度不是组大小的倍数，它还支持填充(不完整的最后一组)或截断(丢弃不完整的最后一组)最后一个:

from iteration_utilities import grouper
seq = list(range(17))
for group in grouper(seq, 4):
    print(group)
# (0, 1, 2, 3)
# (4, 5, 6, 7)
# (8, 9, 10, 11)
# (12, 13, 14, 15)
# (16,)

for group in grouper(seq, 4, fillvalue=None):
    print(group)
# (0, 1, 2, 3)
# (4, 5, 6, 7)
# (8, 9, 10, 11)
# (12, 13, 14, 15)
# (16, None, None, None)

for group in grouper(seq, 4, truncate=True):
    print(group)
# (0, 1, 2, 3)
# (4, 5, 6, 7)
# (8, 9, 10, 11)
# (12, 13, 14, 15)

基准

我还决定比较上面提到的几种方法的运行时间。这是一个对数-对数图，根据不同大小的列表将“10”个元素分组。对于定性结果:较低意味着更快:

至少在这个基准测试中iteration_utilities。石斑鱼表现最好。接着是Craz。

基准是用simple_benchmark1创建的。运行这个基准测试的代码是:

import iteration_utilities
import itertools
from itertools import zip_longest

def consume_all(it):
    return iteration_utilities.consume(it, None)

import simple_benchmark
b = simple_benchmark.BenchmarkBuilder()

@b.add_function()
def grouper(l, n):
    return consume_all(iteration_utilities.grouper(l, n))

def Craz_inner(iterable, n, fillvalue=None):
    args = [iter(iterable)] * n
    return zip_longest(*args, fillvalue=fillvalue)

@b.add_function()
def Craz(iterable, n, fillvalue=None):
    return consume_all(Craz_inner(iterable, n, fillvalue))

def nosklo_inner(seq, size):
    return (seq[pos:pos + size] for pos in range(0, len(seq), size))

@b.add_function()
def nosklo(seq, size):
    return consume_all(nosklo_inner(seq, size))

def SLott_inner(ints, chunk_size):
    for i in range(0, len(ints), chunk_size):
        yield ints[i:i+chunk_size]

@b.add_function()
def SLott(ints, chunk_size):
    return consume_all(SLott_inner(ints, chunk_size))

def MarkusJarderot1_inner(iterable,size):
    it = iter(iterable)
    chunk = tuple(itertools.islice(it,size))
    while chunk:
        yield chunk
        chunk = tuple(itertools.islice(it,size))

@b.add_function()
def MarkusJarderot1(iterable,size):
    return consume_all(MarkusJarderot1_inner(iterable,size))

def MarkusJarderot2_inner(iterable,size,filler=None):
    it = itertools.chain(iterable,itertools.repeat(filler,size-1))
    chunk = tuple(itertools.islice(it,size))
    while len(chunk) == size:
        yield chunk
        chunk = tuple(itertools.islice(it,size))

@b.add_function()
def MarkusJarderot2(iterable,size):
    return consume_all(MarkusJarderot2_inner(iterable,size))

@b.add_arguments()
def argument_provider():
    for exp in range(2, 20):
        size = 2**exp
        yield size, simple_benchmark.MultiArgument([[0] * size, 10])

r = b.run()

1免责声明:我是iteration_utilities和simple_benchmark库的作者。

下面是我的go works on lists,iter和range…懒洋洋地:

def chunker(it,size):
    rv = [] 
    for i,el in enumerate(it,1) :   
        rv.append(el)
        if i % size == 0 : 
            yield rv
            rv = []
    if rv : yield rv        

几乎变成了一句俏皮话;(

In [95]: list(chunker(range(9),2) )                                                                                                                                          
Out[95]: [[0, 1], [2, 3], [4, 5], [6, 7], [8]]

In [96]: list(chunker([1,2,3,4,5],2) )                                                                                                                                       
Out[96]: [[1, 2], [3, 4], [5]]

In [97]: list(chunker(iter(range(9)),2) )                                                                                                                                    
Out[97]: [[0, 1], [2, 3], [4, 5], [6, 7], [8]]

In [98]: list(chunker(range(9),25) )                                                                                                                                         
Out[98]: [[0, 1, 2, 3, 4, 5, 6, 7, 8]]

In [99]: list(chunker(range(9),1) )                                                                                                                                          
Out[99]: [[0], [1], [2], [3], [4], [5], [6], [7], [8]]

In [101]: %timeit list(chunker(range(101),2) )                                                                                                                               
11.3 µs ± 68.2 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)

在Python 3.8中，您可以使用walrus操作符和itertools.islice。

from itertools import islice

list_ = [i for i in range(10, 100)]

def chunker(it, size):
    iterator = iter(it)
    while chunk := list(islice(iterator, size)):
        print(chunk)

In [2]: chunker(list_, 10)                                                         
[10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
[20, 21, 22, 23, 24, 25, 26, 27, 28, 29]
[30, 31, 32, 33, 34, 35, 36, 37, 38, 39]
[40, 41, 42, 43, 44, 45, 46, 47, 48, 49]
[50, 51, 52, 53, 54, 55, 56, 57, 58, 59]
[60, 61, 62, 63, 64, 65, 66, 67, 68, 69]
[70, 71, 72, 73, 74, 75, 76, 77, 78, 79]
[80, 81, 82, 83, 84, 85, 86, 87, 88, 89]
[90, 91, 92, 93, 94, 95, 96, 97, 98, 99]

使用map()而不是zip()修复填充问题在J.F.塞巴斯蒂安的回答:

>>> def chunker(iterable, chunksize):
...   return map(None,*[iter(iterable)]*chunksize)

例子:

>>> s = '1234567890'
>>> chunker(s, 3)
[('1', '2', '3'), ('4', '5', '6'), ('7', '8', '9'), ('0', None, None)]
>>> chunker(s, 4)
[('1', '2', '3', '4'), ('5', '6', '7', '8'), ('9', '0', None, None)]
>>> chunker(s, 5)
[('1', '2', '3', '4', '5'), ('6', '7', '8', '9', '0')]