如何使用时间复杂度优于O（n ^ 2）的STL向量和STL算法进行左连接？

Question

I have 2 vectors, which contain, let's say Person (name, surname, etc) objects. I want to take one of the vectors (let's name it "large") then for each element in this vector find corresponding element in second one ("small") and merge some data from "small" vector element to the "large" vector element. This operation is very similar to left join in SQL terms, but with additional merge of the data. The easiest way is to make 2 cycles, but that will lead to O(n^2) time complexity. Can I do better with STL algorithms?

Answer 1

如果对小向量进行排序 ，则可以通过扫描大向量获取合并部分的O（n log n），并使用binary_search查找小向量中的元素。

Answer 2

Yes! You can do it in O(nlogn) time complexity. Sort the second vector, which takes O(nlogn) time. for each element in first vector, find corresponding element in second one using binary search (STL has binary_search algorithm) and merge data to the element in first vector. for each element in first vector, we are spending O(logn) time. So the running time complexity of this approach is O(nlogn).

Answer 3

If your lists don't change often, you can sort both lists and then do the merge in linear time by simply walking both lists.

If your lists are changing all the time, you're probably better off making the "small" container sorted, such as map or set . In that case just use find on the set for each item in the big list you want to join.