perhaps i should have posted to sage-devel instead ? i don't even know
how to "properly" report a bug.

On 10 sep, 18:53, Pierre <[EMAIL PROTECTED]> wrote:
> hi there,
>
> i thought i'd report this bug, even though it's hard to reproduce and
> not well-identified. I have a ring R of type:
>
> Fraction Field of Multivariate Polynomial Ring in x0, x1, x2, x3, x4,
> x0_iv, x1_iv, x2_iv, x3_iv, x4_iv over Cyclotomic Field of order 4 and
> degree 2
>
> (the _iv variables will play no role in the sequel -- not that it
> matters.)
>
> And I have a list L of 64 matrices in MatrixSpace(R, 2).
> When i try sum(L), I get:
>
> /home/pedro/Bureau/sage-3.0.2-ubuntu32-intel-i686-Linux/local/lib/
> python2.5/site-packages/sage/rings/fraction_field_element.py in
> _add_(self, right)
>     298                         numer = numer // new_gcd
>     299                         denom = denom // new_gcd
> --> 300                     return FractionFieldElement(self.parent(),
> numer, denom, coerce=False, reduce=False)
>     301                 # else: no reduction necessary
>     302             except AttributeError: # missing gcd or quo_rem,
> don't reduce
>
> /home/pedro/Bureau/sage-3.0.2-ubuntu32-intel-i686-Linux/local/lib/
> python2.5/site-packages/sage/rings/fraction_field_element.py in
> __init__(self, parent, numerator, denominator, coerce, reduce)
>      65                 pass
>      66         if self.__denominator.is_zero():
> ---> 67             raise ZeroDivisionError, "fraction field element
> division by zero"
>      68
>      69     def reduce(self):
>
> ZeroDivisionError: fraction field element division by zero
>
> (This is sage 3.1.1 despite the folder name.)
>
> If i try to do a for loop and sum the matrices one by one, i realize
> that the problem is with the addition of sum(L[:49]) (call it M) and
> L[49] (call it N). I can compute them separately and get the same
> error message by trying M+N.
>
> Of course you'd like to see M and N. Well be ready for a
> disappointment: M is
>
> [  ((37/32*I - 5/16)*x0^2*x1*x2*x3 + (-5/64*I + 7/32)*x1*x2^2*x3^2 +
> (-19/32*I + 7/16)*x0^2*x1*x2*x4 + (-9/32*I - 1/16)*x0^2*x1*x3*x4 +
> (-5/32*I + 13/16)*x0^2*x2*x3*x4 + (-19/8*I - 89/16)*x1^2*x2*x3*x4 +
> (15/32*I - 3/8)*x1*x2^2*x3*x4 + (31/64*I - 27/64)*x2^3*x3*x4 +
> (-31/32*I + 5/16)*x1*x2*x3^2*x4 + (21/32*I - 13/64)*x2*x3^3*x4 +
> (-5/16*I + 15/64)*x1*x2^2*x4^2 + (-3/32*I + 1/4)*x1*x2*x3*x4^2 +
> (-7/64*I + 3/64)*x1*x3^2*x4^2 + (13/64*I + 5/8)*x2*x3*x4^3)/
> (x1*x2*x3*x4)      ((5/32*I - 11/32)*x0^2*x1*x2*x3 + (1/8*I -
> 1/4)*x1*x2^2*x3^2 + (7/32*I + 15/32)*x0^2*x1*x2*x4 + (3/32*I -
> 1/32)*x0^2*x1*x3*x4 + (9/32*I - 3/32)*x0^2*x2*x3*x4 + (-29/32*I +
> 303/128)*x1^2*x2*x3*x4 + (5/16*I + 11/16)*x1*x2^2*x3*x4 + (-1/64*I +
> 29/128)*x2^3*x3*x4 + (11/8*I + 1/8)*x1*x2*x3^2*x4 + (21/64*I -
> 13/128)*x2*x3^3*x4 + (-17/64*I - 7/32)*x1*x2^2*x4^2 + (-7/8*I +
> 7/8)*x1*x2*x3*x4^2 + (-9/64*I - 7/64)*x1*x3^2*x4^2 + (5/16*I -
> 13/128)*x2*x3*x4^3)/(x1*x2*x3*x4)]
> [((-5/32*I + 11/32)*x0^2*x1*x2*x3 + (-1/8*I + 1/4)*x1*x2^2*x3^2 +
> (-7/32*I - 15/32)*x0^2*x1*x2*x4 + (-3/32*I + 1/32)*x0^2*x1*x3*x4 +
> (-9/32*I + 3/32)*x0^2*x2*x3*x4 + (-267/32*I + 161/128)*x1^2*x2*x3*x4 +
> (-5/16*I + 23/16)*x1*x2^2*x3*x4 + (1/64*I - 29/128)*x2^3*x3*x4 +
> (-1/8*I + 3/2)*x1*x2*x3^2*x4 + (-21/64*I + 13/128)*x2*x3^3*x4 +
> (17/64*I + 7/32)*x1*x2^2*x4^2 + (7/8*I + 7/4)*x1*x2*x3*x4^2 + (9/64*I
> + 7/64)*x1*x3^2*x4^2 + (-5/16*I + 13/128)*x2*x3*x4^3)/(x1*x2*x3*x4)
> ((47/32*I - 7/8)*x0^2*x1*x2*x3 + (5/64*I + 15/32)*x1*x2^2*x3^2 +
> (47/32*I + 3/4)*x0^2*x1*x2*x4 + (-15/32*I + 1/2)*x0^2*x1*x3*x4 +
> (-27/32*I + 1/2)*x0^2*x2*x3*x4 + (41/8*I + 49/16)*x1^2*x2*x3*x4 +
> (-13/32*I + 1/16)*x1*x2^2*x3*x4 + (-27/64*I - 31/64)*x2^3*x3*x4 +
> (33/32*I + 13/8)*x1*x2*x3^2*x4 + (21/32*I - 13/64)*x2*x3^3*x4 + (1/8*I
> - 29/64)*x1*x2^2*x4^2 + (-83/32*I - 15/16)*x1*x2*x3*x4^2 + (-1/64*I -
> 25/64)*x1*x3^2*x4^2 + (-13/64*I - 5/8)*x2*x3*x4^3)/(x1*x2*x3*x4)]
>
> while N is
>
> [                                                                             
>                                                                               
>                                                                               
>                                                                               
>                                                                               
>                                                                             0
> ((-5/32*I + 11/32)*x0^2*x1*x2*x3 + (-1/8*I + 1/4)*x1*x2^2*x3^2 +
> (-7/32*I - 15/32)*x0^2*x1*x2*x4 + (-3/32*I + 1/32)*x0^2*x1*x3*x4 +
> (-9/32*I + 3/32)*x0^2*x2*x3*x4 + (23/32*I + 129/128)*x1^2*x2*x3*x4 +
> (1/2*I - 3/16)*x1*x2^2*x3*x4 + (1/64*I - 29/128)*x2^3*x3*x4 + (-3/16*I
> + 3/8)*x1*x2*x3^2*x4 + (-21/64*I + 13/128)*x2*x3^3*x4 + (17/64*I +
> 7/32)*x1*x2^2*x4^2 + (1/16*I + 1/4)*x1*x2*x3*x4^2 + (9/64*I +
> 7/64)*x1*x3^2*x4^2 + (-5/16*I + 13/128)*x2*x3*x4^3)/(x1*x2*x3*x4)]
> [  ((5/32*I - 11/32)*x0^2*x1*x2*x3 + (1/8*I - 1/4)*x1*x2^2*x3^2 +
> (7/32*I + 15/32)*x0^2*x1*x2*x4 + (3/32*I - 1/32)*x0^2*x1*x3*x4 +
> (9/32*I - 3/32)*x0^2*x2*x3*x4 + (-23/32*I - 129/128)*x1^2*x2*x3*x4 +
> (-1/2*I + 3/16)*x1*x2^2*x3*x4 + (-1/64*I + 29/128)*x2^3*x3*x4 +
> (3/16*I - 3/8)*x1*x2*x3^2*x4 + (21/64*I - 13/128)*x2*x3^3*x4 +
> (-17/64*I - 7/32)*x1*x2^2*x4^2 + (-1/16*I - 1/4)*x1*x2*x3*x4^2 +
> (-9/64*I - 7/64)*x1*x3^2*x4^2 + (5/16*I - 13/128)*x2*x3*x4^3)/
> (x1*x2*x3*x4)
>
> 0]
>
> (I can provide the dumpstrings if anyone is interested in trying)
> (oh and here I is sqrt(-1))
>
> HOWEVER, it is in fact possible to do sum(L[30:]) + sum(L[:30]) !!
> don't know whether the answer is meaningful though. In fact, other
> experiments with these matrices lead me to believe that sometimes
> sum() just gets the wrong answer (things that were supposed to commute
> didn't commute; but they were sums of things which did commute).
>
> Thoughts anyone ?
>
> Pierre
--~--~---------~--~----~------------~-------~--~----~
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-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Reply via email to