#!/usr/bin/env python # coding=utf-8 ################################################################ # LeetCode: 69 Sqrt(x) # Implement int sqrt(int x). # Compute and return the square root of x, where x is guaranteedto be a non-negative integer. # Since the return type is an integer, the decimal digits aretruncated and only the integer part of the result is returned. ################################################################ import math class Solution: def mySqrt(self, x): """ (Knowledge) 函数功能描述: 传入一个数字,返回其平方根的整数部分 思路: 1. 首先用特判,处理掉x=1的特殊情况; 2. 接着用二分法,找到x平方根的整数部分; 二分法的结束条件:跳出循环之前的最后一次循环,left == right, 此时得到 mid == left,因此: - 如果mid > x * mid,则 x 的平方根必然为比left略小一点的某个值,所以最后结果返回 left - 1 - 如果mid <= x * mid, 则 x 的平方根必然为比left略大一点某个值,又因为执行了 left = mid +1, 所以最后结果返回 left - 1 """ if x < 2: return x left, right = 1, x // 2 while left <= right: mid = left + (right - left) // 2 if mid > x / mid: right = mid - 1 else: left = mid + 1 return left - 1 def mySqrt2(self, x): """ (Knowledge) 算法说明:使用牛顿迭代法,近似求平方根 牛顿迭代法求平方跟的证明详见 => https://www.cnblogs.com/upcan/p/9907402.html 最终得到迭代公式: next = 1/2 * (next + input / next) - next => 等式左边为下一个近似值,等式右的next表示上一个近似值 - input => 输入值 """ # 首先猜测一个初始的预测值 next = x / 2 # 运行20次迭代(这个迭代次数取决于需要的精度,迭代次数越多,得到的结果越精确) for k in range(20): next = 1 / 2 * (next + x / next) # 根据题意,进行向下取整 return math.floor(next) if __name__ == '__main__': solution = Solution() print(solution.mySqrt(9)) print(solution.mySqrt(8)) print(solution.mySqrt(10)) print(solution.mySqrt(7)) print(solution.mySqrt2(9)) print(solution.mySqrt2(8)) print(solution.mySqrt2(10)) print(solution.mySqrt2(7))