Scribd
Upload
51 views

7 Amortized Analysis

This document discusses amortized analysis and various data structures. It contains 4 problems: 1) Analyzing operations on different data structures and maintaining O(1) amortized costs using credits. Data structures include convex hull, counter, stack, and dynamic array. 2) Specifying a credit invariant for each data structure to ensure no operation violates it. 3) Proving that an algorithm using different operations will have O(∑ai*ci) worst case runtime, where ai is the number of times operation i is used and ci is its amortized cost. 4) Stating the complexity of an algorithm using a heap with specified amortized costs for its operations is Θ(m

Uploaded by

Anurag Parothia
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
51 views

7 Amortized Analysis

This document discusses amortized analysis and various data structures. It contains 4 problems: 1) Analyzing operations on different data structures and maintaining O(1) amortized costs using credits. Data structures include convex hull, counter, stack, and dynamic array. 2) Specifying a credit invariant for each data structure to ensure no operation violates it. 3) Proving that an algorithm using different operations will have O(∑ai*ci) worst case runtime, where ai is the number of times operation i is used and ci is its amortized cost. 4) Stating the complexity of an algorithm using a heap with specified amortized costs for its operations is Θ(m

Uploaded by

Anurag Parothia
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Amortized Analysis -

Accounting Method Homework Lecturer: Varun Rajan

1. In each of the following problems, a set of operations permissible on the underlying data
structure are given. For each of the operations, compute their best and worst case com-
plexities. Also, compute the actual cost (in terms of credits) per operation and manage
credits such that each operation has O (1) amortized cost.
(a) Convex Hull problem.
1. updateHull ( H, p) - Add point p to the set of points under consideration and up-
date the current hull H, accordingly.
(b) n-Bit Counter problem.
1. increment(χ) - Increment the counter χ by 1.
2. reset(χ) - Set χ to 0.
Hint: Maintain a pointer to the most significant bit of the counter.
(c) Stack operations.
1. Push(S, x ) - Push element x into the stack S.
2. Pop(S) - Pop the topmost element from S.
3. MultiPop(S, k ) - Pop the topmost k elements from S; pop all elements if k ≥ |S|.
Hint: Try to store the credits in such a way that Pop and MultiPop are "free of cost".
(d) Dynamic array problem.
1. addElement( A, x ) - Add the element x into the first free space in the (possibly
unsorted) array A. If A has no free space, create a new array of size | A|, copy all
the elements from A into the new array, free the space allocated to A and set A to
point to the new array.
2. deleteElementAtIndex ( A, i ) - Delete the element at A[i ].
Can you further adapt deleteElementAtIndex ( A, i ) so that it reduces the size of the
array if "too many" indices are empty, while still have only an O (1) amortized cost?

2. A credit invariant is a statement about the credit availability in the data structure. The amor-
tized costs (that is, number of credits required by the operation) of an operation must be
assigned in such a manner that even after its execution, the credit invariant is maintained.
For each part in the preceding question, specify the credit invariant for the data structure
and argue that no operation violates it.

3. Suppose that an algorithm has access to l different operations, such that the ith operation
has an amortized cost of ci . If the algorithm has executed k = ∑ ai operations, where ai
1≤ i ≤ l
is the number of times the ith operation was executed, then prove the following.
To the point at which k operations have been executed, the algorithm will have
run for O (∑i ai ci ) steps in the worst case.

4. Consider a specialized variant of the binary heap in which deleteMin and decreaseKey op-
erations, respectively, have amortized costs of Θ(log n) and Θ(1). Suppose that, over the
course of the algorithm, these operations are respectively executed n, m times.
We can then conclude that the algorithm’s complexity is Θ(m + n log n).
Complete the statement using one (or more) of the following options.
1. amortized 2. worst case 3. average case

You might also like