CMPS 144   Fall 2023
Final Exam Preview

You can expect that topics that did not appear in the mid-term exams will be emphasized more than those that did.

• Knight Distances and Split-Coalesce Sorting: The vast majority of students submitted Java classes that were correct, or had only minor errors, for Prog. Assgs. #3 and #4. Be able to devise Java code that carries out a task that is similar to one from either of those assignments.
• Parallel Computing and Threads:
  • Parallel Algorithms: With respect to the relevant web page, be able to answer questions about the algorithms shown there (summing the elements of an array, RankSort), or to give a high-level sketch of a parallel algorithm that works in a manner similar to an example shown there.
• Hashing and Hash Tables: (See Lab #14) Given a hash table (where open addressing is used with the linear probing collision resolution strategy),
  • compute the expected number of probes in both successful and unsuccessful searches (for that specific table), and be able to compare the latter number with how many probes could be expected to be made in an unsuccessful search, based only upon the table's packing density.
  • show at which address a new <key,value>-pair would be inserted (given its key's home address).
  • show how any <key,value>-pairs would move within the table as a consequence of removing one of them.
• Binary Trees:
  • Huffman Trees/Coding: Given a list of symbols and their respective frequencies-of-occurrence, build a corresponding Huffman Tree and show the symbol-to-codeword mapping that the tree induces. (See Lab #14.)
  • Heaps: Interpreting an array as representing a complete binary tree, show steps involved in "heapifying" it, so as to turn it into a min- or max-heap, or in preserving its heap-ness in the face of removing the item from its root. Be able to trace execution of the HeapSort algorithm. See the relevant web page.
  • Binary Search Tree: Given one, show how it would change as a result of inserting or deleting specified keys. (See Lab #13.)

• Recursion: (See Labs #11 and #12.) Given a relatively simple recursive method, be able to trace its execution (e.g., by showing the tree depicting all instances of the method that get created) and to describe its effect (e.g., what does it print, or what value does it return) when called with specified parameter values.

  Given a relatively simple problem —involving numbers or an array, say— that is amenable to a recursive solution, be able to describe such a solution. (See Lab #10.)

• Lists:
  • Positional Paradigm: Given a relatively simple problem about lists, develop a Java method that uses the methods in the PositionalListWithCursors and PositionalListCursor interfaces to solve it. (See Lab #10.)
• Class Hierarchies/Extension
• To provide extra functionality: Given a relatively simple Java class that lacks some desired functionality, develop a child class that provides that functionality. (See Lab #5: the Counter class hierarchy.)
• To provide alternative behaviors of some key method(s): Given a relatively simple Java class having an abstract method, develop concrete child classes that implement that method in ways that result in different behaviors. (See the Counter Class Hierarchy (Lab #5) and Lab #4.)

  A similar effect (of producing objects of the same class having different behaviors) can be achieved by using an instance variable that provides some particular functionality (e.g., comparison, classification) as described in a Java interface (e.g., java.util.Comparator). (See Generic Comparator-based Sorting in Java and Lab #11 (Red/Blue classification).) Indeed, this is usually a better approach than using extension.

• Stacks, Queues: applications and implementations (array-based and Link1-based). See Labs #6, 7, and 9.
• Loop Invariants: Given some task to carry out, and a loop invariant to guide you, develop code to solve the problem in a manner consistent with the given loop invariant. (Remember the 2- and 3-color partitioning problems?) (See Lab #3.)