I am absolutely no expert in sage, and I know next to nothing about Python, so I apologize if my question is stupid or ill-formulated (and even if it so, I am likely to learn something from any answer).
I am puzzled by the following: I have a list B of q-expansions of modular forms over GF(2), which is a basis of a space of modular forms stable by the Hecke operators, and a list B' which contains the same elements as B but in a different order. I am trying to compute the matrix of the Hecke operator T_3 in the basis B and B', using hecke_operator_on_basis(B,3,2) and the same with B'. I get a correct answer for B and an error message for B'. How is that possible? I suspect that it is a question of degree of precision, but what exactly is happening here? Please see http://www.sagenb.org/home/pub/3733 for a concrete example.