你可以使用列表扩展方法. 它显示是最快的:

flat_list = []
for sublist in l:
    flat_list.extend(sublist)

表演:

import functools
import itertools
import numpy
import operator
import perfplot


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(numpy.array(a).flat)


def extend(a):
    n = []

    list(map(n.extend, a))

    return n


perfplot.show(
    setup = lambda n: [list(range(10))] * n,
    kernels = [
        functools_reduce_iconcat, extend, itertools_chain, numpy_flat
        ],
    n_range = [2**k for k in range(16)],
    xlabel = 'num lists',
    )

出口:

此分類上一篇

def flatten_array(arr):
  result = []
  for item in arr:
    if isinstance(item, list):
      for num in item:
        result.append(num)
    else:
      result.append(item)
  return result

print(flatten_array([1, 2, [3, 4, 5], 6, [7, 8], 9]))
// output: [1, 2, 3, 4, 5, 6, 7, 8, 9]

对于包含多个列表的列表,这里是一个重复的解决方案,为我工作,我希望它是正确的:

# Question 4
def flatten(input_ls=[]) -> []:
    res_ls = []
    res_ls = flatten_recursive(input_ls, res_ls)

    print("Final flatten list solution is: \n", res_ls)

    return res_ls


def flatten_recursive(input_ls=[], res_ls=[]) -> []:
    tmp_ls = []

    for i in input_ls:
        if isinstance(i, int):
            res_ls.append(i)
        else:
            tmp_ls = i
            tmp_ls.append(flatten_recursive(i, res_ls))

    print(res_ls)
    return res_ls


flatten([0, 1, [2, 3], 4, [5, 6]])  # test
flatten([0, [[[1]]], [[2, 3], [4, [[5, 6]]]]])

出口:

[0, 1, 2, 3]
[0, 1, 2, 3, 4, 5, 6]
[0, 1, 2, 3, 4, 5, 6]
Final flatten list solution is: 
 [0, 1, 2, 3, 4, 5, 6]
[0, 1]
[0, 1]
[0, 1]
[0, 1, 2, 3]
[0, 1, 2, 3, 4, 5, 6]
[0, 1, 2, 3, 4, 5, 6]
[0, 1, 2, 3, 4, 5, 6]
[0, 1, 2, 3, 4, 5, 6]
[0, 1, 2, 3, 4, 5, 6]
Final flatten list solution is: 
 [0, 1, 2, 3, 4, 5, 6]

注意: 下面适用于 Python 3.3+ 因为它使用 yield_from. six 也是第三方包,尽管它是稳定的。

在obj = [1, 2,], [3, 4], [5, 6]的情况下,这里的所有解决方案都很好,包括列表理解和 itertools.chain.from_iterable。

但是,考虑这个稍微复杂的案例:

>>> obj = [[1, 2, 3], [4, 5], 6, 'abc', [7], [8, [9, 10]]]

这里有几个问题:

您可以以以下方式解决此问题:

>>> from collections import Iterable
>>> from six import string_types

>>> def flatten(obj):
...     for i in obj:
...         if isinstance(i, Iterable) and not isinstance(i, string_types):
...             yield from flatten(i)
...         else:
...             yield i


>>> list(flatten(obj))
[1, 2, 3, 4, 5, 6, 'abc', 7, 8, 9, 10]

在这里,您可以检查(一)的子元素(一)与(一)的Iterable(一)无效,从(一)的ABC,但也希望确保(二)的元素(一)不是“类似于(一)的”。

您也可以使用NumPy的公寓:

import numpy as np
list(np.array(l).flat)

它只有在超级列表具有相同的尺寸时才有效。

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