Hi all,
We recently ran into two cases of confusing behaviour of Factorization objects. The first: sage: f = factor(-1); f -1 sage: f^2 -1 I traced this to a special case in the powering code for taking powers of zero, which f is treated as: sage: bool(f) False This is because f.__len__() == 0, and f.__nonzero__ is not defined. What do you think the behaviour of __nonzero__() should be? Always return True? Return 'not self.value()' ? Return 'len(self.__x) > 0 or bool(self.__unit)' ? It's not at all clear to me how this should behave. My main confusion stems from the fact that it's easy to manually create Factorization objects that have value 0, even though some methods in Factorization don't seem to work properly for such manually created (non-prime) factorizations. E.g., Factorization([0,1]) and Factorization([(Integers(4)(2),2)]) have value 0, and: sage: Factorization([(6,1)]).gcd(factor(2)) 1 is unexpected, though understandable. One solution to this might be documenting the exact behaviour of the Factorization object and functions like gcd/lcm to make it clear what (not) to expect. In that case, making __nonzero__() always return True might be acceptable. (Although I'm not sure if that's acceptable for Factorizations of spaces.) The other is a non-commutative +: sage: f = factor(2) sage: f + 7 9 sage: 7 + f ... TypeError: unsupported operand parent(s) for '+': 'Integer Ring' and '<class 'sage.structure.factorization.Factorization'>' Maybe add/sub should be removed entirely as John Cremona suggested in http://trac.sagemath.org/sage_trac/ticket/3927 ? Alternatively, could Factorization objects somehow be hooked into the coercion system through the natural map to their universe where appropriate? Or does them not having a parent make this impossible/hard/undesirable ? -Willem Jan P.S. See also http://groups.google.com/group/sage-devel/browse_thread/thread/425a8614d5765130 for a previous discussion on this topic. --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
