RE: Challenge for Sort programs.
- Posted by kbochert at ix.netcom.com Jun 03, 2002
- 411 views
-------Phoenix-Boundary-07081998- You wrote on 6/3/02 3:47:47 PM: > >Hi all: >I was attempting to improve the performance of the Shell sort that comes >with Euphoria. <snip> >Well, after this rant, you will find the code for all these sorts, plus two >slightly improved versions of Bubble sort, and Selection sort, one of the >basic ones (Rob: why is it not included in Allsorts.ex=3F). >The challenge to find faster sorts remain open. I hope to have contributed >some ideas on this subject. This got me thinking about sorts and how the fastest of all sorts is the partition sort, also known as quicksort. I wrote one for Euphoria, and although it does out-perform the others, its advantage is not as great as I might have expected. I just wrote it with no concern for Euphoria optimizations, so there might be room for considerable improvement. I have also not tested it on a variety of data (of course). I am attaching the code for the quicksort, a test harness and the code from your message. Karl Bochert -------Phoenix-Boundary-07081998- Content-type: application/octet-stream; Name=rforno.e