The documentation bug has been corrected on the master branch. Concerning the components data of A(i, j) + A(j, i), I thought that the faster solution it to remap the axes of the numpy.ndarray to the lexicographically sorted indices. That is, the components data of both A(i, j) and A(j, i) would both be sorted as (i, j). This would break compatibility with A(j, i).data, which would get transposed on future versions of SymPy.
An alternative is to make indexing mandatory when there is no evident index order. That is, A(i, j).data is OK, while (A(i, j) + A(j, i)).data would raise an error. Maybe a new method could be introduced, say .get_data_by_index_order(i, j) -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sympy. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/79d42911-a907-4ed7-95cb-c5dae492566c%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
