On Sun, Jan 3, 2010 at 1:57 PM, Alex Ghitza <aghi...@gmail.com> wrote: > > Hi Kilian, > > I am forwarding this to the sage-nt mailing list as well since you might > get a larger audience. > > > Best, > Alex > > > On Sun, 3 Jan 2010 13:51:19 -0800 (PST), Kilian <kkil...@googlemail.com> > wrote: >> Hello, >> >> i have a problem with sage and modular symbols for Gamma1(4) and odd >> weight k, where the cusp 1/2 is irregular. >> >> According to Merel, there is (for k>2) an exact sequence: >> >> 0-> S_k -> M_k -> B_k -> 0 >> >> Here B_k is the boundary space and S_k is the cuspidal subspace. >> >> Let the weight k be 7. >> >> If I compute the appropriate dimensions with SAGE, I get 4,6 and 3 >> which can't be. Furthermore, computing the boundary map, gives a >> matrix which is definitely _not_ surjective. >> >> On the other hand, Merel explicitely states that the dimension of B_k >> is the number of cusps, i.e. 3, so the failure must already be in >> Merel's paper, or am I missing something?
The B_k in Merel's paper has dimension 2. Merel does not state that dim(B_k) is the number of cusps in general. That's only true when the weight is even. Sage does have a very small bug, which is that it computes the correct space B_k but embeds it (trivially) in a bigger space. There is no need to do this, and I can see how it could be confusing. The correct relations are used, the correct map is computed, it's just that there are extra 0's tacked on. For example, in your example we have the following matrix for the boundary map: [-1 0 0] [ 0 -1 0] [ 0 -1 0] [ 0 -1 0] [ 0 0 0] [ 0 1 0] notice that the extra dimension -- the 0 in the last column -- isn't involved. The fix for this bug is to remove all the cusp classes that are equivalent to 0 because of the relation [Gamma(lambda u; lambda v)] ~ sign(lambda)^k[Gamma (u;v)] For example, in your example that would be the class (u;v) = (1;2). See trac 7837 http://trac.sagemath.org/sage_trac/ticket/7837 >> I assume that 4 and 6 are correct, as a comparison with the usual >> dimension tables for modular forms suggest. Yes. >> What is even more confusing is that Merel states that the isomorphism >> between the boundary space and the space B_k(Gamma) is an >> _isomorphism_, whereas in the SAGE sourcecode and in William Stein's >> book it is only stated that it's injective. Injectivity is all that is needed for the algorithm. -- William >> >> Thanks in advance, >> Kilian. >> > > > -- > Alex Ghitza -- Lecturer in Mathematics -- The University of Melbourne > -- Australia -- http://www.ms.unimelb.edu.au/~aghitza/ > > -- > > You received this message because you are subscribed to the Google Groups > "sage-nt" group. > To post to this group, send an email to sage...@googlegroups.com. > To unsubscribe from this group, send email to > sage-nt+unsubscr...@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/sage-nt?hl=en-GB. > > > -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org -- 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