Abstract
An array preparation method has been discovered that can optimize almost any existing sorting algorithm. The preparation routine is called Sift, and the algorithm organizes the values of the array through the use of heap structures and standard comparison and swap functions. The algorithm adds only O(n) to the data comparison and move totals, and experiments have consistently reduced O(n2) sorting algorithm times to fractions of their previous amounts. The Sift Algorithm can also be used on O(n log n) algorithms to guard against O(n2) behavior by shifting median values towards the middle of the array and arranging the list in an ascending manner.
More Information
I will post more information about this algorithm after I decide how I'd like to go about publishing the information. Much of my research for the McNair Scholarship program will focus on patterns and Sift and sorting algorithms which should be available in the Fall of 2005. These publications will be available to those who wish to find them at the Midgett Building at MTSU, room 106.
Downloads
Algorithm's first implementation: 1/10/2005 | Webpage creation: 4/10/2005 | Last Update: 4/10/2005