Le jeudi 5 janvier 2017 10:35:27 UTC+1, Dima Pasechnik a écrit : > > > Thanks for your suggestion; however, I am not sure if this could fully >> work: some computations require to take derivatives, i.e. to evaluate >> d/dx (sqrt(-z)), where z is the rational function of (x,y) discussed >> above. Could this work in our framework? >> > > d/dx (-z)^{1/2}=-1/(2(-z)^{1/2})dz/dx=-(dz/dx)/(2w), and so you can carry > one staying within rational functions in your variables original variables > and w. > > Yes indeed, but then what about dealing with a second volume element sqrt(-z') (arising from another metric): we cannot work both in the polynomial polynomial ring modulo the ideal generated by w^2+z and that modulo the ideal generated by w'^2 + z', can we?
However, let me take the opportunity of this discussion to mention that all the manifold stuff, as implemented in src/sage/manifolds, depends very weakly on the Symbolic Ring: all the coordinate calculus is encapsulated in the abstract base class CoordFunction <http://doc.sagemath.org/html/en/reference/manifolds/sage/manifolds/coord_func.html#sage.manifolds.coord_func.CoordFunction>, which, at the moment, has a single concrete derived class: CoordFunctionSymb <http://doc.sagemath.org/html/en/reference/manifolds/sage/manifolds/coord_func_symb.html#sage.manifolds.coord_func_symb.CoordFunctionSymb>. Only the latter involves the Symbolic Ring. In the future, we may conceive other concrete derived classes, based on fraction fields of polynomial rings, as Samuel and you suggest, or sympy, or Giac, or even numerical methods. Best wishes, Eric. -- 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 https://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.