C# 优先级队列

前6行是优先队列,后6行是C#原生的queue
Min Heap Priority Queue
The above code is not really a true Priority Queue as it does not allow duplicate keys; also, the SortedList on which it is based does not have O(log n) insertions and removals for random data as a true Priority Queue does. The below code implements a true Min Heap Priority Queue:
namespace PriorityQ {
using KeyT = UInt32;
using System;
using System.Collections.Generic;
using System.Linq;
class Tuple<K, V> { // for DotNet 3.5 without Tuple's
public K Item1; public V Item2;
public Tuple(K k, V v) { Item1 = k; Item2 = v; }
public override string ToString() {
return "(" + Item1.ToString() + ", " + Item2.ToString() + ")";
}
}
class MinHeapPQ<V> {
private struct HeapEntry {
public KeyT k; public V v;
public HeapEntry(KeyT k, V v) { this.k = k; this.v = v; }
}
private List<HeapEntry> pq;
private MinHeapPQ() { this.pq = new List<HeapEntry>(); }
; } }
private int sz {
get {
var cnt = pq.Count;
) ? : cnt - ;
}
}
private Tuple<KeyT, V> pkmn {
get {
) return null;
else {
];
return new Tuple<KeyT, V>(mn.k, mn.v);
}
}
}
private void psh(KeyT k, V v) { // add extra very high item if none
) pq.Add(new HeapEntry(UInt32.MaxValue, v));
]); // copy bottom item...
; ni > ; i >>= , ni >>= ) {
];
] = t; else break;
}
pq[i - ] = new HeapEntry(k, v);
}
private void siftdown(KeyT k, V v, int ndx) {
; var i = ndx;
; ni < cnt; ni = ni + ni + ) {
].k;
var nk = k;
if (k > lk) { i = ni; nk = lk; }
; i = ni; }
if (i != oi) pq[oi] = pq[i]; else break;
}
pq[i] = new HeapEntry(k, v);
}
private void rplcmin(KeyT k, V v) {
) siftdown(k, v, );
}
private void dltmin() {
;
) pq.Clear();
else {
var lkv = pq[lsti];
pq.RemoveAt(lsti); siftdown(lkv.k, lkv.v, );
}
}
private void reheap(int i) {
;
if (lfti < sz) {
; reheap(lfti); reheap(rghti);
var ckv = pq[i]; siftdown(ckv.k, ckv.v, i);
}
}
private void bld(IEnumerable<Tuple<KeyT, V>> sq) {
var sqm = from e in sq
select new HeapEntry(e.Item1, e.Item2);
pq = sqm.ToList<HeapEntry>();
var sz = pq.Count;
) {
];
pq.Add(new HeapEntry(KeyT.MaxValue, lkv.v));
reheap();
}
}
private IEnumerable<Tuple<KeyT, V>> sq() {
return from e in pq
where e.k != KeyT.MaxValue
select new Tuple<KeyT, V>(e.k, e.v); }
private void adj(Func<KeyT, V, Tuple<KeyT, V>> f) {
;
; i < cnt; ++i) {
var e = pq[i];
var r = f(e.k, e.v);
pq[i] = new HeapEntry(r.Item1, r.Item2);
}
reheap();
}
public static MinHeapPQ<V> empty { get { return new MinHeapPQ<V>(); } }
public static bool isEmpty(MinHeapPQ<V> pq) { return pq.mt; }
public static int size(MinHeapPQ<V> pq) { return pq.sz; }
public static Tuple<KeyT, V> peekMin(MinHeapPQ<V> pq) { return pq.pkmn; }
public static MinHeapPQ<V> push(KeyT k, V v, MinHeapPQ<V> pq) {
pq.psh(k, v); return pq; }
public static MinHeapPQ<V> replaceMin(KeyT k, V v, MinHeapPQ<V> pq) {
pq.rplcmin(k, v); return pq; }
public static MinHeapPQ<V> deleteMin(MinHeapPQ<V> pq) { pq.dltmin(); return pq; }
public static MinHeapPQ<V> merge(MinHeapPQ<V> pq1, MinHeapPQ<V> pq2) {
return fromSeq(pq1.sq().Concat(pq2.sq())); }
public static MinHeapPQ<V> adjust(Func<KeyT, V, Tuple<KeyT, V>> f, MinHeapPQ<V> pq) {
pq.adj(f); return pq; }
public static MinHeapPQ<V> fromSeq(IEnumerable<Tuple<KeyT, V>> sq) {
var pq = new MinHeapPQ<V>(); pq.bld(sq); return pq; }
public static Tuple<Tuple<KeyT, V>, MinHeapPQ<V>> popMin(MinHeapPQ<V> pq) {
var rslt = pq.pkmn; if (rslt == null) return null;
pq.dltmin(); return new Tuple<Tuple<KeyT, V>, MinHeapPQ<V>>(rslt, pq); }
public static IEnumerable<Tuple<KeyT, V>> toSeq(MinHeapPQ<V> pq) {
for (; !pq.mt; pq.dltmin()) yield return pq.pkmn; }
public static IEnumerable<Tuple<KeyT, V>> sort(IEnumerable<Tuple<KeyT, V>> sq) {
return toSeq(fromSeq(sq)); }
}
}
The above class code offers a full set of static methods and properties:
1. "empty" to create a new empty queue, 2. "isEmpty" to test if a queue is empty, 3. "size" to get the number of elements in the queue, 4. "peekMin" to retrieve the lowest priority key/value pair entry as a Tuple (possibly null for empty queues), 5. "push" to insert an entry, 6. "deleteMin" to remove the lowest priority entry, 7. "replaceMin" to replace the lowest priority and adjust the queue according to the value (faster than a "deleteMin" followed by a "push"), 8. "adjust" to apply a function to every key/value entry pair and reheapify the result, 9. "merge" to merge two queues into a single reheapified result, 10. "fromSeq" to build a queue from a sequence of key/value pair tuples, 11. "popMin" which is a convenience function combining a "peekMin" with a "deleteMin", returning null if the queue is empty and a tuple of the result otherwise, 12. "toSeq" to output an ordered sequence of the queue contents as Tuple's of the key/value pairs, and 13. "sort" which is a convenience function combining "fromSeq" followed by "toSeq".
The first four are all O(1) and the remainder O(log n) except "adjust" and "fromSeq" are O(n), "merge" is O(m + n) where m and n are the sizes of the two queues, and "toSeq" and "sort" are O(n log n); "replaceMin" is still O(log n) but faster than a "deleteMin" followed by a "push" by a constant factor.
Note that the Key type "KeyT" is not generic in order to give better comparison efficiency than using generic comparison using the IComparible interface but can be changed to different numeric types using the "using KeyT = ???" type alias.
The above code can be tested as per the page specification by the following code:
static void Main(string[] args) {
Tuple<uint, string>[] ins = { new Tuple<uint,string>(3u, "Clear drains"),
new Tuple<uint,string>(4u, "Feed cat"),
new Tuple<uint,string>(5u, "Make tea"),
new Tuple<uint,string>(1u, "Solve RC tasks"),
new Tuple<uint,string>(2u, "Tax return") };
var spq = ins.Aggregate(MinHeapPQ<string>.empty, (pq, t) => MinHeapPQ<string>.push(t.Item1, t.Item2, pq));
foreach (var e in MinHeapPQ<string>.toSeq(spq)) Console.WriteLine(e); Console.WriteLine();
foreach (var e in MinHeapPQ<string>.sort(ins)) Console.WriteLine(e); Console.WriteLine();
var npq = MinHeapPQ<string>.fromSeq(ins);
foreach (var e in MinHeapPQ<string>.toSeq(MinHeapPQ<string>.merge(npq, npq)))
Console.WriteLine(e); Console.WriteLine();
var npq = MinHeapPQ<string>.fromSeq(ins);
foreach (var e in MinHeapPQ<string>.toSeq(MinHeapPQ<string>.merge(npq, npq)))
Console.WriteLine(e);
foreach (var e in MinHeapPQ<string>.toSeq(MinHeapPQ<string>.adjust((k, v) => new Tuple<uint,string>(6u - k, v), npq)))
Console.WriteLine(e); Console.WriteLine();
}
It tests building the queue the slow way using repeated "push"'s - O(n log n), the faster "fromSeq" (included in the "sort") - O(n), and also tests the "merge" and "adjust" methods.
The output of the above test is as follows:
- Output:
(1, Solve RC tasks) (2, Tax return) (3, Clear drains) (4, Feed cat) (5, Make tea) (1, Solve RC tasks) (2, Tax return) (3, Clear drains) (4, Feed cat) (5, Make tea) (1, Solve RC tasks) (1, Solve RC tasks) (2, Tax return) (2, Tax return) (3, Clear drains) (3, Clear drains) (4, Feed cat) (4, Feed cat) (5, Make tea) (5, Make tea) (1, Make tea) (2, Feed cat) (3, Clear drains) (4, Tax return) (5, Solve RC tasks) 再贴上自己写的一点扩展方法
public static class MinHeapPQEX
{
/// <summary>
/// 创建一个空的优先级队列O(1)
/// </summary>
/// <typeparam name="T"></typeparam>
/// <returns></returns>
public static PriorityQueue<T> CreatPriorityQueue<T>()
{
return PriorityQueue<T>.empty;
}
/// <summary>
/// 进队 O(log n)
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="pq"></param>
/// <param name="priority"></param>
/// <param name="model"></param>
public static void Enqueue<T>(this PriorityQueue<T> pq, UInt32 priority, T model)
{
PriorityQueue<T>.push(priority, model, pq);
}
/// <summary>
/// 出队 peek+delete
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="pq"></param>
/// <returns></returns>
public static T Dequeue<T>(this PriorityQueue<T> pq)
{
return PriorityQueue<T>.popMin(pq).Item1.Item2;
}
/// <summary>
/// 检索,但不出队 O(1)
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="pq"></param>
/// <returns></returns>
public static T Peek<T>(this PriorityQueue<T> pq)
{
return PriorityQueue<T>.peekMin(pq).Item2;
}
/// <summary>
/// 判断队列是否为空 O(1)
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="pq"></param>
/// <returns></returns>
public static bool IsEmpty<T>(this PriorityQueue<T> pq)
{
return PriorityQueue<T>.isEmpty(pq);
}
/// <summary>
/// 统计 O(1)
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="pq"></param>
/// <returns></returns>
public static int Count<T>(this PriorityQueue<T> pq)
{
return PriorityQueue<T>.size(pq);
}
/// <summary>
/// 删除即将出队的元素
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="pq"></param>
public static void Delete<T>(this PriorityQueue<T> pq)
{
PriorityQueue<T>.deleteMin(pq);
}
}
C# 优先级队列的更多相关文章
- 体验Rabbitmq强大的【优先级队列】之轻松面对现实业务场景
说到队列的话,大家一定不会陌生,但是扯到优先级队列的话,还是有一部分同学是不清楚的,可能是不知道怎么去实现吧,其实呢,,,这东西已 经烂大街了...很简单,用“堆”去实现的,在我们系统中有一个订单催付 ...
- Java中的队列Queue,优先级队列PriorityQueue
队列Queue 在java5中新增加了java.util.Queue接口,用以支持队列的常见操作.该接口扩展了java.util.Collection接口. Queue使用时要尽量避免Collecti ...
- 如何基于RabbitMQ实现优先级队列
概述 由于种种原因,RabbitMQ到目前为止,官方还没有实现优先级队列,只实现了Consumer的优先级处理. 但是,迫于种种原因,应用层面上又需要优先级队列,因此需求来了:如何为RabbitMQ加 ...
- ACM/ICPC 之 优先级队列+设置IO缓存区(TSH OJ-Schedule(任务调度))
一个裸的优先级队列(最大堆)题,但也有其他普通队列的做法.这道题我做了两天,结果发现是输入输出太过频繁,一直只能A掉55%的数据,其他都是TLE,如果将输入输出的数据放入缓存区,然后满区输出,可以将I ...
- java中PriorityQueue优先级队列使用方法
优先级队列是不同于先进先出队列的另一种队列.每次从队列中取出的是具有最高优先权的元素. PriorityQueue是从JDK1.5开始提供的新的数据结构接口. 如果不提供Comparator的话,优先 ...
- stl的优先级队列
#include <iostream> #include <vector> #include <queue> using namespace std; class ...
- 【python cookbook】【数据结构与算法】5.实现优先级队列
问题:要实现一个队列,它能够以给定的优先级对元素排序,且每次pop操作时都会返回优先级最高的那个元素: 解决方案:采用heapq模块实现一个简单的优先级队列 # example.py # # Exam ...
- POJ 2227 The Wedding Juicer (优先级队列+bfs+dfs)
思路描述来自:http://hi.baidu.com/perfectcai_/item/701f2efa460cedcb0dd1c820也可以参考黑书P89的积水. 题意:Farmer John有一个 ...
- 《Java数据结构与算法》笔记-CH4-6优先级队列
/** * 优先级队列 * 效率:插入O(n),删除O(1).第12章介绍如何通过堆来改进insert时间 */ class PriorityQueue { private int maxSize; ...
- STL学习系列七:优先级队列priority_queue容器
1.简介 最大值优先级队列.最小值优先级队列 优先级队列适配器 STL priority_queue 用来开发一些特殊的应用,请对stl的类库,多做扩展性学习 这里给个例子: #include< ...
随机推荐
- HTML Email 编写指南(转)
作者: 阮一峰 日期: 2013年6月16日 今天,我想写一个"低技术"问题. 话说我订阅了不少了新闻邮件(Newsletter),比如JavaScript Weekly.每周 ...
- jquery file upload示例
原文链接:http://blog.csdn.net/qq_37936542/article/details/79258158 jquery file upload是一款实用的上传文件插件,项目中刚好用 ...
- 【b094&&z14】靶形数独
[问题描述] 小城和小华都是热爱数学的好学生,最近,他们不约而同地迷上了数独游戏,好胜的分们想用数独来一比高低.但普通的数独对他们来说都过于简单了,于是他们向Z博士请教,Z博士拿出了他最近发明的&qu ...
- hdu 4865 dp
Peter's Hobby Time Limit: 2000/1000 MS (Java/Others) Memory Limit: 32768/32768 K (Java/Others) To ...
- Xcode崩溃定位:异常位置Exception的断点
1.全局Exception断点 在XCode界面中按cmd + 6跳到Breakpoint的tab(或者点击下图左上第7个图标),然后点击左下角的+号,增加一个Exception的断点,如下图所示.这 ...
- 一题多解(五) —— topK(数组中第 k 大/小的数)
根据对称性,第 k 大和第 k 小,在实现上,是一致的,我们就以第 k 小为例,进行说明: 法 1 直接排序(sort(A, A+N)),当使用一般时间复杂度的排序算法时,其时间复杂度为 O(N2) ...
- React事件处理函数传参问题
React事件处理函数参数 HTML标签与React 组件是不同的,事件对象e是HTML标签元素的,组件没有的.
- C#中的yield关键字
迭代器,是一个连续的集合,出现多个yield return其实就是将这多个的yield return元素按照出现的顺序存储在迭代器的集合中而已.形如下面的形式: public class CityCo ...
- 十个最有“钱景”的IT技能, 你掌握了哪个?
IT行业的失业率仍然徘徊在历史低点,其中某些岗位(如网络和安全工程师和软件开发商)的失业率在1%左右. Robert Half Technology最近的一项调查显示,大多数CIO将扩大IT团队或专注 ...
- Android菜鸟的成长笔记(18)——绑定本地Service并与之通信
在上一篇中介绍了Service与Activity的区别及Service两种启动方式中的第一种启动方式startService(). 我们会发现用startService().stopService() ...