我在一次工作面试中被问到这个问题，我想知道其他人是如何解决这个问题的。我最擅长使用Java，但也欢迎使用其他语言的解决方案。
给定一个数字数组nums，返回一个数字数组products，其中products[i]是所有nums[j]的乘积，j != i。 输入:[1,2,3,4,5] 输出:[(2 * 3 * 4 * 5),(1 * 3 * 4 * 5),(1 * 2 * 4 * 5),(1 * 2 * 3 * 5),(1 * 2 * 3 * 4)] = [120, 60, 40, 30, 24] 你必须在O(N)中不使用除法来做这个。
当前回答
JavaScript中使用reduce的变体
const getProduct = arr => arr。Reduce ((acc, value) => acc * value); const arrayWithExclusion = (arr, node) => 加勒比海盗。Reduce ((acc, val, j) => (node !== j ?)[…Acc, val]: Acc)， []); const getproductwitheexclusion = arr => { Let result = []; 对于(设I = 0;I < arrr .length;I += 1) { 结果。推动(getProduct (arrayWithExclusion(加勒比海盗,我))); ｝ 返回结果; }；
其他回答
在这里添加我的javascript解决方案，因为我没有发现任何人建议这样做。 除法是什么，除了数从另一个数中得到一个数的次数吗?我计算了整个数组的乘积，然后遍历每个元素，并减去当前元素直到0:
//No division operation allowed
// keep substracting divisor from dividend, until dividend is zero or less than divisor
function calculateProducsExceptCurrent_NoDivision(input){
var res = [];
var totalProduct = 1;
//calculate the total product
for(var i = 0; i < input.length; i++){
totalProduct = totalProduct * input[i];
}
//populate the result array by "dividing" each value
for(var i = 0; i < input.length; i++){
var timesSubstracted = 0;
var divisor = input[i];
var dividend = totalProduct;
while(divisor <= dividend){
dividend = dividend - divisor;
timesSubstracted++;
}
res.push(timesSubstracted);
}
return res;
}
//这是Java中的递归解决方案 //从main product(a,1,0)调用如下;
public static double product(double[] a, double fwdprod, int index){
double revprod = 1;
if (index < a.length){
revprod = product2(a, fwdprod*a[index], index+1);
double cur = a[index];
a[index] = fwdprod * revprod;
revprod *= cur;
}
return revprod;
}
下面是一个使用c#的函数式示例:
Func<long>[] backwards = new Func<long>[input.Length];
Func<long>[] forwards = new Func<long>[input.Length];
for (int i = 0; i < input.Length; ++i)
{
var localIndex = i;
backwards[i] = () => (localIndex > 0 ? backwards[localIndex - 1]() : 1) * input[localIndex];
forwards[i] = () => (localIndex < input.Length - 1 ? forwards[localIndex + 1]() : 1) * input[localIndex];
}
var output = new long[input.Length];
for (int i = 0; i < input.Length; ++i)
{
if (0 == i)
{
output[i] = forwards[i + 1]();
}
else if (input.Length - 1 == i)
{
output[i] = backwards[i - 1]();
}
else
{
output[i] = forwards[i + 1]() * backwards[i - 1]();
}
}
我不完全确定这是O(n)，因为所创建的Funcs是半递归的，但我的测试似乎表明它在时间上是O(n)。
import java.util.Arrays;
public class Pratik
{
public static void main(String[] args)
{
int[] array = {2, 3, 4, 5, 6}; // OUTPUT: 360 240 180 144 120
int[] products = new int[array.length];
arrayProduct(array, products);
System.out.println(Arrays.toString(products));
}
public static void arrayProduct(int array[], int products[])
{
double sum = 0, EPSILON = 1e-9;
for(int i = 0; i < array.length; i++)
sum += Math.log(array[i]);
for(int i = 0; i < array.length; i++)
products[i] = (int) (EPSILON + Math.exp(sum - Math.log(array[i])));
}
}
输出:
[360, 240, 180, 144, 120]
时间复杂度:O(n) 空间复杂度:O(1)
我的第一次尝试，用Python。O (2 n):
def product(l):
product = 1
num_zeroes = 0
pos_zero = -1
# Multiply all and set positions
for i, x in enumerate(l):
if x != 0:
product *= x
l[i] = 1.0/x
else:
num_zeroes += 1
pos_zero = i
# Warning! Zeroes ahead!
if num_zeroes > 0:
l = [0] * len(l)
if num_zeroes == 1:
l[pos_zero] = product
else:
# Now set the definitive elements
for i in range(len(l)):
l[i] = int(l[i] * product)
return l
if __name__ == "__main__":
print("[0, 0, 4] = " + str(product([0, 0, 4])))
print("[3, 0, 4] = " + str(product([3, 0, 4])))
print("[1, 2, 3] = " + str(product([1, 2, 3])))
print("[2, 3, 4, 5, 6] = " + str(product([2, 3, 4, 5, 6])))
print("[2, 1, 2, 2, 3] = " + str(product([2, 1, 2, 2, 3])))
输出:
[0, 0, 4] = [0, 0, 0]
[3, 0, 4] = [0, 12, 0]
[1, 2, 3] = [6, 3, 2]
[2, 3, 4, 5, 6] = [360, 240, 180, 144, 120]
[2, 1, 2, 2, 3] = [12, 24, 12, 12, 8]
