Count permutations with exactly n inversions
Company: Turing
Role: Software Engineer
Category: Coding & Algorithms
Difficulty: medium
Interview Round: Technical Screen
You are given two integers **m** and **n**.
Consider all permutations of the array `[1, 2, ..., m]` (each number from 1 to m appears exactly once). A permutation `perm` (0-indexed) has an **inversion** pair `(x, y)` if:
- `0 <= x < y < m`, and
- `perm[x] > perm[y]`
A permutation is called **special** if it contains **exactly `n` inversions**.
Return the **number of special permutations** modulo **10^9 + 7**.
#### Examples
1) **Input:** `m = 3, n = 0`
**Output:** `1`
**Explanation:** Only `[1, 2, 3]` has 0 inversions.
2) **Input:** `m = 3, n = 1`
**Output:** `2`
**Explanation:** `[1, 3, 2]` and `[2, 1, 3]` each have exactly 1 inversion.
#### Constraints
- `1 <= m <= 1000`
- `0 <= n <= 1000`
Quick Answer: This Coding & Algorithms question evaluates combinatorial counting and algorithmic skills related to permutations and inversion counts, emphasizing dynamic programming and modular arithmetic for large-result handling.
You are given two integers `m` and `n`.
Consider all permutations of the array `[1, 2, ..., m]` (each number from 1 to m appears exactly once). A permutation `perm` (0-indexed) has an **inversion** pair `(x, y)` if:
- `0 <= x < y < m`, and
- `perm[x] > perm[y]`
A permutation is called **special** if it contains **exactly `n` inversions**.
Return the **number of special permutations** modulo **10^9 + 7**.
#### Examples
1) **Input:** `m = 3, n = 0` -> **Output:** `1` (only `[1, 2, 3]` has 0 inversions).
2) **Input:** `m = 3, n = 1` -> **Output:** `2` (`[1, 3, 2]` and `[2, 1, 3]`).
#### Constraints
- `1 <= m <= 1000`
- `0 <= n <= 1000`
#### Approach
Let `dp[i][j]` be the number of permutations of `i` elements with exactly `j` inversions. Inserting the `i`-th element into a permutation of `i-1` elements can create between `0` and `i-1` new inversions, so `dp[i][j] = sum_{k=0}^{min(j, i-1)} dp[i-1][j-k]`. The inner sum is a sliding window, so maintain a prefix-sum array to evaluate each state in O(1), giving an overall O(m*n) time / O(n) space algorithm.
Constraints
- 1 <= m <= 1000
- 0 <= n <= 1000
- Answer is returned modulo 10^9 + 7
Examples
Input: (3, 0)
Expected Output: 1
Explanation: Only [1, 2, 3] has 0 inversions.
Input: (3, 1)
Expected Output: 2
Explanation: [1, 3, 2] and [2, 1, 3] each have exactly 1 inversion.
Hints
- Define dp[i][j] = number of permutations of i elements with exactly j inversions. The answer is dp[m][n].
- Inserting the largest element into a permutation of i-1 elements at the position that leaves k elements to its right adds exactly k new inversions, with 0 <= k <= i-1. So dp[i][j] = sum_{k=0}^{min(j, i-1)} dp[i-1][j-k].
- The inner summation is a sliding window of width i over the previous row. Use a prefix-sum array to compute each new[j] in O(1), reducing the total work to O(m*n). Take care with the modulo on subtraction (add MOD or rely on Python's non-negative modulo).