General ideas
Ondrej Certik
I am following the issues and progress of sympycore, but I generally
agree with what you are trying. Here are some ideas what I would like
to bring to SymPy first. After discussing it in SymPy Issues:
* dealing with unapplied functions etc. There will probably not much
discussion, I think we all agree on the sympycore approach.
* assumptions. This is I think still unresolved even in sympycore.
* the speed improvements. But I'll wait with this one, first there
will be a lot of discussion about each feature, I am sure abou that
:), and second I think it still isn't settled even in sympycore yet.
The assumptions are the most urgent and also the most difficult, since
currently SymPy is sometimes not calculating correctly.
Ondrej
Pearu Peterson
> Hi,
>
> I am following the issues and progress of sympycore, but I generally
> agree with what you are trying. Here are some ideas what I would like
> to bring to SymPy first. After discussing it in SymPy Issues:
>
> * dealing with unapplied functions etc. There will probably not much
> discussion, I think we all agree on the sympycore approach.
yes, it is quite solid approach that is in sympycore, I think.
> * assumptions. This is I think still unresolved even in sympycore.
Yes, I got in a way of doing other stuff like implementing
sets and logic support first, plus doing the restructuring
sympycore tree which takes lots of effort but I think
it is worth it.
> * the speed improvements. But I'll wait with this one, first there
> will be a lot of discussion about each feature, I am sure abou that
> :), and second I think it still isn't settled even in sympycore yet.
Yes, improving speed is an important task for sympycore: if there
arises some idea of doing things faster I'd rather implement it
as soon as possible.
> The assumptions are the most urgent and also the most difficult, since
> currently SymPy is sometimes not calculating correctly.
I agree.
Pearu