【Lintcode】191. Maximum Product Subarray

题目地址：

https://www.lintcode.com/problem/maximum-product-subarray/description

给定一个长$n$的数组$A$，求其最大乘积子数组，返回那个最大乘积。

思路是动态规划。设$f[i]$是以$A[i]$结尾的最大乘积子数组的乘积，那么如果$A[i]\ge 0$，则 f [ i ] = max ⁡ { A [ i ] , A [ i ] f [ i − 1 ] } f[i]=\max\{A[i],A[i]f[i-1]\} f[i]=max{A[i],A[i]f[i−1]}，但如果$A[i]<0$，那么$f[i]$应该是或者就是$A[i]$本身，或者由$A[i]$乘以以$A[i-1]$为结尾的最小乘积子数组的乘积得到，所以我们还需要开一个数组$g$，$g[i]$存以$A[i]$结尾的最小乘积子数组的乘积。这样就有 { f [ i ] = max ⁡ { A [ i ] , A [ i ] f [ i − 1 ] , A [ i ] g [ i − 1 ] } g [ i ] = min ⁡ { A [ i ] , A [ i ] f [ i − 1 ] , A [ i ] g [ i − 1 ] } \begin{cases}f[i]=\max\{A[i],A[i]f[i-1],A[i]g[i-1]\}\\g[i]=\min\{A[i],A[i]f[i-1],A[i]g[i-1]\}\end{cases} {f[i]=max{A[i],A[i]f[i−1],A[i]g[i−1]}g[i]=min{A[i],A[i]f[i−1],A[i]g[i−1]}​初始条件$f[0]=g[0]=A[0]$，答案就是$f$和$g$所有数字的最大值。代码如下：

public class Solution {
    /**
     * @param nums: An array of integers
     * @return: An integer
     */
    public int maxProduct(int[] nums) {
        // write your code here
        int res = Integer.MIN_VALUE;
        int[] maxdp = new int[nums.length], mindp = new int[nums.length];
        maxdp[0] = mindp[0] = nums[0];
        res = nums[0];
        for (int i = 1; i < nums.length; i++) {
            int a  = nums[i], b = nums[i] * maxdp[i - 1], c = nums[i] * mindp[i - 1];
            maxdp[i] = Math.max(a, Math.max(b, c));
            mindp[i] = Math.min(a, Math.min(b, c));
            
            res = Math.max(res, Math.max(maxdp[i], mindp[i]));
        }
        
        return res;
    }
}

时空复杂度$O(n)$。