Reputation: 409
Without providing custom comparator the priority queue inserts elements in ascending order, however, after removing a particular element the order is changed.
PriorityQueue<Integer> pq = new PriorityQueue<>();
pq.add(10);
pq.add(1);
pq.add(2);
pq.add(2);
pq.remove(2);
for(int x: pq) {
System.out.println(x);
}
//outputs: 1 10 2, instead of expected: 1 2 10
Any ideas?
Thanks.
Upvotes: 5
Views: 1241
Reputation: 16628
This is because PriorityQueue is an implementation of min heap (in Java by default), this means the root element will be always lower than the other elements in the list.
To take the example from your question:
1 < 10
the element at position 1 is now 1 and 10 it shifted to index 21 < 2
there is no need to shift the element, and 2 is inserted at index 310 > 2
a shift is needed, 2 is inserted at 2 and 10 is shited to the index 4After all inserts the array looks like this, also check the video below:
array: [NULL, 1, 2, 2, 10]
Now when you remove 2, a shift is required for 10, and 10 is moved back to index 2, and the state of array is [NULL, 1, 10, 2]
and this is what you see as the output when you iterate. Also, check the following method that is implemented inside PriorityQueue and is used when you don't provide a Comparator
.
private void siftUpComparable(int k, E x) {
Comparable<? super E> key = (Comparable<? super E>) x;
while (k > 0) {
int parent = (k - 1) >>> 1;
Object e = queue[parent];
if (key.compareTo((E) e) >= 0)
break;
queue[k] = e;
k = parent;
}
queue[k] = key;
}
Upvotes: 0
Reputation: 11120
Don't iterate on your PriorityQueue<T>
as you do on a collections/arrays; use .poll()
, instead:
while(pq.peek()!=null) {
System.out.println(pq.poll());
}
Priority Queue is an Abstract Data Type, that is often implemented as a Binary Heap data structure, which, in turn, is (commonly) implemented with the array. There are some other ways to implement binary heap, but an ordinary array, is the fastest, simplest and best way for it.
An example of how the array represents a binary heap, looks like this:
Queue order is not being changed, in your case; rather, you're just utilizing data structure in a wrong way, by merely iterating on it in a traditional for-each
/iterative way, as when you iterate on a basic array, not considering, that your Priority Queue backing array is not sorted with its ith index; rather it maintains the top element on top of tree (either Min Heap or Max Heap case) and you can't just get the .poll()
effect by iterating on it in a traditional way.
Upvotes: 4