Problems with linsolve: quotient by polynomial of higher degree

Can anybody please try to debug the computation of linsolve for my example
to see where the problem is generated? (Maybe it is in gcd, as Richard
Fateman suggested). I am having some difficulties in the use of the
debugger, it is not simple for me...

Stefano



2012/5/26 Stefano Ferri <ferriste at gmail.com>

>
> Q - Is it possible to use linsolve as a general linear equations systems
>>> solvers,
>>>
>> If the system is indeed linear, it should work if the demands of memory
>> and time do not prevent it from finishing.
>>
>
>
> Of course, the system is linear. It is in the file attached in my first
> message. All systems generating this problem are linear. It doesn't seem a
> problem of memory or time, since the error comes after one or two seconds.
>
>
>
>> pquotient is apparently free of bugs. It appears that some subtle bug in
>> the use of GCD is the cause.  If Maxima
>> computes  G=polynomialGCD(A,B),  then the degree of G must be less than
>> the degrees of A or B.
>> But it isn't.
>>
>> It may be that there is some conflict when terms with sqrt are simplified
>> and some processing decreases
>> the degree of A  after the GCD.
>>
>> Finding this bug would be nice, if indeed the system is linear etc.
>>
>> In some cases a work-around is to use a different GCD algorithm, e.g.
>> setting gcd:red
>>
>
>
> Setting gcd:red here has the only effect of slowing down the computation,
> it gives the same error. (Seems like there are many values for gcd, but are
> they undocumented?) Setting gcd to ez gives an error:
>
> (%i6) gcd:ez;
> (%o6)                                 ez
> (%i7) linsolve(_eq_list,_unknowns_list);
>
> Maxima encountered a Lisp error:
>
>  Error in PROGN [or a callee]: Caught fatal error [memory may be damaged]
>
> Automatically continuing.
> To enable the Lisp debugger set *debugger-hook* to nil.
>
> Unfortunately I'm not running the latest version of Maxima, now I'm on
> Ubuntu 11.10.
> (%i8) build_info();
>
> Maxima version: 5.24.0
> Maxima build date: 10:27 5/17/2011
> Host type: i686-pc-linux-gnu
> Lisp implementation type: GNU Common Lisp (GCL)
> Lisp implementation version: GCL 2.6.7
>
>
> I've forgotten to say that I have set algebraic to true, but even setting
> it to false doesn't change nothing.
>
> Could this bea related issue?
> http://www.math.utexas.edu/pipermail/maxima/2003/003934.html
>
>
> Stefano
>
>