The fact that the discrepancy is 46273*46153 (both primes) makes me suspect that there's a factor of 2 missing in the CRT bounds, to allow for $\pm$. But I don't have the source here to check.
On Thursday, 26 April 2012 11:19:49 UTC+1, Alastair Irving wrote: > > On 25/04/2012 16:45, Graham Gerrard wrote: > > Finding occasional inconsistencies when using matrices with cyclotomic > > entries, though works well most of the time... > > > > sage: s=CyclotomicField(24,'s').gen() > > sage: (8*s^6-1)^10 > > -1098715216*s^6 - 372960063 > > sage: xb=matrix(1,1,[8*s^6-1]) > > sage: xb^10 > > [1036922553*s^6 - 372960063] > I can confirm this behaviour. > > > > Above example is trivial computation, though problem recurs in several > > contexts. I believe that scalar multiplication is correct and xb^10 > > incorrect. > Yes, I've checked with Magma and you're right. > > > > Any suggestions? > Sage has a special algorithm for multiplying matrices over cyclotomic > fields which works modulo primes and then does a Chinese remainder to > put the answer together. There is presumably a problem somewhere in > this algorithm but I don't know where. A rather unpleasant workaround > is to do > xb.sparse_matrix()^10 > Sage just uses a generic algorithm for multiplying the sparse matrices > so this gives the correct answer. However, the performance probably > won't be as good. > > Alastair > -- To post to this group, send email to sage-support@googlegroups.com To unsubscribe from this group, send email to sage-support+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-support URL: http://www.sagemath.org
