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

首先，我将它设计为将字符串拆分为子字符串以解析包含十六进制的字符串。 今天我把它变成复杂的，但仍然简单的生成器。

def chunker(iterable, size, reductor, condition):
    it = iter(iterable)
    def chunk_generator():
        return (next(it) for _ in range(size))
    chunk = reductor(chunk_generator())
    while condition(chunk):
        yield chunk
        chunk = reductor(chunk_generator())

参数:

明显的

Iterable是任何包含/生成/迭代输入数据的Iterable /迭代器/生成器， 当然，大小是你想要得到的块的大小，

更有趣的

reductor is a callable, which receives generator iterating over content of chunk. I'd expect it to return sequence or string, but I don't demand that. You can pass as this argument for example list, tuple, set, frozenset, or anything fancier. I'd pass this function, returning string (provided that iterable contains / generates / iterates over strings): def concatenate(iterable): return ''.join(iterable) Note that reductor can cause closing generator by raising exception. condition is a callable which receives anything what reductor returned. It decides to approve & yield it (by returning anything evaluating to True), or to decline it & finish generator's work (by returning anything other or raising exception). When number of elements in iterable is not divisible by size, when it gets exhausted, reductor will receive generator generating less elements than size. Let's call these elements lasts elements. I invited two functions to pass as this argument: lambda x:x - the lasts elements will be yielded. lambda x: len(x)==<size> - the lasts elements will be rejected. replace <size> using number equal to size

from itertools import izip_longest

def chunker(iterable, chunksize, filler):
    return izip_longest(*[iter(iterable)]*chunksize, fillvalue=filler)

在我的特殊情况下，我需要填充项目，以重复最后一个元素，直到它达到大小，所以我改变了这个答案，以适应我的需要。

大小为4的输入输出示例:

Input = [1,2,3,4,5,6,7,8]
Output= [[1,2,3,4], [5,6,7,8]]

Input = [[1,2,3,4,5,6,7]]
Output= [[1,2,3,4], [5,6,7,7]]

Input = [1,2,3,4,5]
Output= [[1,2,3,4], [5,5,5,5]]

def chunker(seq, size):
    res = []
    for el in seq:
        res.append(el)
        if len(res) == size:
            yield res
            res = []
    if res:
        res = res + (size - len(res)) * [res[-1]] 
        yield res

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。

chunk_size = 4
for i in range(0, len(ints), chunk_size):
    chunk = ints[i:i+chunk_size]
    # process chunk of size <= chunk_size