3

Problem: 3SUM

Input: Three lists A, B and C of integers and an integer k. Each list contain $n$ numbers. Task: Decide whether there exists a tuple (a, b, c) ∈ A × B × C such that a + b + c = k.

Question : Is it possible to solve 3SUM in $O(n^2)$ time using constant space ? Prove or disprove it

Complexity
  • 1,187
  • 9
  • 23

2 Answers2

5

A $O(n^2)$ algorithm (with $O(1)$ space) is as follows:

  • Sort $A$, $B$, and $C$ individually in $O(n \log n)$.
  • For each $a \in A$:
    • Search a pair of $b \in B$ and $c \in C$ such that $b + c = k - a$. This can be done in $O(n)$ by traversing $B$ from the smallest to the largest and $C$ from the largest to the smallest. (Tip: Comparing $b + c$ with $k-a$ each time.)
hengxin
  • 9,541
  • 3
  • 36
  • 73
2

This question is not a 3sum problem. The question can be solved with hash tables. For each element in C traverse A and see if the corresponding element exists in B. Time complexity would be $O(|C|*|A|*1) = O(|C||A|)$

Edit:
Answer to the updated question:

This problem can be treated as 2SUM problem for each element in C (or any other list as they are symmetrical). That is for each $c \in C$ we find solutions for the sum $k - c$ using elements from A and B.

For a $O(1)$ auxiliary space algorithm one can first sort A and B. Then use 2-pointer algorithm (Tutorial) to solve the 2SUM problem in $O(n)$.

Comment if you have doubts on implementation. You can also see https://cs.stackexchange.com/a/13586/81272

Nilesh Hirani
  • 61
  • 6