On Thu, Mar 17, 2011 at 4:06 AM, David Roe <r...@math.harvard.edu> wrote: >> > for is a function of the hash of its elements to >> > satisfy equal matrices over different parents having the same hash. >> >> Err, sorry, I don't manage to parse the end of the sentence. Do you >> mind reformulating? > > I think what Robert was trying to say is that if R and S are different > rings, and M is a matrix over R then > hash(M.change_ring(S)) == hash(M) as much as possible.
Exactly. On Thu, Mar 17, 2011 at 7:29 AM, Jeroen Demeyer <jdeme...@cage.ugent.be> wrote: > On 2011-03-17 15:04, Jason Grout wrote: >> def hash(self): >> h=0 >> for i,j,entry in m.nonzero_entries(): # nonzero entries for sparse >> matrices >> h^=hash(entry)^i^j >> return h > > Since you're only xorring, this will give a lot of collisions. I think > something like > > def hash(self): > n = m.ncols() > h = (m.nrows() + m.ncols())**2 + n > for i,j,entry in m.nonzero_entries(): > h ^= hash(entry)*(i*n + j + 1) > return h > > has less chance of collisions. Of course everything should be computed > as a C long. This is almost the current definition, see http://hg.sagemath.org/sage-main/file/361a4ad7d52c/sage/matrix/matrix_dense.pyx#l37 Initializing h with the dimension and not ignoring the 0th entry would be an improvement. We still need to support specialized implementations such as http://hg.sagemath.org/sage-main/file/361a4ad7d52c/sage/matrix/matrix_mod2_dense.pyx#l299 which is 100s of times faster than the generic implementation (and dense mod 2 matrices tend to be large). - Robert -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
