[Date Index] [Thread Index] [Author Index]

Inequalities

• To: mathgroup at yoda.physics.unc.edu
• Subject: Inequalities
• From: mstankus at oba.ucsd.edu (Mark Stankus)
• Date: Fri, 16 Oct 92 14:18:52 PDT

In relation to a previous posting about geq0:

The problem is that 2 x^2 is not of the form x_^2.

A good start may be:   

mySgn[c_?NumberQ] := Sgn[c];
mySgn[x_^n_?(EvenQ[#]&)] := 1;
mySgn[c_ d_] := mySgn[c] mySgn[d];
mySgn[c_ + d_] := mySgn[c] /; mySgn[c] == mySgn[d]

geq0[expr_] := If[mySgn[x] == 1, True,False,unknown];

I have written some code which takes a list of 
recorded facts (such as x>=1 and y <= 2) and
if you ask if ask if x>= 2 it says unknown and if you 
ask x>= 0 , it says True. 

This code is beta-ish, but it works for my application
(a symbolic variant of Mora's algorithm -- noncommutative
Groebner basis).

Send me a note if you want a copy of my beta code.

Mark Stankus