Dear all,

I'm copying this to open-axiom, in the hope that somebody there is also
interested in looking at the algorithm that chooses the signatures.  Note that
I did already a very little research on that topic, see

http://groups.google.at/group/fricas-devel/msg/7133b4ee2b13055b


Now the main report:

I find the following behaviour very confusing:

(1) -> f(x,y)==x+y
                                                                   Type: Void
(2) -> draw(x +-> f(x,2), 0..1)
   Compiling function f with type (DoubleFloat,PositiveInteger) -> 
      Point DoubleFloat 
   Conversion failed in the compiled user function f .
 
   Cannot convert from type DoubleFloat to Point DoubleFloat for value
   2.0

Why is Axiom trying to compile f as function to Point DFLOAT?  Here is some
more info:

(2) -> )di op draw

There are 31 exposed functions called draw :
   [1] ((DoubleFloat -> DoubleFloat),Segment Float,List DrawOption) -> 
            TwoDimensionalViewport
            from TopLevelDrawFunctionsForCompiledFunctions
   [2] ((DoubleFloat -> DoubleFloat),Segment Float) -> 
            TwoDimensionalViewport
            from TopLevelDrawFunctionsForCompiledFunctions

[...]

   [7] ((DoubleFloat -> Point DoubleFloat),Segment Float,List 
            DrawOption) -> ThreeDimensionalViewport
            from TopLevelDrawFunctionsForCompiledFunctions
   [8] ((DoubleFloat -> Point DoubleFloat),Segment Float) -> 
            ThreeDimensionalViewport
            from TopLevelDrawFunctionsForCompiledFunctions
[...]

i.e., the interpreter seems to prefer signature [8] over [2]...  (this is also
confirmed by )se me bo on).

I guess that this is a similar problem as in #105 (watch out for 

"Axiom chooses the wrong signature (it should use
integrate$RationalFunctionDefiniteIntegration)"


Should I attach this to #105 or open a separate issue?

Martin


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
open-axiom-devel mailing list
open-axiom-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/open-axiom-devel

Reply via email to