二叉堆,是一个满二叉树,满足堆的性质。即父节点大于等于子节点(max heap)或者是父节点小于等于子节点(min heap)。二叉堆的如上性质常用于优先队列(priority queue)或是用于堆排序。
由于max heap 与min heap类似,下文只针对min heap进行讨论和实现。
如上图,是根据字母的ASCII码建立的最小堆。
我们用数组对满二叉树采用宽度优先遍历存储堆结构,如下图所示:
从数组下标1开始存储堆,这样的处理方式可以得到如下性质:
1.堆中的每个父节点k,他的两个子节点为k*2和k*2+1
2.堆中的每个子节点k,他的父节点为k/2
堆的插入:当向堆中插入一个元素时,首先将新添加的元素放入数组的末尾,然后使用percolatingUp方法,将新添加的元素逐层向上移动至合适位置。如下图所示:
堆的删除:当将堆顶元素删除时,首先将数组中末尾的元素放入对顶,然后使用percolatingDown方法,将堆顶元素逐层向下移动至合适位置。如下图所示:
堆的构建:将对大小为k的数组,从第k/2个元素开始(第1个到第k/2个元素有子节点),依次使用porcelatingUp将元素向上调整,其时间复杂度O(k)。
堆排序的思想是:将待排序的数组转成堆(堆的构建),删除root节点(获取最小元素),重构(porcelatingDown),删除root节点(获取第二小元素),重构(porcelatingDown),...一直到堆中不再有元素。
堆操作的时间复杂度(N表示元素个数):
建堆:O(N)
添加:O(logN)
删除:O(logN)
堆排序:O(NlogN)
堆及堆排序的java实现:
/**************************************************************************** * This demonstrates binary heap operations along with the heapSort. * *****************************************************************************/ import java.util.*; @SuppressWarnings("unchecked") public class Heap<AnyType extends Comparable<AnyType>> { private static final int CAPACITY = 2; private int size; // Number of elements in heap private AnyType[] heap; // The heap array public Heap() { size = 0; heap = (AnyType[]) new Comparable[CAPACITY]; } /** * Construct the binary heap given an array of items. */ public Heap(AnyType[] array) { size = array.length; heap = (AnyType[]) new Comparable[array.length+1]; System.arraycopy(array, 0, heap, 1, array.length);//we do not use 0 index buildHeap(); } /** * runs at O(size) */ private void buildHeap() { for (int k = size/2; k > 0; k--) { percolatingDown(k); } } private void percolatingDown(int k) { AnyType tmp = heap[k]; int child; for(; 2*k <= size; k = child) { child = 2*k; if(child != size && heap[child].compareTo(heap[child + 1]) > 0) child++; if(tmp.compareTo(heap[child]) > 0) heap[k] = heap[child]; else break; } heap[k] = tmp; } /** * Sorts a given array of items. */ public void heapSort(AnyType[] array) { size = array.length; heap = (AnyType[]) new Comparable[size+1]; System.arraycopy(array, 0, heap, 1, size); buildHeap(); for (int i = size; i > 0; i--) { AnyType tmp = heap[i]; //move top item to the end of the heap array heap[i] = heap[1]; heap[1] = tmp; size--; percolatingDown(1); } for(int k = 0; k < heap.length-1; k++) array[k] = heap[heap.length - 1 - k]; } /** * Deletes the top item */ public AnyType deleteMin() throws RuntimeException { if (size == 0) throw new RuntimeException(); AnyType min = heap[1]; heap[1] = heap[size--]; percolatingDown(1); return min; } /** * Inserts a new item */ public void insert(AnyType x) { if(size == heap.length - 1) doubleSize(); //Insert a new item to the end of the array int pos = ++size; //Percolate up for(; pos > 1 && x.compareTo(heap[pos/2]) < 0; pos = pos/2 ) heap[pos] = heap[pos/2]; heap[pos] = x; } private void doubleSize() { AnyType [] old = heap; heap = (AnyType []) new Comparable[heap.length * 2]; System.arraycopy(old, 1, heap, 1, size); } public String toString() { String out = ""; for(int k = 1; k <= size; k++) out += heap[k]+" "; return out; } public static void main(String[] args) { Heap<String> h = new Heap<String>(); h.insert("p"); h.insert("r"); h.insert("i"); h.insert("o"); System.out.println(h); h.deleteMin(); System.out.println(h); Heap<Integer> tmp = new Heap<Integer>(); Integer[] a = {4,7,7,7,5,0,2,3,5,1}; tmp.heapSort(a); System.out.println(Arrays.toString(a)); } }
1.http://www.cs.cmu.edu/~adamchik/15-121/lectures/Binary%20Heaps/heaps.html
2.http://en.wikipedia.org/wiki/Binary_heap
原文地址:http://blog.csdn.net/hooting/article/details/44116301