Tuesday, August 2, 2011

Insertion Sort

Insertion Sort - Algorithm

Insertion sort algorithm somewhat resembles selection sort. Array is imaginary divided into two parts - sorted one and unsorted one. At the beginning, sorted part contains first element of the array and unsorted one contains the rest. At every step, algorithm takes first element in the unsorted part and inserts it to the right place of the sorted one. When unsorted part becomes empty, algorithm stops.

Insertion Sort - Implementation

package main.java.algo.sorting;

public class InsertionSort {

	public static void main(String[] args) {
		InsertionSort s = new InsertionSort();
		int[] array = { 7, -5, 2, 16, 4 };
		array = s.insertionSort(array);
		for (int a : array) {
			System.out.println(a);
		}
	}

	private int[] insertionSort(int[] array) {
		int length = array.length;
		int tmp = 0;
		for (int i = 1; i < length; i++) {
			for (int j = i; j > 0; j--) {
				if (array[j -1] > array[j]) {
					tmp = array[j];
					array[j] = array[j-1];
					array[j-1] = tmp;
				}
			}
		}
		return array;
	}
	
}


Insertion Sort - Complexity Analysis

The outer loop goes n times
Inner nexted loop goes i-1 times. where i is the counter of outer loop.
Since i ranges 0 -to->n , for complexity we take the worst case for i=n
Thus total iterations = n * (n-1)
Complexity = n square

Space complexity - It is inplace with use of very few constants.
Thus -> In Place. O(1)

Best Case = O(n) // occurs when the list is already sorted ascending - it picks up the next element and just inserts into the last spot of sorted array.
Worst case = O(n^2) // occurs when the list is already sorted descending - when it picks up the new element, it has to loop through entire sorted array and place the element at 1st place
Average Case - O(n^2) // although the average-case running time is approximately half of the worst-case running time, it is still a quadratic function of n.


Insertion Sort - Comparison with Selection sort

1) Lookups
- In every iteration, Selection for always looks through the remaining elements to find min.
- But Insertion sort, picks the element and just loops through k-1 times to fit the element at its place.
- This Insertion sort is faster in performance.

2) Writes
- Because of above reason, Selection sort always does 1 swap per iteration
- Insertion does swap k-1 times till it has found the rough place for picked element
- Thus insertion sort should be avoided when writes are expensive compared to reads.

3) Variants
- Selection sort does not vary in best/worst case or the oredering of list
- Insertion sort performs better in desc ordered rather than asc ordered.

4) Use selection fort 10-20 elements
Use Insertion sort for 20-80 elements.

No comments:

Post a Comment