 Open Access
 Total Downloads : 592
 Authors : Vishweshwarayya C Hallur, Ramesh K, Basavaraj A Goudannavar
 Paper ID : IJERTV1IS5246
 Volume & Issue : Volume 01, Issue 05 (July 2012)
 Published (First Online): 02082012
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Index Selection Sorting Algorithm
INDEX SELECTION SORTING ALGORITHEM
Vishweshwarayya C Hallur1
Angadi Institute of Technology & Management, Belgaum
Ramesh K2 Assistant Professor
Department Computer Science Karnatak University, Dharwad580003
Basavaraj A Goudannavar3
Department of P.G. Studies and Research in Computer Science Karnatak University, Dharwad580003
Abstract:
One of the most frequent operations performed on database is searching. To perform this operation we have different kinds of searching techniques. These all searching algorithms work only on data, which are previously sorted. An efficient algorithm is required to make searching algorithm fast and efficient. This paper presents a new sorting algorithm named as Index Selection Sorting Algorithm (ISSA). This ISSA is designed to perform sorting quickly and easily and also efficient as existing algorithms in sorting.
Key Words: Algorithm, Sorting, ISSA, Worst Case, Average Case, and Best Case.

Introduction
Using a computer to solve problem involves directing it on what step it must follow to get the problem to be solved. The step it must follow is called an algorithm. The common sorting algorithm can be divided into two classes by the difficulty of their algorithms. There is a direct correlation between the complexity and effectiveness of an algorithm [1].
The complexity of an algorithm generally written in the form of Big O(n) notation, where O represents the complexity of the algorithm and value n represents the size of
the list. The two groups of sorting algorithm are O(n2), which includes bubble sort, insertion
sort, selection sort, and shell sort. And O(n log(n)) which includes the heap sort, merge sort and quick sort[2].
Since the drastic advancement in computing, most of the research is done to solve the sorting problem, perhaps due to the complexity of solving it efficiently dispite its simple and familiar statement.
It is always very
difficult to say that one sorting technique is better than another. Performance of the various sorting algorithms depends upon the data being sorted. Sorting is used in most of the applications and there have been plenty of performance analyses [3][4].
There has been growing interest on enhancements to sorting algorithms that do not have an effect on their asymptotic complexity but rather tend to improve performance by enhancing data locality [2][3][5].

Proposed System
In ISSA technique the first number will be compared with all the elements in the list, at the end of each pass selection of proper index of new list is done and then element is copied it to that position in the new list. And this step will be repeated for n number of times.
The best case time complexity of ISSA is Omega (n2), the average case of the ISSA is theta (n2) and worst case of ISSA is Big O(n2).
Diagrammatic representation of ISSA:
Unsorted List with size a[5]
345
565
49
23
232
a[0] a[1] a[2] a[3] a[4]
New List i.e. b[5]
a[0] a[1] a[2] a[3] a[4]
After the last pass, the index of 232 is calculated and then 232 is copied into that position in the new list. Therefore index of 232 is 2. Hence,
23
49
345
565
a[0] a[1] a[2] a[3] a[4]

Algorithm Algorithm ISSA(a,b,n) for i 0 to n1
k 0
item a[i]
for j 0 to n1
if (item > a[j]) then
increment k
After the first pass, the index of 345 is calculated and then 345 is copied in to that position in the new list. Therefore index of 345 is 3. Hence,
345
a[0] a[1] a[2] a[3] a[4]
After the second pass, the index of the 565 is calculated and then 565 is copied into that position in the new list. Therefore index of 565 is 4. Hence,
345
565
a[0] a[1] a[2] a[3] a[4]
b[k] item

Comparisons of ISSA with other sorting technique
Below are the tables representing the calculated running time for n values and their graphs in various cases.
Best Case
Insertion
Quick
Shell
Index
Selection
8
8
7.22
6.52
64
16
16
19.26
23.19
256
32
32
48.16
72.49
1024
64
64
115.59
208.78
4096
128
128
269.72
568.36
16384

Best Case:
After the third pass, the index of 49 is calculated and then 49 is copied into that position in new list. Therefore index of 49 is 1. Hence,
49
345
565
a[0] a[1] a[2] a[3] a[4]
After the fourth pass, the index of 23 is calculated and then 23 is copied into that position in the new list. Therefore index of 23 is 0. Hence,
23
49
345
565
a[0] a[1] a[2] a[3] a[4]
20000
Running Time
15000
10000
5000
0
8
16
32
64
128
No. of elements
Insertion Quick Shell

Average Case:
Avarage Case
Insertion
Quick
Shell
Index
Selection
8
64
7.22
13.45
64
16
256
19.26
32
256
32
1024
48.16
76.10
1024
64
4096
115.59
181.01
4096
128
16384
269.72
430.53
16384
20000
Insertion
No. of elements
Shell
Quick
10000
0
Running Time
8
16
32
64
128

Worst Case:
Worst Case
Insertion
Quick
Shell
Index Selection
8
64
64
22.62
64
16
256
256
64
256
32
1024
1024
181.01
1024
64
4096
4096
512
4096
p>128 16384
16384
1448.1
16384
No. of elements
Shell
Quick
Insertion
20000
15000
10000
5000
0
Running Time
8
16
32
64
128
From the above graphs one can easily observe that with all the cases it takes same time and in worst case it takes same time like other sorting algorithms except quick sort technique.


Conclusion
Logic of ISSA is based on the logic of selection sort and insertion sort. In those techniques either the smallest or largest elements are taken and then placed them in appropriate position but in this ISSA first element, second element, third element and so on from the unsorted list is taken and then it is placed in its appropriate position in new list.

References