On Nov 23, 8:38 am, William Stein <wst...@gmail.com> wrote: > On Mon, Nov 23, 2009 at 8:28 AM, rjf <fate...@gmail.com> wrote: > > From the proposal > > > ... and which has sophisti- > > cated interfaces to nearly all other mathematics software, including > > Mathematica, Maple, > > MATLAB and Magma. ... > > > Maxima just gets no respect. :) > > Most of the facilities mentioned are already in Maxima. > > I wish...
I seem to recall that Sage people had their knickers in a knot when the interface to Maxima's eigenvalue package changed. Of course you might believe that Maxima has no linear algebra, and no ODE solvers and no ... whatever. Maybe your wish came true when you weren't looking? > > > And why is Cython much more than a Python to C translator? (This is > > not sarcasm. I honestly have no idea that it was more. > > I thought it was, if anything, less.) > > It is much, much more than that. So now you know. That's helpful. Forgive me for asking for information. I see, somewhat later, that this has to do with adding type declarations. Just the ticket. To me is suggests that Python is inappropriate for numerical work -- for which C is more appropriate. And so the veneer of Python is insufficient. Of course using (optional) declarations in Lisp helps some compilers to produce better numerical code. Some of that code generation (depending on the Lisp system) is even done with C as an intermediary. And has been done that way since, oh, 1984 or so. > > "venerable" Maxima is mentioned once, suggesting that the only thing > > it can do is symbolic integration and numeric integration. > > Actually, while Maxima includes library access to Fortran methods, it > > is far inferior to what could be done in numeric integration, > > as demonstrated by recent Mathematica versions. You would hardly get a > > hint that 75% of the sage-support messages are about Maxima. > > No they aren't. I have supplied the calculation already. Open up a sage-support window. There are 3480 or so messages. Now search for "maxima" and you will have a hit on 75% of them. > > Moreover, there isn't anything in the proposal that uses Maxima at > all. The proposal is about numpy/scipy/PDE/linear algebra/algebraic > topology/group theory/differential algebra/pynac; this is all > independent from Maxima. You would think "Unifying Mathematical Software for Scientists, Engineers and Mathematicians" might have something to do with Scientists and Engineers. But my mistake. Apparently its purpose is to allow those other people access to algebraic topology/group theory/differential algebra. That's what unifying means to a mathematician ? > > > Maybe what is needed is a Fortran to Python translator. > > :-) > > > I think that if NSF sent the proposal over to computer science and > > engineering, it might not get a great reception, but it is hard to > > predict such things. > > As I mentioned at the top of this thread, the proposal is to the the > computational mathematics program, which is part of DMS = "Division of > Mathematical Sciences". As you may or may not know, the various program directors at NSF are free to "shop around" proposal that seem to them to be wrongly categorized, interdisciplinary, or otherwise in need of others' guidance. RJF > > -- William -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
