Several methods of implementing Cartesian product in Python

Introduction: During the interview, the interviewer asked this question. It was not very good at the time. After the interview, I checked it and found that most of the blogs are implemented using toolkits, and most of the blog content is still complete. Similarly, even the numbers haven't changed, and I haven't found a few useful blogs after searching for a long time. In fact, this is also the current trend of most blogs, plagiarizing each other, without any thoughts of my own, I don't understand the meaning of writing such blogs. So I plan to implement a solution that does not use a toolkit, so with the help of someone else's blog, I realized the use of backtracking to solve the Cartesian product. Here are a few ways to solve this problem:

1. Toolkit

from itertools import product
res = list(product(['a','b],['c'],['d','e','g']))
print(res)

2. Recursive generator, this method refers to the content of the blog https://www.lfhacks.com/tech/python-recursive-generator , and the code is copied directly

def combi(seq):
    if not seq:
        yield []
    else:
        for element in seq[0]:
            for rest in combi(seq[1:]):
                yield [element] + rest

Test with the following code

n=[[1,2,3],[4],[5,6,7]]
print list(combi(n))

The operating mode is:

拆出第一个序列，得到1,2,3
对于1                      （暂存结果[1]）
拆出第2个序列，得到4
    对于4                      （暂存结果[1, 4]）
    拆出第3个序列，得到5,6,7
        对于5                       （暂存结果[1, 4, 5]）
        拆出第4个序列（空），得到（空）       
        把5加到（空）前面，返回结果    （返回暂存结果[1, 4, 5]）
        对于6                       （暂存结果[1, 4, 6]）
        拆出第4个序列（空），得到（空）        
        把6加到（空）前面，返回结果    （返回暂存结果[1, 4, 6]）
        对于7                       （暂存结果[1, 4, 7]）
        拆出第4个序列（空），得到（空）       
        把7加到（空）前面，返回结果    （返回暂存结果[1, 4, 7]）

3. Implementation of the retrospective method

arr = [['a','b],['c'],['d','e','g']]
res = []
layer = len(arr)
def func(arr,index,temp,layer):
    if len(temp)==layer:
        res.append(temp[:])
    else:
        if not arr[index]:
            layer -= 1
            func(arr, index + 1, temp,layer)
        else:
            for i in range(len(arr[index])):
                temp.append(arr[index][i])
                func(arr,index+1,temp,layer)
                temp.pop()

func(arr,0,[],layer)
print(res)

The reason for using layer here is to prevent the sub-list from appearing as an empty list. If the sub-list is not an empty list, you can add temp to the result set when the length of temp is equal to the length of arr.

Reference: https://blog.csdn.net/ylyg050518/article/details/78787432

Several methods of implementing Cartesian product in Python

Guess you like