On Wednesday, January 28, 2015 at 4:45:54 PM UTC-6, Simon King wrote: > > Hi Stefan, > > On 2015-01-28, Stefan <stefan...@gmail.com <javascript:>> wrote: > > The question, now, is why the coercion system is invoked at all: I'm > doing > > arithmetic within my own ring. The culprit, I guess, is the expression > > 1/self in element.pyx. > > > > Does a ring keep track of its own multiplicative identity element? > > It looks like when you do 1/self, then 1 is coerced into self.parent(), > which involves the coercion system. On the other hand, coercions from ZZ > to self.parent() are ubiquituous, hence, it is surprising that changing > 1/self to self.parent().one()/self helps. >
No coercion is still faster than a fast coercion. > > Note that replacing 1/self by ~self may help as well. But, as you said > in your other posting, the default implementation of __invert__ is not > always perfect. > The line "1/self" occurs INSIDE the __invert__ special method of element.pyx. Should I open a ticket to replace that 1 by self.parent().one() ? Also, why isn't there an _invert_ method just like _add_ and friends? --Stefan -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
