Then does unsorted_multi_pass use extra memory than sorted? If yes, how much? If I sort myself, I guess I have to sort the bundled property map of the edges along with the edges array. Dan
Date: Fri, 16 Apr 2010 20:44:23 -0400 From: jewillco@osl.iu.edu To: boost@lists.boost.org Subject: Re: [boost] Requesting permission to merge r61326 into release branch
On Fri, 16 Apr 2010, Dan Jiang wrote:
Hi Jaremiah, My CSR graph now works with push_relabel_max_flow. Many thanks to your help without which I would have still be struggling! Right now I still use edges_are_unsorted, I already see more than 30% memory saving. I'll sort the edges myself and see how much more I can save. Thanks again and BOOST rocks!
I'm not sure you'd want to sort the edges yourself unless you can do it out of core (such as in a database query) or some other special case. The CSR graph's sorting routines are fairly fast (linear time); the constants are better with the unsorted_multi_pass versions than the straight unsorted versions.
-- Jeremiah Willcock _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
_________________________________________________________________ Hotmail & Messenger. Get them on your phone now. http://go.microsoft.com/?linkid=9724463