05-05-2016 00:09

Different methods for sorting, such as sorting by counting, by insertion (straight insertion, binary insertion and two-way insertion), shell’s sorting, bubble sort, quick sort, heap sort, sorting by merging (straight two-way merge sorting, list merge sort), and sorting by distribution (radix sorting) are presented in the last few years. In this work, a specific implementation of “A Modified Sorting Algorithm Using Undefined Arrays Size” is presented whose advantage are that (1) the method is very easy to implementation, (2) the first algorithm has number of comparisons less than the others sorting method which has O(N2) comparisons, (3) the second algorithm has a number of comparisons which is less than the other method except the Quick sort method, (4) the second algorithm has number of comparisons approximately equal to  the number of comparisons of Quick sort method, (5) The distribution process for a group of items into subarrays, solve the problem of array size (i.e. if the number of items exceed more than 30,000 then there will be no problem[2]). In this work, comparisons tables to display the number of comparisons at N items, and a comparisons figures are presented. In a worst-case the new method is faster than Quick sort, because the new sorting method is faster than any method having O(N2) comparisons.