Re: recognition puzzle
- To: mathgroup at christensen.cybernetics.net
- Subject: [mg226] Re: [mg193] recognition puzzle
- From: sfeir at mines.u-nancy.fr
- Date: Wed, 23 Nov 94 12:11:02 +0100
>Dear Mma puzzle fans:
>
>I have a list of several expressions; for example,
>
>{x^2 + 2 I x y - y^2 , 4 x^2 - 8 I x y - 4 y^2 }
>
>Mma has the list {x,y} of primitive variables. I want it to recognize that
>these expressions are linear in only TWO cooked-up variables (namely,
>x^2-y^2 and x*y), and I want it to return the list {x^2-y^2, x*y} to me.
>
>Is this an impossible pattern recognition task? Or is there a way to make
>Mma recognize the combination x^2-y^2 by its own "thought" processes?
>
>Regards- Martin McClain, Chemistry, Wayne State, Deetroit
The package PatternRearrange in the Mathematica Journal 4.3 by Harris
addresses exactly this question. I beleive the package is also on
MathSouyrce
_____________________________/____________________________________
A.A. Sfeir e-mail: sfeir at mines.u-nancy.fr
Ecole des Mines Tel : (33) 83 58 40 15
Parc de Saurupt (33) 83 37 64 41
54042 Nancy Cedex FRANCE Fax: (33) 83 57 97 94
_________________________________\________________________________