On Tuesday, September 2, 2014 11:17:35 PM UTC-7, Joachim Durchholz wrote: > > Am 03.09.2014 um 02:07 schrieb Richard Fateman: > > Sure. Unlikely to be easy to do by simply hacking on trees. Here's a > > classic pattern: > > > > a*x^2+b*x+c. a,b,c are pattern variables. x, in this context, > is a > > symbolic constant. > > You might like to also impose a,b,c free of x, and a is non-zero, to > > complete the pattern "quadratic in x" > > > > match this against the expression > > > > (45*x+10)^2/5 +q*x+pi > > > > Yes this can be done, but can you do it by placing attributes on + and > *? > > Easily, actually. > Just store the entire pattern, including preconditions, in the attribute > of the top-level node (+ in this case). > This sounds like "assume that we have a solution" . Theorem. "We have a solution". If you store the entire pattern, then you are hypothesizing that you have a program that matches quadratics.
> Actually Python makes this relatively easy since reifying functions is > fussless and reifying expressions as lambdas is lightweight (can't beat > Haskell or Lisp in that area, of course). > > This would probably also make it easier to check for pattern overlap. > I don't understand why you say this. > > > I suppose there is also the question of whether you want this to match > or > > not. > > That's a strategy question: Which of multiple applicable rules do we > want to apply? > If you are writing a program that does something with quadratics, e.g. integrating 1/quadratic_in_x, then how do you want to match quadratics? > RUBI requires that at most one rule be applicable at a time (i.e. no > pattern overlap), so strategy is encoded in the preconditions of the > rules and it's a non-issue. > You seem to be asserting that this requirement is trivially satisfied. Are you sure? The other requirement that I mentioned, related to termination, is distinctly not trivially satisfied unless you can prove things about simplification. -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to sympy+unsubscr...@googlegroups.com. To post to this group, send email to sympy@googlegroups.com. Visit this group at http://groups.google.com/group/sympy. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/3f495ba5-20f5-4825-a978-f8961ed7c1f0%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
