Scribd
Upload
2 views2 pages

2 Running Time

The document analyzes the running time of insertion sort, indicating that the best case is Q(n) for already sorted arrays and the worst case is Q(n^2) for reverse sorted arrays. It introduces the divide and conquer approach, exemplified by merge sort, which has a running time of Q(n log n). Other sorting algorithms and their complexities, such as bubble sort and binary search, are also briefly mentioned.

Uploaded by

viraguber3
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
2 views2 pages

2 Running Time

The document analyzes the running time of insertion sort, indicating that the best case is Q(n) for already sorted arrays and the worst case is Q(n^2) for reverse sorted arrays. It introduces the divide and conquer approach, exemplified by merge sort, which has a running time of Q(n log n). Other sorting algorithms and their complexities, such as bubble sort and binary search, are also briefly mentioned.

Uploaded by

viraguber3
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

LECTURE 2

ANALYSIS OF INTSERTION SORT

Running time T(n) =


c1n + c2(n-1) + c4(n-1) + c5S tj + c6S (tj-1) + c7S (tj-1) + c8(n-1)
Here:
1. Subscripts refer to statements in the pseudocode.
2. tj is the number of times the while loop test is executed for that value of j.
3. Summation S runs over all values of j, that is j=2 to j=n.

Best case occurs for already sorted array, in which case no values need to
moved from one place to another. Running time is Q(n).

Worst case occurs for reverse sorted array, in which case tj, the number of
executions of the loop test, is at its maximum. Running time is Q(n2).

Even if we assume that tj = j/2 on average, we still get running time Q(n2).

DIVIDE AND CONQUER APPROACH

1. Divide the problem into a number of sub-problems.


2. Solve the sub-problems recursively, but if a sub-problem is “small enough",
solve it in a “straightforward” manner, without further sub-division.
3. Combine the solutions to sub-problems into the solution to the original
problem.

EXAMPLE:

MERGE-SORT( A, p, r )
if p < r
q = floor((p+r)/2)
MERGE-SORT( A, p, q )
MERGE-SORT( A, q+1, r )
MERGE( A, p, q, r )
For the function MERGE, please see [CLRS] <-- SELF STUDY.

ANALYSIS OF RUNNING TIME

DIVIDE step: Constant running time, denoted as D(n) = Q(1).

RECURSION step: T(n) = 2T(n/2)

MERGE step: C(n) = Q(n).

Combining these:

Q(1), if n = 1
T(n) =
2T(n/2) + Q(n), if n > 1

Solution is T(n) = Q(n log2n) ) = Q(n lg n).

[We accept without proof, but we can always check by back-substitution.]

Solution can be understood by making use of a balanced binary tree, assuming


for that purpose that n = 2k, for some integer k.

A so-called “master theorem" sets out the general case, as we will see later.

OTHER EXAMPLES

In Bubble-sort, we count the number of exchanges. Running time Q(n2).

In matrix multiplication, we count the number of scalar multiplications. Simple


algorithm gives running time Q(n3).

Binary search in a sorted array takes running time Q(log2n).

Finding the maximum or minimum element in an unsorted array takes running


time Q(n).

You might also like