Here are some lovely graphs borrowed from http://linux.wku.edu/~lamonml/algor/sort/sort.html. The first graph shows execution time, represented in seconds on the vertical axis and the number of items in a data set on the horizontal axis. The second graph shows execution time in tenths of a second. You can perhaps guess which sort of complexity is represented by each graph!

The n log n algorithms absolutely crush the quadratic algorithms in terms of speed.

Conclusion

In this article we have reviewed the primary sorting algorithms and their general properties. We have reviewed how quickly they complete relative to the size of the data set to be sorted. We have also discussed how to choose the appropriate algorithm. The metrics presented in this article were borrowed from http://linux.wku.edu/~lamonml/algor/sort/sort.html, who was kind enough to compile them, knowing that eventually I would need them-–so, thanks to wku.edu.

As I wrote this article, it occurred to me that we could go one step further and throw in an implementation of the Strategy design pattern using the contents of this article. We will create a SortStrategy class and create a method in the base Collection class to select a SortStrategy class based on the data it contains. Since we have reviewed the best and worst of our two classes of algorithms and know how complex (or simple) they are, we will pick the “best” from each class and go from there.