我想在 for 循环中添加所有列表并制作一个 ndarray

Question

例如我有一个列表

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

我想在for循环中添加所有列表并制作一个ndarray，

关注

for list in lists:
    sum function...

结果我想要的是

[22, 26, 30]

我怎样才能把它写成漂亮的代码？

Answer 1

如果你使用的是 NumPy，那很简单：

import numpy as np

l = [[1,2,3], [4,5,6], [7,8,9], [10,11,12]]
arr = np.array(l)
result = arr.sum(axis=0)
print(result)
# [22 26 30]

对于纯 Python 解决方案：

l = [[1,2,3], [4,5,6], [7,8,9], [10,11,12]]
result = [sum(column) for column in zip(*l)]
print(result)
# [22, 26, 30]

NumPy 解决方案会更快。

In [1]: import numpy as np

In [2]: arr = np.arange(10 ** 6).reshape(10 ** 3, 10 ** 3)

In [3]: %timeit arr.sum(axis=0)
881 µs ± 22.2 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [4]: l = arr.tolist()

In [5]: %timeit [sum(x) for x in zip(*l)]
76.8 ms ± 1.84 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

Answer 2

你不需要 Numpy 就可以做到这一点，只需要使用 python 的基础。（所有子列表的长度必须相同）

l =  [[1,2,3], [4,5,6], [7,8,9], [10,11,12]]
result = [sum(subL[i] for subL in l) for i in range(len(l[0]))]

或者，没有列表理解：

result = []
for i in range(len(l[0])):
    n = 0
    for x in l:
        n += subL[i]
    result.append(n)

两者都产生输出 [22, 26, 30]。

如果你对性能感兴趣，我写了这个（不确定是否准确）：

from timeit import timeit
import numpy as np

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

def listComprehension():
    result = [sum(subL[i] for subL in l) for i in range(len(l[0]))]

def basic():
    result = []
    for i in range(len(l[0])):
        n = 0
        for subL in l:
            n += subL[i]
        result.append(n)

def zipped():
    result = [sum(column) for column in zip(*l)]

def numpyied():
    arr = np.array(l)
    result = arr.sum(axis=0)

print(timeit("listComprehension()", setup = "from __main__ import listComprehension"))
# 3.738487364
print(timeit("basic()", setup = "from __main__ import basic"))
# 1.953782115
print(timeit("zipped()", setup = "from __main__ import zipped"))
# 1.413262091
print(timeit("numpyied()", setup = "from __main__ import numpyied"))
# 9.576366059999998

令人惊讶的是 numpy 更慢，但我不能说为什么。（最后两个函数取自）

我想在 for 循环中添加所有列表并制作一个 ndarray

I want to add all lists in for loop and make one ndarray

python

numpy

numpy-ndarray