还有一个O(N^(3/2))非最优解。不过，这很有趣。

首先预处理大小为N^0.5的每个部分乘法(这在O(N)时间复杂度中完成)。然后，计算每个数字的其他值的倍数可以在2*O(N^0.5)时间内完成(为什么?因为您只需要将其他((N^0.5) - 1)数字的最后一个元素相乘，并将结果与属于当前数字组的((N^0.5) - 1)数字相乘。对每一个数都这样做，可以得到O(N^(3/2))时间。

例子:

4, 6, 7, 2, 3, 1, 9, 5, 8

部分结果: 4*6*7 = 168 2*3*1 = 6 9*5*8 = 360

要计算3的值，需要将其他组的值乘以168*360，然后乘以2*1。

我习惯使用c#:

    public int[] ProductExceptSelf(int[] nums)
    {
        int[] returnArray = new int[nums.Length];
        List<int> auxList = new List<int>();
        int multTotal = 0;

        // If no zeros are contained in the array you only have to calculate it once
        if(!nums.Contains(0))
        {
            multTotal = nums.ToList().Aggregate((a, b) => a * b);

            for (int i = 0; i < nums.Length; i++)
            {
                returnArray[i] = multTotal / nums[i];
            }
        }
        else
        {
            for (int i = 0; i < nums.Length; i++)
            {
                auxList = nums.ToList();
                auxList.RemoveAt(i);
                if (!auxList.Contains(0))
                {
                    returnArray[i] = auxList.Aggregate((a, b) => a * b);
                }
                else
                {
                    returnArray[i] = 0;
                }
            }
        }            

        return returnArray;
    }

试试这个!

import java.util.*;
class arrProduct
{
 public static void main(String args[])
     {
         //getting the size of the array
         Scanner s = new Scanner(System.in);
            int noe = s.nextInt();

        int out[]=new int[noe];
         int arr[] = new int[noe];

         // getting the input array
         for(int k=0;k<noe;k++)
         {
             arr[k]=s.nextInt();
         }

         int val1 = 1,val2=1;
         for(int i=0;i<noe;i++)
         {
             int res=1;

                 for(int j=1;j<noe;j++)
                 {
                if((i+j)>(noe-1))
                {

                    int diff = (i+j)-(noe);

                    if(arr[diff]!=0)
                    {
                    res = res * arr[diff];
                    }
                }

                else
                {
                    if(arr[i+j]!=0)
                    {
                    res= res*arr[i+j];
                    }
                }


             out[i]=res;

         }
         }

         //printing result
         System.out.print("Array of Product: [");
         for(int l=0;l<out.length;l++)
         {
             if(l!=out.length-1)
             {
            System.out.print(out[l]+",");
             }
             else
             {
                 System.out.print(out[l]);
             }
         }
         System.out.print("]");
     }

}

上下两次。在O(N)完成的工作

private static int[] multiply(int[] numbers) {
        int[] multiplied = new int[numbers.length];
        int total = 1;

        multiplied[0] = 1;
        for (int i = 1; i < numbers.length; i++) {
            multiplied[i] = numbers[i - 1] * multiplied[i - 1];
        }

        for (int j = numbers.length - 2; j >= 0; j--) {
            total *= numbers[j + 1];
            multiplied[j] = total * multiplied[j];
        }

        return multiplied;
    }

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。

这里有一个小的递归函数(在c++中)来进行修改。它需要O(n)额外的空间(在堆栈上)。假设数组在a中，N表示数组长度，我们有:

int multiply(int *a, int fwdProduct, int indx) {
    int revProduct = 1;
    if (indx < N) {
       revProduct = multiply(a, fwdProduct*a[indx], indx+1);
       int cur = a[indx];
       a[indx] = fwdProduct * revProduct;
       revProduct *= cur;
    }
    return revProduct;
}