Heap

• A Heap data structure is a binary tree with the following properties :

1. It is a complete binary tree; that is, each level of the tree is completely filled, except possibly the bottom level.
At this level, it is filled from left to right.

2. It satisfies the heap-order property: The data item stored in each node is greater than or equal to
the data items stored in its children.

• Examples:
(a) (b) (c)
9 9 9

8 4 8 4 6 4

6 2 3 6 2 3 8 2 3

• In the above examples only (a) is a heap. (b) is not a heap as it is not complete and (c) is complete
but does not satisfy the second property defined for heaps.

2
Heap as an Array
• An array A that represents a heap is an array with two attributes
– length, the number of elements in the array
– heap-size, the number of heap elements stored in the array
• Viewed as a binary tree and as an array :

116
2 3
14 5 6
10
4 7
8 7 9 3
8 9 10
2 4 1

1 2 3 4 5 6 7 8 9 10
16 14 10 8 7 9 3 2 4 1

•The root of the tree is stored at A[0], its left-child at A[1], its right child at A[2]
etc.
4
 Accessing the Heap Values
• Given the index i of a node, the indices of its parent Parent(i), left-child
LeftChild(i) and right child RightChild(i) can be computed simply :

Parent(i) LeftChild(i) RightChild(i)

return i/2 return 2i return 2i+1

• Heaps must also satisfy the heap property for every node, i, other than the
root.
A[Parent(i)] A[i]

• Therefore, the largest element in a heap is stored at the root, and the
subtrees rooted at a node contain smaller values than does the node itself.

5
Inserting into a Heap
Algorithm
Bharti Nathani [email protected]
 Heap as an Array
• We could implement heaps using a linked list like structure, like we did with
binary trees but in this instance it is actually easier to implement heaps using
arrays.

• We simply number the nodes in the heap from top to bottom, numbering the nodes
on each level from left to right and store the ith node in the ith location of the
array. (of course remembering that array indexing in Java starts at zero).
1

2 3

4 5 6 7

Programing and Data Structures 9

 Heap Operations
• The height of a node in a tree is the number of
edges on the longest simple downward path from
the node to a leaf. (i.e. maximum depth from that
node – Remember?)

• The height of an n-element heap based on a

binary tree is lg n

• The basic operations on heaps run in time at

most proportional to the height of the tree and
thus take O(lg n) time.
1
0
 Maintaining the Heap Property
• One of the more basic heap operations is converting a complete binary
tree to a heap.

• Such an operation is called “Heapify”.

• Its inputs are an array A and an index i into the array.

• When Heapify is called, it is assumed that the binary trees rooted at

LeftChild(i) and RightChild(i) are heaps, but that A[i] may be smaller
than its children, thus violating the 2nd heap property.

• The function of Heapify is to let the value at A[i] “float down” in the
heap so that the subtree rooted at index i becomes a heap.

• The action required from Heapify is as follows:

1
1
 Example of Heapify 1
16 • Subtree to be heapified is marked by the shaded
2 3
i 4 5 10 7
triangle.
4 6
14 7 9 3 • Node i if smaller than its children is swapped with
8 9 10 1 the greater of its two children – in this case the left
2 8 child with value 14 is swapped for value 4 in node I.

1
16 3 • If the next subtree down (again marked with the
2
14 10 shaded triangle region) is not a heap, we “percolate”
5 6 7 down to the next subtree level and find the larger
i 4
4 7 9 3 of the two children again to swap with current root
8 9 10 node.
2 8 1

1
2 16 • We keep “percolating” down until either the
3
14 10 subtree conforms to the heap properties or a leaf
4 5 6 7 has been reached.
8 7 9 3
8 9 10 1
2 4
Programming and Data Structures 7
 The Heapify Algorithm
Begin
Heapify(A[], i) // Heapify takes in our heap and index of current root node of subtree to be heapified
{
left = LeftChild(i); // index of left child
right = RightChild(i); // index of right child

if left A.heap-size AND A[left] > { // if still within bounds AND left child
A[i] largest = left; // greather than parent – remember largest as left child
}
else {
largest = i; // else parent still has largest value for now
}

if right A.heap-size AND A[right] > A[largest] { // if still within bounds AND right child
largest = right // greather than parent – remember largest as right child
}
if largest NOT EQUAL i { // if parent does not hold largest value
swap A[i] and A[largest] // swap parent with child that does have largest value
Heapify(A, largest) // Percolate down and Heapify next subtree
}
}
End
Programming and Data Structures 13
 The Heapify Algorithm
• At each step, the index of the largest of the elements A[i], A[Left(i)],
and A[Right(i)] is stored in the variable largest.

• If A[i] is largest, then the subtree rooted at node i is a heap and the
procedure ends.

• Otherwise, one of the two children has the largest element, and A[i] is
swapped with A[largest], which causes node i and its children to satisfy
the heap property.

• The node largest, however, now has the original value A[i], and thus
the subtree rooted at largest may violate the heap property.

• Therefore, Heapify must be called recursively on that subtree.

Programming and Data Structures 14

 Analysis of Heapify
• The running time of Heapify on a subtree of size n rooted at a given
node i is the O(1) time to fix up the relationships among the elements
A[i], A[Left(i)] and A[Right(i)], plus the time to run Heapify on a
subtree rooted at one of the children of node i.

• The subtrees can have size of at most 2n/3, and the running time of
Heapify can therefore be described by the recurrence,
T(n) T(2n/3) + O(1)

• Using either the iteration method or the master’s theorem we can find
that the solution to this works out as:
T(n) = O(log n)

Programming and Data Structures 15

 Building a Heap
• For the general case of converting a complete binary tree to a heap, we begin at the last
node that is not a leaf, apply the “percolate down” routine to convert the subtree rooted at
this current root node to a heap.

• We then move onto the preceding node and “percolate down” that subtree.

• We continue on in this manner, working up the tree until we reach the root of the given tree.

• We use the Heapify procedure here in a bottom up manner. This means also means that to
convert an array A[1..n], where n = A.length, into a heap we can apply the same procedure
as described above.

• Now, since the subarray A[( n/2 + 1)..n] are all leaves of the tree, each is a 1-element heap.

• The procedure “Build-Heap” goes through the remaining nodes and runs Heapify on each.

• The order of processing guarantees that the subtrees rooted at children of node i are heaps
before Heapify is run at that node.

Programming and Data Structures 16

 Build Heap Algorithm
Begin:
Build-Heap (A[]) { // takes in an array to be heapified
A.heap-size = A.length; // heap size is number of elements in the array
for i = A.length/2 downto 1 // starting at first subtree up to root full tree
{
Heapify(A, i); // Heapify the current subtree
}
}
End

• Each call to Heapify costs O(lg n) time, and there are O(n) such calls.
• Thus, the running time is at most O(n lg n)
• A more complex analysis, however, gives us a tighter upper bound of O(n).
• Hence, we can build a heap from an unordered array in linear time.

• Lets have a look at a visual example of this operation (have a look at the next slide).
• The current subtrees that are being worked on are marked inside a shaded triangle.
• Any swapping that goes on is marked with a thicker edge.
Programming and Data Structures 17
 Example of Build Heap
A 4 1 3 2 16 9 10 14 8 7

14 1
4
2 3 2 3
1 3 1 5 3 7
4 5 6 7 4 6
2 i 16 9 10 i 2 16 9 10
8 14 9 10 9 10
8 7 8 14 8 7
(a) (b)
14 1
2 4
2 1
5 i 33 7 i 1 10 3
6 5 6 7
4 14 16 9 10 4 16 9 3
14
82 9 10 7 8 9 10 7
8 2 8
(c) (d)
1 1
i 4 16
3 2 3
216 10 14 10
5 6 7 5 6 7
4 7 9 3 4 8 7 9 3
14 10
8 2 9 10 8 9 1
8 1 2 4
(f)
(e)
18