You are given two non-increasing 0-indexed integer arrays nums1 and nums2.
A pair of indices (i, j), where 0 <= i < nums1.length and 0 <= j < nums2.length, is valid if both i <= j and nums1[i] <= nums2[j]. The distance of the pair is j - i.
Return the maximum distance of any valid pair (i, j). If there are no valid pairs, return 0.
An array arr is non-increasing if arr[i-1] >= arr[i] for every 1 <= i < arr.length.
Example 1:
Input: nums1 = [55,30,5,4,2], nums2 = [100,20,10,10,5]
Output: 2
Explanation: The valid pairs are (0,0), (2,2), (2,3), (2,4), (3,3), (3,4), and (4,4).
The maximum distance is 2 with pair (2,4).
Example 2:
Input: nums1 = [2,2,2], nums2 = [10,10,1]
Output: 1
Explanation: The valid pairs are (0,0), (0,1), and (1,1).
The maximum distance is 1 with pair (0,1).
Example 3:
Input: nums1 = [30,29,19,5], nums2 = [25,25,25,25,25]
Output: 2
Explanation: The valid pairs are (2,2), (2,3), (2,4), (3,3), and (3,4).
The maximum distance is 2 with pair (2,4).
Constraints:
1 <= nums1.length, nums2.length <= 105
1 <= nums1[i], nums2[j] <= 105
Both nums1 and nums2 are non-increasing.
Solutions
Solution 1: Binary Search
Thinking
Both arrays are nonincreasing; we want the largest \(j-i\) with \(i\le j\) and \(\textit{nums1}[i]\le \textit{nums2}[j]\). Trying every pair is \(O(mn)\) and too slow for \(n\le 10^5\).
For a fixed \(i\), the farthest legal \(j\) is the last index in \(\textit{nums2}[i:]\) that is at least \(\textit{nums1}[i]\). Reversing \(\textit{nums2}\) lets us find that index by binary search.
Assume the lengths of \(\textit{nums1}\) and \(\textit{nums2}\) are \(m\) and \(n\) respectively.
Traverse array \(\textit{nums1}\), for each number \(\textit{nums1}[i]\), perform a binary search for numbers in \(\textit{nums2}\) in the range \([i,n)\), find the last position \(j\) that is greater than or equal to \(\textit{nums1}[i]\), calculate the distance between this position and \(i\), and update the maximum distance value \(\textit{ans}\).
The time complexity is \(O(m \log n)\), and the space complexity is \(O(1)\), where \(m\) and \(n\) are the lengths of \(\textit{nums1}\) and \(\textit{nums2}\) respectively.
Solution 1 uses only the monotonicity of \(\textit{nums2}\). \(\textit{nums1}\) is also nonincreasing, so the farthest \(j\) never moves left as \(i\) grows. Two pointers advance together in linear time.