以此为本,以此为本,以此为本,以此为本,以此为本,以此为本,以此为本。


如果您的列表来自列表理解,问题可以通过纠正理解更简单/直接解决;请参见Python列表理解;压缩列表?

在这里,最受欢迎的解决方案通常只包含一个“层”的清单。 查看清单不规则(自愿清单)的清单,解决方案完全清单一个深度清单的结构(重复,一般)。


当前回答

你的功能不起作用的原因是因为延伸延伸一个序列在现场,并且不会返回它。

reduce(lambda x,y: x.extend(y) or x, l)

注意:扩展比 + 列表更有效。

其他回答

我用 perfplot 测试了大多数建议的解决方案(我的宠物项目,基本上是时间周围的插槽),并发现

import functools
import operator
functools.reduce(operator.iconcat, a, [])

要成为最快的解决方案,无论是许多小列表还是很少的长列表都被混合(operator.iadd 同样快)。

更简单、更可接受的选择是

out = []
for sublist in a:
    out.extend(sublist)

如果字体列表的数量很大,这表现得比上面的建议略糟糕。

此分類上一篇

此分類上一篇


重复字符的代码:

import functools
import itertools
import operator

import numpy as np
import perfplot


def forfor(a):
    return [item for sublist in a for item in sublist]


def sum_brackets(a):
    return sum(a, [])


def functools_reduce(a):
    return functools.reduce(operator.concat, a)


def functools_reduce_iconcat(a):
    return functools.reduce(operator.iconcat, a, [])


def itertools_chain(a):
    return list(itertools.chain.from_iterable(a))


def numpy_flat(a):
    return list(np.array(a).flat)


def numpy_concatenate(a):
    return list(np.concatenate(a))


def extend(a):
    out = []
    for sublist in a:
        out.extend(sublist)
    return out


b = perfplot.bench(
    setup=lambda n: [list(range(10))] * n,
    # setup=lambda n: [list(range(n))] * 10,
    kernels=[
        forfor,
        sum_brackets,
        functools_reduce,
        functools_reduce_iconcat,
        itertools_chain,
        numpy_flat,
        numpy_concatenate,
        extend,
    ],
    n_range=[2 ** k for k in range(16)],
    xlabel="num lists (of length 10)",
    # xlabel="len lists (10 lists total)"
)
b.save("out.png")
b.show()

matplotlib.cbook.flatten() 将为粘贴列表工作,即使它们比示例更深地粘贴。

import matplotlib
l = [[1, 2, 3], [4, 5, 6], [7], [8, 9]]
print(list(matplotlib.cbook.flatten(l)))
l2 = [[1, 2, 3], [4, 5, 6], [7], [8, [9, 10, [11, 12, [13]]]]]
print(list(matplotlib.cbook.flatten(l2)))

结果:

[1, 2, 3, 4, 5, 6, 7, 8, 9]
[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]

这比 underscore 快 18 倍。

Average time over 1000 trials of matplotlib.cbook.flatten: 2.55e-05 sec
Average time over 1000 trials of underscore._.flatten: 4.63e-04 sec
(time for underscore._)/(time for matplotlib.cbook) = 18.1233394636

我会建议使用发电机与产量声明和产量从。

from collections.abc import Iterable

def flatten(items, ignore_types=(bytes, str)):
    """
       Flatten all of the nested lists to the one. Ignoring flatting of iterable types str and bytes by default.
    """
    for x in items:
        if isinstance(x, Iterable) and not isinstance(x, ignore_types):
            yield from flatten(x)
        else:
            yield x

values = [7, [4, 3, 5, [7, 3], (3, 4), ('A', {'B', 'C'})]]

for v in flatten(values):
    print(v)

我创建了一点功能,基本上可以平滑任何东西. 你可以用管道:管道安装平滑一切

from flatten_everything import flatten_everything
withoutprotection=list(
    flatten_everything(
        [
            1,
            1,
            2,
            [3, 4, 5, [6, 3, [2, 5, ["sfs", "sdfsfdsf",]]]],
            1,
            3,
            34,
            [
                55,
                {"brand": "Ford", "model": "Mustang", "year": 1964, "yearxx": 2020},
                pd.DataFrame({"col1": [1, 2], "col2": [3, 4]}),
                {"col1": [1, 2], "col2": [3, 4]},
                55,
                {"k32", 34},
                np.array([[[1, 2], [3, 4]], [[5, 6], [7, 8]]]),
                (np.arange(22), np.eye(2, 2), 33),
            ],
        ]
    )
)
print(withoutprotection)
output:
[1, 1, 2, 3, 4, 5, 6, 3, 2, 5, 'sfs', 'sdfsfdsf', 1, 3, 34, 55, 'Ford', 'Mustang', 1964, 2020, 1, 2, 3, 4, 1, 2, 3, 4, 55, 34, 'k32', 1, 2, 3, 4, 5, 6, 7, 8, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 1.0, 0.0, 0.0, 1.0, 33]

你甚至可以保护物体免受闪烁:

from flatten_everything import ProtectedDict,ProtectedList,ProtectedTuple
withprotection=list(
    flatten_everything(
        [
            1,
            1,
            2,
            [3, 4, 5, [6, 3, [2, 5, ProtectedList(["sfs", "sdfsfdsf",])]]],
            1,
            3,
            34,
            [
                55,
                ProtectedDict({"brand": "Ford", "model": "Mustang", "year": 1964, "yearxx": 2020}),
                pd.DataFrame({"col1": [1, 2], "col2": [3, 4]}),
                {"col1": [1, 2], "col2": [3, 4]},
                55,
                {"k32", 34},
                np.array([[[1, 2], [3, 4]], [[5, 6], [7, 8]]]),
                ProtectedTuple((np.arange(22), np.eye(2, 2), 33)),
            ],
        ]
    )
)
print(withprotection)
output:
[1, 1, 2, 3, 4, 5, 6, 3, 2, 5, ['sfs', 'sdfsfdsf'], 1, 3, 34, 55, {'brand': 'Ford', 'model': 'Mustang', 'year': 1964, 'yearxx': 2020}, 1, 2, 3, 4, 1, 2, 3, 4, 55, 34, 'k32', 1, 2, 3, 4, 5, 6, 7, 8, (array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,17, 18, 19, 20, 21]), array([[1., 0.], [0., 1.]]), 33)]

这可能不是最有效的方式,但我认为要放一个单线(实际上是一个双线)。两种版本都会在任意的序列列列表上工作,并利用语言功能(Python 3.5)和回归。

def make_list_flat (l):
    flist = []
    flist.extend ([l]) if (type (l) is not list) else [flist.extend (make_list_flat (e)) for e in l]
    return flist

a = [[1, 2], [[[[3, 4, 5], 6]]], 7, [8, [9, [10, 11], 12, [13, 14, [15, [[16, 17], 18]]]]]]
flist = make_list_flat(a)
print (flist)

产量是

[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]

它首先以深度工作. 旋转会下降,直到它找到一个非列表元素,然后延伸当地变量板,然后转向父母。 每当板回来时,它在列表理解中延伸到父母的板。

上面的一个创建了几个地方列表,并返回它们,这些列表被用来扩展父母的列表,我认为这一点的路径可能是创建一个可怕的板块,如下。

a = [[1, 2], [[[[3, 4, 5], 6]]], 7, [8, [9, [10, 11], 12, [13, 14, [15, [[16, 17], 18]]]]]]
flist = []
def make_list_flat (l):
    flist.extend ([l]) if (type (l) is not list) else [make_list_flat (e) for e in l]

make_list_flat(a)
print (flist)

产量再次

[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]

虽然我目前对效率不确定。