Implement pow(x, n).
题解:注意两点:
- 普通的递归把n降为n-1会超时,要用二分的方法,每次把xn = x[n/2] * x[n/2] * xn-[n/2]*2, [n/2]表示n除以2下取整。
- n有可能取负数,负数的时候,先计算pow(x,-n),然后返回1/pow(x,-n);
代码如下:
1 public class Solution { 2 public double pow(double x, int n) { 3 if(n == 0) 4 return 1; 5 if(n == 1) 6 return x; 7 8 boolean isNeg = false; 9 10 //if n is negative,we first caculate 1/pow(x,n) = pow(x,-n) 11 if(n<0){ 12 isNeg = true; 13 n *= -1; 14 } 15 16 //binary 17 int mid = n /2; 18 int left = n - mid*2; 19 double front = pow(x, mid); 20 double end = pow(x, left); 21 22 if(isNeg) 23 return 1/(front*front*end); 24 else { 25 return front*front*end; 26 } 27 } 28 }