惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

SecWiki News
SecWiki News
I
InfoQ
The Cloudflare Blog
人人都是产品经理
人人都是产品经理
博客园 - Franky
T
Tailwind CSS Blog
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
量子位
博客园_首页
罗磊的独立博客
V
V2EX
李成银的技术随笔
大猫的无限游戏
大猫的无限游戏
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
T
True Tiger Recordings
Vercel News
Vercel News
Cyberwarzone
Cyberwarzone
Cisco Talos Blog
Cisco Talos Blog
F
Fox-IT International blog
D
Darknet – Hacking Tools, Hacker News & Cyber Security
M
Microsoft Research Blog - Microsoft Research
Know Your Adversary
Know Your Adversary
爱范儿
爱范儿
The Register - Security
The Register - Security
G
Google Developers Blog
The Hacker News
The Hacker News
Malwarebytes
Malwarebytes
S
Securelist
博客园 - 三生石上(FineUI控件)
Jina AI
Jina AI
T
Threat Research - Cisco Blogs
T
The Exploit Database - CXSecurity.com
S
SegmentFault 最新的问题
博客园 - 叶小钗
F
Fortinet All Blogs
Apple Machine Learning Research
Apple Machine Learning Research
宝玉的分享
宝玉的分享
博客园 - 聂微东
T
Threatpost
博客园 - 【当耐特】
D
Docker
P
Privacy & Cybersecurity Law Blog
www.infosecurity-magazine.com
www.infosecurity-magazine.com
G
GRAHAM CLULEY
V
Visual Studio Blog
C
Cisco Blogs
IT之家
IT之家
S
Security Archives - TechRepublic
Latest news
Latest news
阮一峰的网络日志
阮一峰的网络日志

jdhao's digital space

Manage uv.lock file with Renovate Set up Python Provider for Neovim Ripgrep Config to Search Hidden Files Pre-commit Setup for Your Project I read the nvim v0.12 release note so you don't have to Return Different Values for Each Call of A Mock Migrate Python Project from Pip to Uv 德语常用不规则动词 葱油鸡腿制作 Check Trailing White Spaces in Your Project 菜谱:茄子肉丁 object vs nested type in data mapping in Elasticsearch Node, Index, Shard in Elasticsearch Logging setup for Pytest Select fields in Elasticsearch: _source, fields and stored_fields 中式葱花饼制作 菜谱: 凉拌苤蓝(卜留克/kohlrabi) 我也有高考 PTSD Garmin Course Syncing Not Working? Prevent Accidental Index Delete in Elasticsearch How to Import GPX File into Garmin Watch Python system PATH issues When We Use Pytest 菜谱:泰式打抛牛肉 菜谱:烤箱羊肉串 How to Filter Warnings in Python/pytest 家常烤箱烤鸡腿 Comparison between Several Desktop Speakers How to Use LuaRocks Package in Neovim Macbook 外接显示器 家常萝卜炖羊排 Run the Job Immediately after Starting Scheduler in Python APScheduler Retry for Google Cloud Client 菜谱:土豆金枪鱼沙拉 菜谱:椰香咖喱鸡 凉拌绿豆宽粉制作 Make Python logging Work in GCP Liveness and Readiness Check in Kubernetes Notes on Using GCP Logging 西班牙土豆饼制作 Elasticsearch Version Conflict Error How to Use the Elasticsearch task API Speed up document indexing in Elasticsearch via bulk indexing Index refresh issue in Elasticsearch Google Cloud Storage Usage 家常煎羊排制作 凉拌茄子制作 Configure Python logging with dictConfig Debugging Wezterm Issues Black Formatter Setup for Python Project Git line ending config Garmin Forerunner 965 Essential Tips and Setups How to Download Files from Google Cloud Storage in the Databricks Workspace Notebook Databricks Cli Usage Working with Databricks Workspace Files 手抓羊肉饭制作 Databricks Init Scripts Using Virutal Environment in Python with venv File Systems in Databricks LATERAL VIEW EXPLODE in Spark 菜谱:麻婆豆腐 在德国做台湾卤肉饭 FastAPI testing and OpenAPI doc generation Change Timezone in Databricks Spark How to Profile Your Python Script/Module 菜谱:茄子肉沫 Migrating from Packer.nvim to Lazy.nvim How to Extract PDF file on macOS How to Deploy Fastapi Application with Docker Nerdfont Icon Missing after Wezterm Upgrade Pylsp setup for Neovim in 2023 How to Parse Query Param With Multiple Values in FastAPI 菜谱:土豆胡萝卜烧牛肉 Zsh Startup Files in macOS PATH Variable Changed inside Tmux on macOS? Work with JSON File in Neovim Running/importing Python code/module in Databricks Agile and Scrum 菜谱:凉拌牛肉 Awesome Command Line Tools Written in Rust How to get or set Databricks spark configuration Set Up German Version macOS Add A Custom Search Engine for Vimium 中国大陆小米手机如何使用 Google Pay 春节回乡记 滇西之行 2023 贵阳行 2023 程序员海外工作---语言篇 2023 长沙行 2023 西安行 德国工签申请指南 2022 年博客回顾 感染 omicron 记录 How to Override Default Options in Neovim Variadic Arguments in Lua How to Enable Method Autocompletion for OpenCV How to Read Local CSV File to Table in MySQL I read the nvim v0.8 release note so you do not have to Creating A Trigger in PostgreSQL Cost of Living in Shenzhen You Do Not Need a Plugin for This Feature
快排(quick sort) C++ 实现
2021-07-10 · via jdhao's digital space

快排是一种快速排序算法,原理是从数组中选择一个 pivot,数组中小于等于 pivot 的元素放到左边,大于 pivot 的元素放到右边。然后对左边和右边数组分别递归进行快速排序,最后整个数组就成为排序好的数组。

非原地算法#

下面是 quicksort 的一种实现方法,非原地操作,每次会申请额外的空间存放左子数组和右子数组,但是理解上非常简单,没有任何难理解的地方:

#include <iostream>
#include <random>
#include <string>
#include <vector>
#include <ctime>

using std::random_device;
using std::cout;
using std::endl;
using std::vector;
using std::string;

void printList(vector<int>&, const string&);

vector<int> quickSort(vector<int> arr){
  if (arr.size() <= 1){
    return arr;
  }

  int N = arr.size();
  // arr[mid] 是 pivot
  int mid = (N-1) / 2;

  vector<int> left;
  vector<int> right;

  for (int i = 0; i < N; i++){
    if (i == mid) continue;

    // 对于原始数组,小于等于 pivot 的元素放左子数组,大于 pivot 的元素放右子数组
    if (arr[i] <= arr[mid]){
      left.push_back(arr[i]);
    }
    else{
      right.push_back(arr[i]);
    }
  }

  // 对左右子数组,再分别进行快排
  left = quickSort(left);
  right = quickSort(right);

  vector<int> new_arr;
  for (auto & x: left){
    new_arr.push_back(x);
  }
  new_arr.push_back(arr[mid]);
  for (auto & x: right){
    new_arr.push_back(x);
  }

  return new_arr;
}

// Generate a random sequence of length len, in range(low, high) (inclusive).
// need to #include<random>
vector<int> genRandom(int low, int high, int len){
  std::random_device rd;
  std::mt19937 gen(rd());
  std::uniform_int_distribution<int> distribution(low, high);

  vector<int> arr(len, 0);
  for (int i = 0; i != len; ++i){
      arr[i] = distribution(gen);
  }

  return arr;
}

void printList(vector<int>& arr, const string& desc){
  cout << desc << ": ";

  for (auto it = arr.begin(); it != arr.end(); it++){
    cout << *it << ((it != arr.end()-1) ? ' ' : '\n');
  }
}

int main()
{
  vector<int> arr = genRandom(1, 1000, 10);
  cout << "Before quick sort\n";
  printList(arr, "arr");

  auto arr1 = quickSort(arr);
  cout << "After quick sort\n";
  printList(arr1, "arr1");

  return 0;
}

快排实现一个坑,如果把数列中间元素作为 pivot,pivot 不能合并到左边的数组里面,否则会造成无限循环,pivot 必须单独拿出来。举个例子,假如数组为 arr = {2, 1}, pivot 是 arr[0],如果把小于等于 pivot 的元素放在左边数组,把大于 pivot 的元素放在右边数组,结果会是左边 {2, 1}, 右边 {}, 然后左右两边数组又调用 quickSort(),quickSort 终止条件是 arr.size() <= 1,左边的数组会继续重复上述的过程,永远循环下去。

正确的做法是把 pivot 单独拿出来,然后对左边数组进行 quickSort 递归,再对右边数组进行 quickSort 递归,然后把排序好的左数组,pivot,排序好的右数组合并起来。还以上面 case 为例,pivot 是 arr[0],左边数组 {1}, 右边数组 {}, 左右数组都达到终止条件返回,然后把左数组,pivot,右数组拼起来,排序好的数组就是 {1, 2}。

原地算法#

还有一种原地实现的算法,不需要开辟额外的空间,不过没有上面的实现那么直观容易理解。

#include <iostream>
#include <vector>
#include <random>

using namespace std;

// Generate a random sequence of length len, in range(low, high) (inclusive).
// need to #include<random>
vector<int> genRandom(int low, int high, int len){
  std::random_device rd;
  std::mt19937 gen(rd());
  std::uniform_int_distribution<int> distribution(low, high);

  vector<int> arr(len, 0);
  for (int i = 0; i != len; ++i){
      arr[i] = distribution(gen);
  }

  return arr;
}

int partion(vector<int> &arr, int low, int high) {
  int pivot = arr[low];
  int i = low;

  for (int j = low+1; j < high; j++){
    if (arr[j] <= pivot){
      i++;
      swap(arr[i], arr[j]);
    }
  }
  // 经过上述 for 循环, i 以及 i 左边都是小于等于 pivot 的元素,最后把 arr[i] 和
  // pivot 元素位置调换一下即可。即可保证我们完成了 partion
  swap(arr[i], arr[low]);

  return i;
}
void quickSort(vector<int>& arr, int low, int high){
  // low < high,才需要排序
  if (low < high){
    int r = partion(arr, low, high);

    quickSort(arr, low, r);
    quickSort(arr, r+1, high);
  }
}

int main() {
  vector<int> arr = genRandom(1, 100, 5);

   for(auto e: arr)
        cout<< e <<" ";
    cout<< endl;

  int N = arr.size();
  quickSort(arr, 0, N);

   for(auto e: arr)
        cout<< e <<" ";
    cout<< endl;

  return 0;
}

上述实现中,最核心的是 partion() 函数,partion 函数选择第一个元素作为 pivot,然后把 pivot 移动到最终它应该在的位置,并保证该位置左边都是小于等于 pivot 的元素,右边都是大于 pivot 的元素,最后返回 pivot 元素在新数组中的位置。pivot 不一定非要选择第一个元素,也可以选择最后一个元素,逻辑对应改动即可。其实 pivot 也可以选择其他元素,不过实现起来感觉比较麻烦,想了很久,也没实现对,总有一些 edge case 没处理正确。

参考#