On Mon, Sep 05, 2011 at 11:59:02PM -0700, Anne Schilling wrote: > I now implemented another version of this method which is recursive > (the two methods are bruhat_upper_cover and bruhat_upper_cover_old). > I am not sure the complexity is any better though.
I like it, and am pretty sure it is much better. Can you make a quick benchmark? Say by computing all bruhat upper covers for S4 / S5 / S6 using the old and new algorithm, as well as using the method for finite coxeter groups? > That is ensured by the conditions on the descents of the left and right > factor. That's what I was pondering: those conditions sure guarantee that there is no immediate cancellation with the left or right factor; it's not immediately obvious to me that it does guarantee that there is no more complicated cancellation involving the left and right factors and the s_i in the middle. > Ok, I deleted this for the old method, but kept it for the new > method since it is recursive. +1! Cheers, Nicolas -- Nicolas M. ThiƩry "Isil" <nthi...@users.sf.net> http://Nicolas.Thiery.name/ -- You received this message because you are subscribed to the Google Groups "sage-combinat-devel" group. To post to this group, send email to sage-combinat-devel@googlegroups.com. To unsubscribe from this group, send email to sage-combinat-devel+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/sage-combinat-devel?hl=en.