Hi all,
I am the lead developer of SymPy and I just wanted to clarify a few
things about SymPy's motivations.
1) Speed is not a (top) priority. Rewriting it in C++ will maybe be
done in the future, but definitely not now, as we are still playing
with the design. However, I think that rewriting the core in C++ could
make it as fast as GiNaC.
2) We want to keep it as simple as possible, however, I want it to be
fast enough to be able to do basics calculations in physics, like
calculations of Ricci tensor from a metric (this SymPy already can and
it's reasonably fast, like 2s on my computer) and calculations in QFT
(currently SymPy is very slow on this).
3) We want to have a simple (but poweful) library, that people can
just install with
apt-get install python-sympy
(similarly on other systems) and use like:
>>> from sympy import Symbol, cos
>>> x=Symbol('x')
>>> e=1/cos(x)
>>> print e.series(x,10)
1+1/2*x^2+5/24*x^4+61/720*x^6+277/8064*x^8+50521/3628800*x^10
And it should be fast enough for at least a first sketch of a new
symbolic algorithm. Of course there are and there will always be more
specialized (and hence a lot faster) libraries, like FORM, GiNaC,
etc.
4) I want SymPy to be the first thing that a user in Python would
choose if he wanted to try to calculate something symbolically. And as
anything in python - if it turns out, it's not already fast enough, he
will then choose some faster, but harder to use library.
5) On the other hand, SAGE, as I understand it, wants to glue all the
best opensource (and maybe commercial) CAS programs and it wants to be
as fast as possible. So some of the motives are common, but not all.
It would be fine, if we could share knowledge, or maybe even some
code, for the common things, if there are any. For example I am not an
expert on groebner bases nor a mentor for that application. However, I
will think of finding a better algorithm for anything in SymPy only
when it turns out, that the current code is not fast enough. And you I
think are writing something similar as SymPy is doing, so I can share
some experience.
Ondrej
On Apr 13, 9:23 pm, "Timothy Clemans" <[EMAIL PROTECTED]>
wrote:
> If SymPy was written in C++ and had a Python interface in it what
> would be the likely hood of it being the backend for symbolic
> computation in SAGE?
>
> On 4/13/07, Timothy Clemans <[EMAIL PROTECTED]> wrote:
>
> > me: There are several SAGE people who are playing around with SymPy
> > and an optional package of SymPy for SAGE has been requested by
> > William
> > Sent at 11:50 AM on Friday
> > Ondrej: I see
> > feel free to ask on the sympy mailing list
>
> > On 4/13/07, Bill Page <[EMAIL PROTECTED]> wrote:
>
> > > On April 13, 2007 11:02 AM William Stein wrote:
>
> > > > On 4/13/07, Mike Hansen wrote:
> > > > > I definitely agree that SAGE's goals are quite a bit higher
> > > > > and more ambitious than those outlined on the SymPy project.
> > > > > While looking over the SymPy website, I was just surprised
> > > > > because I had never heard of the project and their scope
> > > > > seemed to be much wider than I had initially thought (quantum
> > > > > field theory calculations, a port of grtensorii, Groebner
> > > > > bases calculations, symbolic linear algebra, etc.) While
> > > > > there may be a limited (if any) amount we could benefit from
> > > > > their current codebase, I thought it'd be good just to be
> > > > > aware of any other work done with computer algebra in Python.
>
> > > > I certainly agree. At a minimum it would be nice to have a
> > > > SAGE optional package that install SymPy. Anybody want to
> > > > make one?
>
> > > It turns out that running SymPy in Sage is quite easy after
> > > you install the prerequisite pygame package that is already
> > > available as an experimental SAGE package
>
> > > sage -i pygame-1.7.1release
>
> > > I have been experimenting a little with it here:
>
> > >http://wiki.axiom-developer.org/SandBoxSymPy
>
> > > It might be interesting to know (and a bit galling - at least it
> > > was for me as an Axiom developer) to know that the SymPy got 5
> > > Google Summer of Code funded projects! Axiom partnered with LispNYC
> > > for SOC 2007 but we did not get any qualified applications. :-(
>
> > > Regards,
> > > Bill Page.
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [EMAIL PROTECTED]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~----------~----~----~----~------~----~------~--~---
