如何将任意长度的列表拆分为大小相等的块?


请参阅如果数据结果将直接用于循环,并且不需要存储,则如何以块形式遍历列表。

对于字符串输入的同一问题,请参见每n个字符拆分字符串?。相同的技术通常适用,但也有一些变化。


当前回答

这项任务可以在公认答案中使用生成器轻松完成。我正在添加实现长度方法的类实现,这可能对某些人有用。我需要知道进度(使用tqdm),所以生成器应该返回块的数量。

class ChunksIterator(object):
    def __init__(self, data, n):
        self._data = data
        self._l = len(data)
        self._n = n

    def __iter__(self):
        for i in range(0, self._l, self._n):
            yield self._data[i:i + self._n]

    def __len__(self):
        rem = 1 if self._l % self._n != 0 else 0
        return self._l // self._n + rem

用法:

it = ChunksIterator([1,2,3,4,5,6,7,8,9], 2)
print(len(it))
for i in it:
  print(i)

其他回答

非常简单的事情:

def chunks(xs, n):
    n = max(1, n)
    return (xs[i:i+n] for i in range(0, len(xs), n))

对于Python 2,使用xrange()代替range()。

此时,我认为我们需要强制性的匿名递归函数。

Y = lambda f: (lambda x: x(x))(lambda y: f(lambda *args: y(y)(*args)))
chunks = Y(lambda f: lambda n: [n[0][:n[1]]] + f((n[0][n[1]:], n[1])) if len(n[0]) > 0 else [])

这是一个生成大小均匀的块的生成器:

def chunks(lst, n):
    """Yield successive n-sized chunks from lst."""
    for i in range(0, len(lst), n):
        yield lst[i:i + n]
import pprint
pprint.pprint(list(chunks(range(10, 75), 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]]

对于Python 2,使用xrange代替range:

def chunks(lst, n):
    """Yield successive n-sized chunks from lst."""
    for i in xrange(0, len(lst), n):
        yield lst[i:i + n]

下面是一行理解列表。不过,上面的方法更可取,因为使用命名函数使代码更容易理解。对于Python 3:

[lst[i:i + n] for i in range(0, len(lst), n)]

对于Python 2:

[lst[i:i + n] for i in xrange(0, len(lst), n)]

我很惊讶没有人想到使用iter的双参数形式:

from itertools import islice

def chunk(it, size):
    it = iter(it)
    return iter(lambda: tuple(islice(it, size)), ())

演示:

>>> list(chunk(range(14), 3))
[(0, 1, 2), (3, 4, 5), (6, 7, 8), (9, 10, 11), (12, 13)]

这适用于任何可迭代的对象,并延迟生成输出。它返回元组而不是迭代器,但我认为它还是有一定的优雅。它也不会垫;如果您需要填充,上面的一个简单变体就足够了:

from itertools import islice, chain, repeat

def chunk_pad(it, size, padval=None):
    it = chain(iter(it), repeat(padval))
    return iter(lambda: tuple(islice(it, size)), (padval,) * size)

演示:

>>> list(chunk_pad(range(14), 3))
[(0, 1, 2), (3, 4, 5), (6, 7, 8), (9, 10, 11), (12, 13, None)]
>>> list(chunk_pad(range(14), 3, 'a'))
[(0, 1, 2), (3, 4, 5), (6, 7, 8), (9, 10, 11), (12, 13, 'a')]

与基于izip_longest的解决方案一样,上面的解决方案也始终适用。据我所知,对于可选pad的函数,没有单行或双线itertools配方。通过结合以上两种方法,这一方法非常接近:

_no_padding = object()

def chunk(it, size, padval=_no_padding):
    if padval == _no_padding:
        it = iter(it)
        sentinel = ()
    else:
        it = chain(iter(it), repeat(padval))
        sentinel = (padval,) * size
    return iter(lambda: tuple(islice(it, size)), sentinel)

演示:

>>> list(chunk(range(14), 3))
[(0, 1, 2), (3, 4, 5), (6, 7, 8), (9, 10, 11), (12, 13)]
>>> list(chunk(range(14), 3, None))
[(0, 1, 2), (3, 4, 5), (6, 7, 8), (9, 10, 11), (12, 13, None)]
>>> list(chunk(range(14), 3, 'a'))
[(0, 1, 2), (3, 4, 5), (6, 7, 8), (9, 10, 11), (12, 13, 'a')]

我相信这是提议的提供可选填充的最短的分块器。

正如Tomasz Gandor所观察到的,如果两个填充块遇到一长串填充值,它们会意外停止。以下是以合理方式解决该问题的最后一个变体:

_no_padding = object()
def chunk(it, size, padval=_no_padding):
    it = iter(it)
    chunker = iter(lambda: tuple(islice(it, size)), ())
    if padval == _no_padding:
        yield from chunker
    else:
        for ch in chunker:
            yield ch if len(ch) == size else ch + (padval,) * (size - len(ch))

演示:

>>> list(chunk([1, 2, (), (), 5], 2))
[(1, 2), ((), ()), (5,)]
>>> list(chunk([1, 2, None, None, 5], 2, None))
[(1, 2), (None, None), (5, None)]

我很好奇不同方法的性能,这里是:

在Python 3.5.1上测试

import time
batch_size = 7
arr_len = 298937

#---------slice-------------

print("\r\nslice")
start = time.time()
arr = [i for i in range(0, arr_len)]
while True:
    if not arr:
        break

    tmp = arr[0:batch_size]
    arr = arr[batch_size:-1]
print(time.time() - start)

#-----------index-----------

print("\r\nindex")
arr = [i for i in range(0, arr_len)]
start = time.time()
for i in range(0, round(len(arr) / batch_size + 1)):
    tmp = arr[batch_size * i : batch_size * (i + 1)]
print(time.time() - start)

#----------batches 1------------

def batch(iterable, n=1):
    l = len(iterable)
    for ndx in range(0, l, n):
        yield iterable[ndx:min(ndx + n, l)]

print("\r\nbatches 1")
arr = [i for i in range(0, arr_len)]
start = time.time()
for x in batch(arr, batch_size):
    tmp = x
print(time.time() - start)

#----------batches 2------------

from itertools import islice, chain

def batch(iterable, size):
    sourceiter = iter(iterable)
    while True:
        batchiter = islice(sourceiter, size)
        yield chain([next(batchiter)], batchiter)


print("\r\nbatches 2")
arr = [i for i in range(0, arr_len)]
start = time.time()
for x in batch(arr, batch_size):
    tmp = x
print(time.time() - start)

#---------chunks-------------
def chunks(l, n):
    """Yield successive n-sized chunks from l."""
    for i in range(0, len(l), n):
        yield l[i:i + n]
print("\r\nchunks")
arr = [i for i in range(0, arr_len)]
start = time.time()
for x in chunks(arr, batch_size):
    tmp = x
print(time.time() - start)

#-----------grouper-----------

from itertools import zip_longest # for Python 3.x
#from six.moves import zip_longest # for both (uses the six compat library)

def grouper(iterable, n, padvalue=None):
    "grouper(3, 'abcdefg', 'x') --> ('a','b','c'), ('d','e','f'), ('g','x','x')"
    return zip_longest(*[iter(iterable)]*n, fillvalue=padvalue)

arr = [i for i in range(0, arr_len)]
print("\r\ngrouper")
start = time.time()
for x in grouper(arr, batch_size):
    tmp = x
print(time.time() - start)

结果:

slice
31.18285083770752

index
0.02184295654296875

batches 1
0.03503894805908203

batches 2
0.22681021690368652

chunks
0.019841909408569336

grouper
0.006506919860839844