headers
Valery Pipin | 1 Mar 2004 04:36
Picon

Re: Mupad comparision

On Monday 01 March 2004 01:21, C Y wrote:
| Pekka Sorjonen has an interesting comparision between Maxima and Mupad
| at http://homepage.mac.com/peso1/index.html
|
| It looks like many of the Maxima errors are still present in the cvs
| version, so it's probably worth keeping in mind at some point in the
| future when we start addressing these type of errors.
|
| Maybe we could make a new test file - wester.mac or something.
|
| CY
Thanks for the reference!
My impression is that he skips many problem that are out of Mupad abilities, 
eg tensors. Currently I have no time to check it detally.

You mentioned  wester.mac.  Does it include all problems that are solved in 
Wester's review?

best rgds,
Valery
Permalink | Reply | 2 comments |
headers
CY | 1 Mar 2004 06:37
Picon
Favicon

Re: Mupad comparision

On Mon, 1 Mar 2004 11:36:39 +0800
Valery Pipin <pip <at> iszf.irk.ru> wrote:

> Thanks for the reference!
> My impression is that he skips many problem that are out of Mupad abilities, 
> eg tensors. Currently I have no time to check it detally.
> 
> You mentioned  wester.mac.  Does it include all problems that are solved in 
> Wester's review?

I suppose it could - it doesn't exist yet ;-).  It was just a suggestion.  Right now there would be a large # of
failures we aren't in a position to deal with.  (IIRC, isn't the Great Source Downcasing next?  That ought to
be good for a few months of fun.)  I just thought it was something we should keep in mind when we get to the point
where fixing things like our inequalities system is a high priority.

CY
Permalink | Reply | 1 comment |
headers
C Y | 1 Mar 2004 06:39
Picon
Favicon

Worth doing?

Somebody's putting together a software project reference for government
users: Open Source Reference Book 2004.  Worth adding Maxima to it?

__________________________________
Do you Yahoo!?
Get better spam protection with Yahoo! Mail.
http://antispam.yahoo.com/tools
Permalink | Reply | 0 comments |
headers
Valery Pipin | 1 Mar 2004 07:08
Picon

Re: Mupad comparision

On Monday 01 March 2004 13:37, CY wrote:
| On Mon, 1 Mar 2004 11:36:39 +0800
|
| Valery Pipin <pip <at> iszf.irk.ru> wrote:
| > Thanks for the reference!
| > My impression is that he skips many problem that are out of Mupad
| > abilities, eg tensors. Currently I have no time to check it detally.
| >
| > You mentioned  wester.mac.  Does it include all problems that are solved
| > in Wester's review?
|
| I suppose it could - it doesn't exist yet ;-).  It was just a suggestion. 
| Right now there would be a large # of failures we aren't in a position to
| deal with.  (IIRC, isn't the Great Source Downcasing next?  That ought to
| be good for a few months of fun.)  I just thought it was something we
| should keep in mind when we get to the point where fixing things like our
| inequalities system is a high priority.
Ok, 
I think it's  a good idea to bring Wester's tests  to maxima and go through.
I think I will try (but not immideatly, rigth now I'm busy with habilitation). 
Also there is an interesting website connecting to component tensor 
manipulation within maple. It is a good point to check maxima speed on those 
problems. From time to time, I hear the claim that maple is the fastest CAS 
in world.

best rgds,
Valery
Permalink | Reply | 0 comments |
headers
Billinghurst, David (CALCRTS | 2 Mar 2004 00:27
Picon

maxima windows package from cvs

I have built a windows installer from maxima cvs.
 - dosn't support spaces in folder names
 - works (including xmaxima) on NT4 and XP
 - command line maxima sort of works on windows 95 
   (and presumably 98/Me) but need to test some more
 - xmaxima doesn't work on windows 95

I don't have anywhere to host it, but can upload it somewhere and 
update it occasionally.

To get it working I needed to make some monor changes.  Open to suggestions
on better ways to acheive the same result.

1. Get the installer to generate a maxima.bat with the new %prefix%

This is done by creating a bat file that echoes the existing maxima.bat file 
to stdout, but with %prefix% given by a command line arg.  This is run by 
the installer to overwrite the original maxima.bat. 

2. Define MAXIMA_PREFIX=%prefix% in windows.bat

3. Add the full path to gcc in the variable compiler::*cc*

Done by adding the following line to maxima-init.lisp
(setq compiler::*cc* (concatenate 'string *maxima-prefix* "\\bin\\" compiler::*cc*))
Permalink | Reply | 1 comment |
headers
Vadim V. Zhytnikov | 2 Mar 2004 07:06
Picon

Re: xmaxima situation summary

Vadim V. Zhytnikov пишет:
> Valery Pipin пишет:
> 
>> On Thursday 26 February 2004 03:40, Vadim V. Zhytnikov wrote:
>> |snipped|
>> | It is not clear from your post what is the situation with xmaxima
>> | on Unix. I've just tried xmaxima on Linux with gcl, clisp, cmucl
>> | and sbcl. It seems that it works with clisp only.
>> On my system xmaxima works with gcl
>> | BTW maxima server mode don't work on sbcl producing Lisp error.
>> I confirm this. The system is debian-sid and maxima is taken from cvs.
>>
> 
> Never mind sbcl - I forget that xmaxima didn't work
> with sbcl before recent maxima/xmaxima changes.
> 
> But gcl and cmucl still are problem to me.
> I've build Maxima CVS with gcl 2.6.1 CVS, clisp 2.32
> and cmucl 18e.
> 
> xmaxima-local -l clisp - works fine.
> 
> xmaxima-local -l gcl - starts maxima but
> maxima subprocess terminates with the
> message "exiting on eof" after very first
> command.
> 
> xmaxima-local -l cmucl - starts maxima
> but if I try to enter any command - just
> nothing happens - cmucl maxima still
(Continue reading)

Permalink | Reply | 7 comments |
headers
Valery Pipin | 4 Mar 2004 13:26
Picon

does "algebra/solver/solver.mac" broken?

I tried the "test.mac" from the new solver package for maxima but it fails.
Is it broken?
(C1) load("algebra/solver/solver.mac");

(D1)        /usr/share/maxima/5.9.1/share/algebra/solver/solver.mac
(C2) block([s:solver([x+y+x^2=a,x-y=b],[x,y],[a,b])],print(s));

MACSYMA was unable to evaluate the predicate:
ERREXP1
#0: SetupSolver(equations=[y+x^2+x = a,x-y = b],solverparams=[[x,y],[a,b]])
#1: SOLVER(equations=[y+x^2+x = a,x-y = b],solverparams=[[x,y],[a,b]])
(solver.mac line 451)
 -- an error.  Quitting.  To debug this try DEBUGMODE(TRUE);)

Also,
To start the "solver mac" correctly the  call paths inside the package
should be replaced to "algebra/solver" instead ./ 

best rgds,
Valery
Permalink | Reply | 0 comments |
headers
C Y | 5 Mar 2004 04:08
Picon
Favicon

evaluate Maxima string from lisp command line...

Hey guys.  I'm trying to figure out how to take a string in regular
maxima input format, and feed it to Maxima on the lisp level.  I
haven't been able to sort out how to take diff(x^2,x) and evaluate it
on the lisp command line.  Can someone point out the syntax to me? 
Something like:

(setq expr (mread string))

(macsyma-eval expr)

would be what I'm after.

CY

__________________________________
Do you Yahoo!?
Yahoo! Search - Find what you’re looking for faster
http://search.yahoo.com
Permalink | Reply | 7 comments |
headers
CY | 5 Mar 2004 07:03
Picon
Favicon

Re: evaluate Maxima string from lisp command line...

On Thu, 04 Mar 2004 20:51:32 -0800
Richard Fateman <fateman <at> cs.berkeley.edu> wrote:

> The source file would be nparse,  but I suspect that
> it doesn't have exactly that functionality to read from a string.
> It reads character by character from a stream, most usually.

So, how does what is entered on the command line get translated into Maxima internals?  nparse has a lot of
mread, scan-string, and parse type stuff, but it isn't immeditately clear what does what.  

> If it were in lisp you could do (read-from-string  " ....")
> 
> (setf expr (macsyma-parse-from-string  " ....."))  if that existed.
>
> Then to evaluate it, do
> 
> (meval expr)

Is there any documentation somewhere on how Maxima actually does this?  It would seem to be an appropriate
topic for some kind of programmer's manual.

CY
Permalink | Reply | 0 comments |
headers
Stavros Macrakis | 5 Mar 2004 07:05
Picon

RE: evaluate Maxima string from lisp command line...

> ...how to take a string in regular maxima input format,
> and feed it to Maxima on the lisp level

In other words, you want to parse a string as Macsyma language.  The
macsyma-read-string function is supposed to do this, but doesn't work in
5.9.0/GCL.  It also has no error handling.  I have rewritten it from
scratch.  Give it a whirl.

  (meval (macsyma-read-string "diff(x^2,x)"))
    => ((MTIMES SIMP) 2 |$x|)

;;; Copyright 2004 by Stavros Macrakis
;;; Licensed under your choice of the GPL or the modified BSD license

;;; Parse a string as a Maxima expression
;;; Return the parsed form
;;; If there's an error, it returns the symbol parse-error; the
;;; second return value is then the error string.

;;; Usage examples:
;;;
;;; (macsyma-read-string "1/+3*x") =>
;;;   ((MQUOTIENT) 1 ((MPLUS) ((MTIMES) 3 |$x|)))
;;;   NIL
;;;
;;; Note that this is simply parsed -- no simplification.
;;; The simplified form is 
;;; ((MTIMES SIMP) ((RAT SIMP) 1 3) ((MEXPT SIMP) |$x| -1))
;;;
;;; (macsyma-read-string "1.2.3") =>
(Continue reading)

Permalink | Reply | 6 comments |

Gmane