随机select两个列表的某些索引创建两个新列表

Question

假设我有两个列表：

A = ['cat', 'dog', 'cow', 'pig', 'monkey']
B = ['Felix', 'Fido', 'Moo', 'Trump', 'King Kong']

现在我想通过 selecting 3 个随机索引（不替换）创建 lists A 和 list B，同时保持 A 和 A 的值之间的关系B.

例如，

随机 selected 索引：4, 0, 3

因此，

A = ['monkey', 'cat', 'pig']
B = ['King Kong', 'Felix', 'Trump']

有没有一种方法可以做到这一点而无需编写迭代 3 次以 select 随机索引的 for 循环？

Answer 1

Python 有一个示例函数，可以在不替换的情况下进行选择。您可以从指数中抽样并将样本应用于您的输入。

from random import sample

A = ['cat', 'dog', 'cow', 'pig', 'monkey']
B = ['Felix', 'Fido', 'Moo', 'Trump', 'King Kong']
k = 3

samp = sample(range(len(A)), k)
A_p = [A[i] for i in samp]
B_p = [B[i] for i in samp]

Answer 2

您可以zip the lists, then pick 3 random pairs with random.sample，最后再次将这些对拆分为单独的列表：

import random

pairs = list(zip(A, B))  # make pairs out of the two lists
pairs = random.sample(pairs, 3)  # pick 3 random pairs
A1, B1 = zip(*pairs)  # separate the pairs

这是一步一步发生的事情：

>>> list(zip(A, B))
[('cat', 'Felix'), ('dog', 'Fido'), ('cow', 'Moo'), ('pig', 'Trump'), ('monkey', 'King Kong')]
>>> random.sample(_, 3)
[('monkey', 'King Kong'), ('pig', 'Trump'), ('dog', 'Fido')]
>>> list(zip(*_))
[('monkey', 'pig', 'dog'), ('King Kong', 'Trump', 'Fido')]

Answer 3

您可以使用：

import numpy as np

indx = np.random.choice(len(A),3,False)

np.array(A)[indx]
Out[593]: array(['cow', 'pig', 'monkey'], dtype='<U6')

np.array(B)[indx]
Out[594]: array(['Moo', 'Trump', 'King Kong'], dtype='<U9')

Answer 4

我认为我最喜欢 zip 答案，但我会使用 列表理解

来采用这种方法

import random   

l1 = ['cat', 'dog', 'cow', 'pig', 'monkey']
l2 = ['Felix', 'Fido', 'Moo', 'Trump', 'King Kong']
r = [random.randint(0, len(l1)-1) for i in range(3)]

A = [l1[i] for i in r]
B = [l2[i] for i in r]

~/python/Whosebug/9.23$ python3.7 loop.py 
['monkey', 'cow', 'cat']
['King Kong', 'Moo', 'Felix']

Answer 5

尽管可以将循环委托给函数或将其装扮成列表理解，但您无法避免循环。不过，您可以遍历每个列表一次。

#lets assume your randomly selected numbers are 4, 0 and 3
indices = (4,0,3)
# lets assign an ordinal to each index
indices_ordinals = [(ind, ord) for ind, ord in enumerate(indices)]
# now we can re sort the indices, since we have recorded their original sequence
indices_ordinals.sort(key = lambda x: x[0])
# now go through the lists (for simplicity we assume they have equal number of elemwnts) and fill the output dicts
csi = 0 # csi is the index that we are looking for currently
A1 = {}
B1 = {}
for i, elem in enumerate(A):
    if i == indices_ordinals[csi][0]:
        A1[indices_ordinals[1]] : elem
        csi += 1
csi =0
for i, elem in enumerate(B):
    if i == indices_ordinals[csi][0]:
        B1[indices_ordinals[1]] : elem
        csi += 1
#now we only need to sort A1 and B1 to get desired output
A2=[A1[i] for i in sorted(A1)]
B2=[B1[i] for i in sorted(B1)]

随机select两个列表的某些索引创建两个新列表

Randomly select certain indices of two lists to create two new lists

python

list

indices

python-3.x