Java Collections Framework - Building Java Programs: A Back to Basics Approach - page 694

You should not try to construct a TreeSet of objects without a natural ordering,
such as Point objects:

// this code compiles but will lead to a runtime error
Set<Point> points = new TreeSet<Point>();

The preceding code compiles (unfortunately), but it generates an exception when
you run it because it doesn't know how to order the Point objects in the TreeSet :

Exception in thread "main"
java.lang.ClassCastException: java.awt.Point
at java.util.TreeMap.compare(Unknown Source)
at java.util.TreeMap.put(Unknown Source)
at java.util.TreeSet.add(Unknown Source)

You'd be better off using a HashSet in this case.
In summary, the following are some of the major differences between HashSet
and TreeSet :

Collection Strengths

• Extremely fast performance for add, contains, and remove tasks
• Can be used with any type of objects as its elements

HashSet

• Elements are stored in sorted order
• Must be used with elements that can be compared (such as Integer , String )

TreeSet

Set Operations
Consider the task of figuring out how many unique elements appear in two given
sets. You cannot just add the sets' sizes, since they might have some elements in
common that should not be counted twice in your total. Instead, you could count all
elements from the first set and then count only the unique elements of the second, by
checking to see whether each element from the second is also in the first:

// Returns the number of unique elements contained
// in either set1 or set2. Not a good model to follow.
public static int totalElements(Set<String> set1, Set<String> set2) {
int count = set1.size();
for (String element : set2) {
if (!set2.contains(element)) {
count++;

}

}
return count;

}