On Thu, Aug 21, 2008 at 11:18 AM, Ondrej Certik <[EMAIL PROTECTED]> wrote:
>
> On Thu, Aug 21, 2008 at 7:39 PM, Robert Dodier <[EMAIL PROTECTED]> wrote:
>>
>> On Aug 20, 11:18 pm, Tim Lahey <[EMAIL PROTECTED]> wrote:
>>
>>> 1. Commands to parse expressions.
>>
>> Maxima has various expression-hacking functions.
>> Most objects are expressions.
>>
>>> 2. Like M. Hampton, I miss implicit variables.
>>
>> Maxima makes default assumptions about variables,
>> and doesn't require them to be declared before use.
>> This is part of Maxima's general sloppiness but it does make
>> for ease of use.
>>
>>> 3. Somewhat related to #1, is the ability to make new variables/
>>> function names from old ones.
>>
>> Maxima function concat.
>>
>>> To illustrate this, my Maple code for this is:
>>
>> I glanced at the code and it looks like it should be translateable
>> to Maxima. Maybe I'll try it this weekend.
>>
>> I'm pushing Maxima here for two reasons; one is that the solution
>> for the problems stated here might be close at hand after all.
>> More selfishly, I'd like to get more people using and (I hope)
>> developing Maxima rather than reinventing its wheels.
>
> +1, however, how about speed? Are you able to make maxima as fast as
> ginac? And faster, because (unless I am mistaken) Mathematica is even
> faster than ginac?
By the way, with the new ginac stuff in sage, here is a little synthetic
benchmark of something:
MATHEMATICA:
Timing[a = Expand[(x^Sin[x] +y^Cos[y] - z^(x+y))^100];]
{0.180212, Null}
SAGE:
sage: time a = expand((x^sin(x) + y^cos(y) - z^(x+y))^100)
CPU times: user 0.15 s, sys: 0.00 s, total: 0.15 s
So Mathematica isn't always faster than Ginac.
Incidentally Sage/Ginac is faster above since I changed Ginac
to directly use gmp for binomial coefficients instead of using CLN;
with CLN Ginac is slower at the above.
But this is just one stupid tiny datapoint in an massive multidimensional
space of possible benchmarks.
William
On Thu, Aug 21, 2008 at 11:23 AM, William Stein <[EMAIL PROTECTED]> wrote:
> On Thu, Aug 21, 2008 at 11:18 AM, Ondrej Certik <[EMAIL PROTECTED]> wrote:
>>
>> On Thu, Aug 21, 2008 at 7:39 PM, Robert Dodier <[EMAIL PROTECTED]> wrote:
>>>
>>> On Aug 20, 11:18 pm, Tim Lahey <[EMAIL PROTECTED]> wrote:
>>>
>>>> 1. Commands to parse expressions.
>>>
>>> Maxima has various expression-hacking functions.
>>> Most objects are expressions.
>>>
>>>> 2. Like M. Hampton, I miss implicit variables.
>>>
>>> Maxima makes default assumptions about variables,
>>> and doesn't require them to be declared before use.
>>> This is part of Maxima's general sloppiness but it does make
>>> for ease of use.
>>>
>>>> 3. Somewhat related to #1, is the ability to make new variables/
>>>> function names from old ones.
>>>
>>> Maxima function concat.
>>>
>>>> To illustrate this, my Maple code for this is:
>>>
>>> I glanced at the code and it looks like it should be translateable
>>> to Maxima. Maybe I'll try it this weekend.
>>>
>>> I'm pushing Maxima here for two reasons; one is that the solution
>>> for the problems stated here might be close at hand after all.
>>> More selfishly, I'd like to get more people using and (I hope)
>>> developing Maxima rather than reinventing its wheels.
>>
>> +1, however, how about speed? Are you able to make maxima as fast as
>> ginac? And faster, because (unless I am mistaken) Mathematica is even
>> faster than ginac?
>>
>> In SymPy at least, we want to be as fast eventually.
>>
>
> I'm really sick of this. "Fast" is meaningless without specific benchmarks,
> and then only certain ones should matter. Here's a list to get going:
>
> http://wiki.sagemath.org/symbench
>
> It would be nice to have some specific concrete examples involving
> pattern matching, substitution, etc. Please post something concrete
> to this list that matters to you. It doesn't have to be possible in sage now
> (see e.g., W5 at the above page).
>
> William
>
--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
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://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
