Logo Data Structures and Algorithms with Object-Oriented Design Patterns in C++
next up previous contents index

Basics

A priority queue is a container which provides the following three operations:

Enqueue
used to put objects into the container;
FindMin
returns a reference to the smallest object in the container; and
DequeueMin
removes the smallest object from the container.

A priority queue is used to store a finite set of keys drawn from a totally ordered set of keys K. As distinct from search trees, duplicate keys are allowed in priority queues.

Program gif gives the declaration of the PriorityQueue abstract class. The PriorityQueue class is derived from the Container class. In addition to the inherited functions, the public interface of the PriorityQueue class comprises the three functions listed above.

   program23974
Program: PriorityQueue and MergeablePriorityQueue Class Definitions

Program gif also declares one additional class--MergeablePriorityQueue. A mergeable priority queue   is one which provides the ability to merge efficiently two priority queues into one. Of course it is always possible to merge two priority queues by dequeuing the elements of one queue and enqueuing them in the other. However, the mergeable priority queue implementations we will consider allow more efficient merging than this.

   figure23996
Figure: Object Class Hierarchy

It is possible to implement the required functionality using data structures that we have already considered. For example, a priority queue can be implemented simply as a list. If an unsorted list  is used, enqueuing can be accomplished in constant time. However, finding the minimum and removing the minimum each require O(n) time where n is the number of items in the queue. On the other hand, if an sorted list  is used, finding the minimum and removing it is easy--both operations can be done in constant time. However, enqueuing an item in an sorted list requires O(n) time.

Another possibility is to use a search tree. For example, if an AVL tree  is used to implement a priority queue, then all three operations can be done in tex2html_wrap_inline59891 time. However, search trees provide more functionality than we need. Viz., search trees support finding the largest item with FindMax, deletion of arbitrary objects with Withdraw, and the ability to visit in order all the contained objects via DepthFirstTraversal. All these operations can be done as efficiently as the priority queue operations. Because search trees support more functions than we really need for priority queues, it is reasonable to suspect that there are more efficient ways to implement priority queues. And indeed there are!

A number of different priority queue implementations are described in this chapter. All the implementations have one thing in common--they are all based on a special kind of tree called a min heap or simply a heap.

Definition ((Min) Heap)  A (Min) Heap   is a tree,

displaymath66128

with the following properties:

  1. Every subtree of T is a heap; and,
  2. The root of T is less than or equal to the root of every subtree of T. I.e., tex2html_wrap_inline66146, where tex2html_wrap_inline66148 is the root of tex2html_wrap_inline63460.
According to Definition gif, the key in each node of a heap is less than or equal to the roots of all the subtrees of that node. Therefore, by induction, the key in each node is less than or equal to all the keys contained in the subtrees of that node. Note, however, that the definition says nothing about the relative ordering of the keys in the subtrees of a given node. For example, in a binary heap either the left or the right subtree of a given node may have the larger key.


next up previous contents index

Bruno Copyright © 1997 by Bruno R. Preiss, P.Eng. All rights reserved.