On Monday 18 August 2008, Hamish Ivey-Law wrote: > Hello, > > I found a bug that occurs when calling random_element() on a > polynomial or power series ring over a Givaro finite field (the Givaro > finite fields are used when the field is non-prime and has cardinality > < 2^16). The problem is that the polynomial ring assumes that its > base ring's random_element() method takes arguments and it thus > calls the base ring method like this: R.random_element(*args, > **kwds). But the Givaro FF implementation of random_element() only > takes a 'self' argument, and thus raises a TypeError if args is non- > empty. For example: > > --BEGIN-- > sage: P.<x> = PowerSeriesRing(GF(3^3, 'a')) > sage: P.random_element(7) > --------------------------------------------------------------------------- > TypeError Traceback (most recent call > last) > > /Users/hlaw/sage/devel/<ipython console> in <module>() > > /Users/hlaw/sage/local/lib/python2.5/site-packages/sage/rings/ > power_series_ring.py in random_element(self, prec, bound) > 544 1/15 + 19/17*t + 10/3*t^2 + 5/2*t^3 + 1/2*t^4 + > O(t^5) > 545 """ > --> 546 return self(self.__poly_ring.random_element(prec, > bound), prec) > 547 > 548 def __cmp__(self, other): > > /Users/hlaw/sage/local/lib/python2.5/site-packages/sage/rings/ > polynomial/polynomial_ring.py in random_element(self, degree, *args, > **kwds) > 770 """ > 771 R = self.base_ring() > --> 772 return self([R.random_element(*args, **kwds) for _ in > xrange(degree+1)]) > 773 > 774 def _monics_degree( self, of_degree ): > > TypeError: random_element() takes no arguments (1 given) > --END-- > > In this case, the problem occurs because the > PowerSeriesRing_generic.random_element() passes bound to > PolynomialRing_general.random_element() where it is placed in the > args parameter and subsequently given to > FiniteField_givaro.random_element() which is not expecting any > arguments. > > I fixed this problem by adding *args and **kwds parameters to > FiniteField_givaro.random_element(): > > diff -r 717c10d9cd4a sage/rings/finite_field_givaro.pyx > --- a/sage/rings/finite_field_givaro.pyx Fri Jul 11 11:46:02 2008 > -0700 > +++ b/sage/rings/finite_field_givaro.pyx Mon Aug 18 16:10:50 2008 > +0200 > @@ -358,7 +358,7 @@ cdef class FiniteField_givaro(FiniteFiel > else: > return True > > - def random_element(FiniteField_givaro self): > + def random_element(FiniteField_givaro self, *args, **kwds): > """ > Return a random element of self. > > Depending on your policy for the random_element() interface, this > may or may not be the best way to fix the problem. The diff above was > against Sage 3.0.5, though I checked that this issue is still present > in 3.1.1. > > Regards, > Hamish.
Hi there, I agree with your fix. However the problem is bigger than what you reported. We don't have a unified interface for random_element: Some methods expect parameters and some don't: algebras/quaternion_algebra.py: def random_element(self): algebras/quaternion_order.py: def random_element(self): coding/linear_code.py: def random_element(self): combinat/cartesian_product.py: def random_element(self): combinat/choose_nk.py: def random_element(self): combinat/multichoose_nk.py: def random_element(self): combinat/permutation_nk.py: def random_element(self): combinat/permutation.py: def random_element(self): combinat/permutation.py: def random_element(self): combinat/permutation.py: def random_element(self): combinat/permutation.py: def random_element(self): combinat/split_nk.py: def random_element(self): combinat/subset.py: def random_element(self): combinat/subset.py: def random_element(self): combinat/tableau.py: def random_element(self): matrix/matrix_space.py: def random_element(self, density=1, *args, **kwds): modular/dirichlet.py: def random_element(self): modules/free_module.py: def random_element(self, prob=1.0, **kwds): rings/contfrac.py: def random_element(self, num_bound=2, den_bound=2): rings/extended_integer_ring.py: def random_element(self, x=None, y=None): rings/integer_mod_ring.py: def random_element(self, bound=None): .... So I'd suggest: - to open a trac ticket for the particular issue you have (if you don't have a Trac account yet, I can do that for you) - we open another Trac ticket with the task to 'fix' all the random_element methods by adding *args and **kwds. Thoughts? Martin -- name: Martin Albrecht _pgp: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x8EF0DC99 _www: http://www.informatik.uni-bremen.de/~malb _jab: [EMAIL PROTECTED] --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
