Big Java

1
Big Java

• Chapter 16

2
Advanced Data Structures

• Often referred to as the Java Collections
  Framework.
• Set and map data types
• Hash tables
• Binary trees
• Heap
• Priority queue

3
Sets

• Unordered collection
• Fundamental operations
• Add an element
• Remove an element
• Containment test (is object in set?)
• List elements (in arbitrary order)
• Reject duplicates

4
Set Implementation
ltltinterfacegtgt Set
HashSet
TreeSet
5
Example
instantiate with specific implementation. Advantag
e could easily change implementation, rest of
program stays the same.
use methods from interface
could also do if (names.contains(Jane))

6
Example (continued)
can use for each loop with Sets Nodes are
not necessarily visited in order inserted!
7
Reading Assignment

• Read Quality Tip 16.1 pages 704-705. How would
  you feel if you had designed the List interface
  and ArrayList/LinkedList classes?

8
Maps

• A map data type keeps associations between keys
  and values
• Cannot contain duplicate keys
• Operations include
• put
• get
• containsKey/containsValue
• keySet/values return all keys/values
• size
• Java has two implementations HashMap and
  TreeMap.

9
Example
10
Hash Table

• Hash table can be used to implement sets and
  maps
• Hash function computes an integer value (hash
  code) from an object, goal is for different
  objects to yield different hash codes
• int h obj.hashCode()
• hashCode is inherited from Object
• Collision when two or more distinct objects
  have the same hash code

11
HashSet

• Author provides HashSet, an extension of
  AbstractSet
• provides a hash table with buckets (linked lists)
  to hold collisions
• possible hash code for string
• final int HASH_MULTIPLER 31 //prime
• int h0
• for (int i0 ilts.length() i)
• h HASH_MULTIPLIER h s.charAt(i)
• Can use integer fields directly as hash codes
• If hashCode not overridden, Object class computes
  based on memory location of object. Problem if
  define equals but not hashCode.

12
Binary Search Tree

• Review on your own, pages 720-734

13
TreeSet or HashSet?

• With a good hash function, hashing is usually
  faster
• Balanced trees (remember those?) can guarantee
  reasonable performance, HashSet depends entirely
  on performance of hash function
• To use TreeSet, objects being stored must
  implement Comparable interface
• For TreeMap, same requirement for keys
• Can supply a Comparator object to TreeSet/TreeMap
  constructor (takes two objects and returns
  comparison)
• TreeSet is preferable if want to print list of
  items in order

14
Reading Assignment

• Random Fact 16.2, Software Piracy, pages 738-739

15
Priority Queue

• Collects elements which have a priority
• Elements inserted in any order, but retrieved
  according to priority
• Java includes PriorityQueue() class
• Abstract data class
• Often use heap to implement
• Heaps covered in 406 (I think)
• Heapsort is O(n log(n))

16
Chapter 16 Quick Exercise

• Use the HashTable code from the textbook
• Can you store objects with the same hash code but
  different data? Try modifying CoinHashCodePrinter
  to insert two quarters but initialize them
  differently (e.g., a quarter and quarter
  rather than both quarter)
• Review the TreeSet and TreeSetTester code.
  Notice the use of CoinComparator. Modify to use
  with some other class, such as BankAccount