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

我非常喜欢tzot和J.F.Sebastian提出的Python文档版本,但它有两个缺点:

它不是很明确我通常不希望在最后一个块中有填充值

我在代码中经常使用这个:

from itertools import islice

def chunks(n, iterable):
    iterable = iter(iterable)
    while True:
        yield tuple(islice(iterable, n)) or iterable.next()

更新:一个懒块版本:

from itertools import chain, islice

def chunks(n, iterable):
   iterable = iter(iterable)
   while True:
       yield chain([next(iterable)], islice(iterable, n-1))

其他回答

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

itertools模块中的配方提供了两种方法来实现这一点,具体取决于您希望如何处理最终的奇数大小的批次(保留它、用填充值填充它、忽略它或引发异常):

from itertools import islice, izip_longest

def batched(iterable, n):
    "Batch data into lists of length n. The last batch may be shorter."
    # batched('ABCDEFG', 3) --> ABC DEF G
    it = iter(iterable)
    while True:
        batch = list(islice(it, n))
        if not batch:
            return
        yield batch

def grouper(iterable, n, *, incomplete='fill', fillvalue=None):
    "Collect data into non-overlapping fixed-length chunks or blocks"
    # grouper('ABCDEFG', 3, fillvalue='x') --> ABC DEF Gxx
    # grouper('ABCDEFG', 3, incomplete='strict') --> ABC DEF ValueError
    # grouper('ABCDEFG', 3, incomplete='ignore') --> ABC DEF
    args = [iter(iterable)] * n
    if incomplete == 'fill':
        return zip_longest(*args, fillvalue=fillvalue)
    if incomplete == 'strict':
        return zip(*args, strict=True)
    if incomplete == 'ignore':
        return zip(*args)
    else:
        raise ValueError('Expected fill, strict, or ignore')

我专门为此写了一个小图书馆,这里有。库的分块函数特别有效,因为它是作为生成器实现的,因此在某些情况下可以节省大量内存。它也不依赖切片表示法,因此可以使用任意迭代器。

import iterlib

print list(iterlib.chunked(xrange(1, 1000), 10))
# prints [(1, 2, 3, 4, 5, 6, 7, 8, 9, 10), (11, 12, 13, 14, 15, 16, 17, 18, 19, 20), ...]

延迟加载版本

导入pprintpprint.pprint(列表(块(范围(10,75),10))[范围(10、20),范围(20、30),范围(30、40),范围(40、50),范围(50、60),范围(60、70),范围(70,75)]将此实现的结果与接受答案的示例使用结果进行比较。

上面的许多函数都假定整个可迭代函数的长度是预先知道的,或者至少计算起来很便宜。

对于一些流式对象,这意味着首先将完整数据加载到内存中(例如下载整个文件)以获取长度信息。

但是,如果您还不知道完整大小,可以使用以下代码:

def chunks(iterable, size):
    """
    Yield successive chunks from iterable, being `size` long.

    https://stackoverflow.com/a/55776536/3423324
    :param iterable: The object you want to split into pieces.
    :param size: The size each of the resulting pieces should have.
    """
    i = 0
    while True:
        sliced = iterable[i:i + size]
        if len(sliced) == 0:
            # to suppress stuff like `range(max, max)`.
            break
        # end if
        yield sliced
        if len(sliced) < size:
            # our slice is not the full length, so we must have passed the end of the iterator
            break
        # end if
        i += size  # so we start the next chunk at the right place.
    # end while
# end def

这之所以有效,是因为如果您传递了一个iterable的结尾,slice命令将返回less/no元素:

"abc"[0:2] == 'ab'
"abc"[2:4] == 'c'
"abc"[4:6] == ''

我们现在使用切片的结果,并计算生成的块的长度。如果它低于我们的预期,我们知道我们可以结束迭代。

这样,除非访问,否则不会执行迭代器。

def chunk(lst):
    out = []
    for x in xrange(2, len(lst) + 1):
        if not len(lst) % x:
            factor = len(lst) / x
            break
    while lst:
        out.append([lst.pop(0) for x in xrange(factor)])
    return out