Should I use arrays or linked lists in my implementations? In general we don't tell you how to implement your data structures—you can use arrays, linked lists, or maybe even invent your own new structure provided you abide by the specified time and space requirements.
How serious are you about not calling external library functions (other than those that are specifically permitted)? You will receive a substantial deduction. The goal of this assignment is to implement data types from first principles, using resizing arrays and linked lists—feel free to use java.util.LinkedList and java.util.ArrayList on future programming assignments. We also require you to use StdIn (instead of java.util.Scanner) because we will intercept the calls to StdIn in our testing.
Can I add extra public methods to the Deque or RandomizedQueue APIs? Can I use different names for the methods? No, you must implement the API exactly as specified.
What does"uniformly at random" mean? If there are N items in the randomized queue, then you should choose each one with probability 1/N, up to the randomness of StdRandom.uniform(), independent of past decisions. You can generate a pseudo-random integer between 0 and N − 1 using StdRandom.uniform(N).
Given an array, how can I rearrange the entries in random order? Use StdRandom.shuffle()—it implements the Knuth shuffle discussed in lecture and runs in linear time. Note that depending on your implementation, you may not need to call this method.
Can repeated calls to sample() in a randomized queue return the same item more than once? Yes, since you are sampling without removing the item. This is also known as "sampling with replacement."
Should two iterators to the same randomized queue return the items in the same order? No, each iterator should have a different random order. This is what "independent iterators" means.
Why is it called a randomized queue if the items are not removed in FIFO? This is a common name used in queueing theory.
What should the next() method return if the iterator has no more items? According to the API for java.util.Iterator, it should throw a java.util.NoSuchElementException.
What should my deque (or randomized queue) iterator do if the deque (or randomized queue) is structurally modified at any time after the iterator is created (but before it is done iterating)? You don't need to worry about this in your solution. An industrial-strength solution (used in the Java libraries) is to make the iterator fail-fast: throw a java.util.ConcurrentModificationException as soon as this is detected.
Why does the following code lead to a generic array creation compile-time error when Item is a generic type parameter?
Java prohibits the creation of arrays of generic types. See the Q+A in Section 1.3 for a brief discussion. Instead, use a cast.Item a = new Item;
Unfortunately, this leads to an unavoidable compiler warning.Item a = (Item) new Object;
The compiler says that my program uses unchecked or unsafe operations and to recompile with -Xlint:unchecked for details. Usually this means you did a potentially unsafe cast. When implementing a generic stack with an array, this is unavoidable since Java does not allow generic array creation. For example, the compiler outputs the following warning with ResizingArrayStack.java:
% javac-algs4 ResizingArrayStack.java Note: ResizingArrayStack.java uses unchecked or unsafe operations. Note: Recompile with -Xlint:unchecked for details. % javac-algs4 -Xlint:unchecked ResizingArrayStack.java ResizingArrayStack.java:25: warning: [unchecked] unchecked cast found : java.lang.Object required: Item a = (Item) new Object; ^ ResizingArrayStack.java:36: warning: [unchecked] unchecked cast found : java.lang.Object required: Item Item temp = (Item) new Object[capacity]; ^ 2 warnings
You should not receive any other compiler warnings on this assignment (and you should not receive any compiler warnings on any other assignment).
Checkstyle complains that my nested class' instance variables must be private and have accessor methods that are not private. Do I need to make them private? No, but there's no harm in doing so. The access modifier of a nested class' instance variable is irrelevant—regardless of its access modifier, it can be accessed anywhere in the file. (Of course, the enclosing class' instance variables should be private.)
Can a nested class have a constructor? Yes.
What assumptions can I make about the input to Subset? Standard input can contain any sequence of strings. You may assume that there is one integer command-line argument k and it is between 0 and the number of strings on standard input.
Will I lose points for loitering? Yes. See p. 137 of the textbook for a discussion of loitering. Loitering is maintaining a useless reference to an object that could otherwise be garbage collected.
Develop unit tests as you write each method and constructor to allow for testing. As an example for Deque, you know that if you call addFirst() with the numbers 1 through N in ascending order, then call removeLast() N times, you should see the numbers 1 through N in ascending order. As soon as you have those two methods written, you can write a unit test for these methods. In addition use randomized unit tests (which we employ heavily in our correctness testing).
Test intermixed sequence of operations. Sometimes you want to test two methods in isolation, as above. But, you also need to make sure that all of the methods work together with one another.
Test your iterator. And make sure to test that multiple iterators can be used simultaneously. You can test this with a nested foreach loop. The iterators should operate independently of one another.
Make sure to test what happens when your data structures are emptied. One very common bug is for something to go wrong when your data structure goes from non-empty to empty and then back to non-empty. Make sure to include this in your tests.
Don't rely on our automated tests for debugging. As suggested above, write your own unit tests; it's good practice.
These are purely suggestions for how you might make progress. You do not have to follow these steps. These same steps apply to each of the two data types that you will be implementing.