The big algorithms are an issue, yes, but they're still valuable to see even if they're just one step. Various functions use integrate within them. Even if we can't look inside integrate it'd be nice to know that it happened.
Having such a system in place might also encourage us to build smaller functions in the future. A substantial part of the GSoC project might involve seeking out and breaking up large functions when possible. On Sun, Feb 24, 2013 at 4:34 PM, Stefan Krastanov < [email protected]> wrote: > >> Not all of SymPy is written this way (lots of little functions) but a > >> substantial refactor of our giant functions into a few smaller ones > >> might allow this. > > > > I'm not sure how this would play with the "large" algorithms like f.e. > > Gruntz, Risch, MeijerG etc. And even if it would finally work out > > reasonable, the results are more for insight in CAS theory and debugging. > > These algorithms work fundamentally different that what a human would do > > by hand ("step-by-step") to solve the same problem. > > Indeed, but I think that we agree on that. All the > step-by-step/rewrite-rules discussions are about the algorithms like > simplification, etc, that already employ similar style. > > -- > 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?hl=en. > For more options, visit https://groups.google.com/groups/opt_out. > > > -- 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?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
