将Michael Anderson的解决方案翻译成Haskell:

otherProducts xs = zipWith (*) below above

     where below = scanl (*) 1 $ init xs

           above = tail $ scanr (*) 1 xs

以下是线性O(n)时间内的简单Scala版本:

def getProductEff(in:Seq[Int]):Seq[Int] = {

   //create a list which has product of every element to the left of this element
   val fromLeft = in.foldLeft((1, Seq.empty[Int]))((ac, i) => (i * ac._1, ac._2 :+ ac._1))._2

   //create a list which has product of every element to the right of this element, which is the same as the previous step but in reverse
   val fromRight = in.reverse.foldLeft((1,Seq.empty[Int]))((ac,i) => (i * ac._1,ac._2 :+ ac._1))._2.reverse

   //merge the two list by product at index
   in.indices.map(i => fromLeft(i) * fromRight(i))

}

这是可行的，因为本质上答案是一个数组，它是左右所有元素的乘积。

我们正在分解数组的元素，首先从下标之前开始，即前缀，然后是下标或后缀之后

class Solution:

   def productExceptSelf(nums):

      length = len(nums)


      result = [1] * length


      prefix_product = 1


      postfix_product = 1

# we initialize the result and products


      for i in range(length)

      result[i] *= prefix_product


       prefix_product *= nums[i]

#we multiply the result by each number before the index

      for i in range(length-1,-1,-1)

      result[i] *= postfix_product


      postfix_product *= nums[i]

#same for after index
   return result

抱歉，走路时用手机

下面是我尝试用Java来解决这个问题。抱歉格式不规范，但代码有很多重复，这是我能做的最好的，使它可读。

import java.util.Arrays;

public class Products {
    static int[] products(int... nums) {
        final int N = nums.length;
        int[] prods = new int[N];
        Arrays.fill(prods, 1);
        for (int
           i = 0, pi = 1    ,  j = N-1, pj = 1  ;
           (i < N)         && (j >= 0)          ;
           pi *= nums[i++]  ,  pj *= nums[j--]  )
        {
           prods[i] *= pi   ;  prods[j] *= pj   ;
        }
        return prods;
    }
    public static void main(String[] args) {
        System.out.println(
            Arrays.toString(products(1, 2, 3, 4, 5))
        ); // prints "[120, 60, 40, 30, 24]"
    }
}

循环不变量为pi = nums[0] * nums[1] *..* nums[N-2] *..num [j + 1]。左边的i部分是“前缀”逻辑，右边的j部分是“后缀”逻辑。

递归一行程序

Jasmeet给出了一个(漂亮的!)递归解;我把它变成了这样(可怕!)Java一行程序。它进行就地修改，堆栈中有O(N)个临时空间。

static int multiply(int[] nums, int p, int n) {
    return (n == nums.length) ? 1
      : nums[n] * (p = multiply(nums, nums[n] * (nums[n] = p), n + 1))
          + 0*(nums[n] *= p);
}

int[] arr = {1,2,3,4,5};
multiply(arr, 1, 0);
System.out.println(Arrays.toString(arr));
// prints "[120, 60, 40, 30, 24]"

这是我的代码:

int multiply(int a[],int n,int nextproduct,int i)
{
    int prevproduct=1;
    if(i>=n)
        return prevproduct;
    prevproduct=multiply(a,n,nextproduct*a[i],i+1);
    printf(" i=%d > %d\n",i,prevproduct*nextproduct);
    return prevproduct*a[i];
}

int main()
{
    int a[]={2,4,1,3,5};
    multiply(a,5,1,0);
    return 0;
}

预先计算每个元素左右两边数字的乘积。 对于每个元素，期望值都是它相邻元素乘积的乘积。

#include <stdio.h>

unsigned array[5] = { 1,2,3,4,5};

int main(void)
{
unsigned idx;

unsigned left[5]
        , right[5];
left[0] = 1;
right[4] = 1;

        /* calculate products of numbers to the left of [idx] */
for (idx=1; idx < 5; idx++) {
        left[idx] = left[idx-1] * array[idx-1];
        }

        /* calculate products of numbers to the right of [idx] */
for (idx=4; idx-- > 0; ) {
        right[idx] = right[idx+1] * array[idx+1];
        }

for (idx=0; idx <5 ; idx++) {
        printf("[%u] Product(%u*%u) = %u\n"
                , idx, left[idx] , right[idx]  , left[idx] * right[idx]  );
        }

return 0;
}

结果:

$ ./a.out
[0] Product(1*120) = 120
[1] Product(1*60) = 60
[2] Product(2*20) = 40
[3] Product(6*5) = 30
[4] Product(24*1) = 24

(更新:现在我仔细看，这使用与Michael Anderson, Daniel Migowski和上面的聚基因润滑剂相同的方法)