Hi Torsten,
Thanks so much. I have read some from this link: http://web.it.kth.se/~cschulte/papers/SchulteCarlsson_CPH_2006.pdf Is that the book that you are referring? If you are familiar with Search Module of Mozart/Oz, I have a specific question. When one uses {Search.best.bab ..} directly vs if someone uses the script {new Search.object script( ...)} and then call for next solution (or such) - is there any difference? Do the former uses copying vs the latter uses the "trailing" concept? Or they both probably use the copying? I stumbled upon an interesting problem yesterday. I have an input that has, say 14 tasks, and it takes a second to get the solution and finish the search space. But as soon as I add another task (that is very similar to a task before), although it gets a solution quickly enough, but the search space exploration continues for minutes! Is it the problem with filtering? Idempotent propagators? If so how to get around that? I seemed to see similar problems in our real problems, where although a solution is obtained quickly enough but the search space exploration continues for minutes, hours and never finishing? Please let me know I was able to describe my problem! Thanks, Ashis ________________________________ From: [email protected] [mailto:[email protected]] On Behalf Of Torsten Anders Sent: Thursday, June 18, 2009 10:28 AM To: Mozart users Subject: Re: Constraint Definition On 18 Jun 2009, at 13:36, Maity, Ashis K wrote: Are there some data points (references) as to how big a problem Oz could handle and its resource usage statistics? I think the best reference is again Christian's book, he did lots of performance measurements and compared the space-based approach with the more common backtracking-based approach (copying vs trailing). If I remember correctly, copying (i.e. Oz) is as efficient as trailing concerning space consumption if you use adaptive recomputation. If the performance of Oz is not enough for you, then you might be interested to look at Gecode or also GeOz (the latter is work in progress and not released yet). Best Torsten -- Torsten Anders Interdisciplinary Centre for Computer Music Research University of Plymouth Office: +44-1752-586219 Private: +44-1752-558917 http://strasheela.sourceforge.net http://www.torsten-anders.de
_________________________________________________________________________________ mozart-users mailing list [email protected] http://www.mozart-oz.org/mailman/listinfo/mozart-users
