Groebner bases over the integers
- From: "philipp@xxxxxxxxxxxxxx" <philipp@xxxxxxxxxxxxxx>
- Date: 29 Apr 2006 09:07:46 -0700
Hi,
I'm currently working with computations in the ring Z[X1...Xn] of
integer polynomials over finitely many variables (as a method for
handling non-linear arithmetic in a theorem prover). The problems that
we are dealing with usually (or often) have the shape
f1(X1,...,Xn)=0, ..., fk(X1,...,Xn)=0 |- g(X1,...,Xn)=0
where f1, ..., fx, g are polynomials over the integers. I.e., we want
to prove that whenever the functions that are described through
polynomials f1, ..., fk disappear for integers X1,...,Xn, also the
function described by g disappears. From Hilbert's 10th problem it is
clear that this problem is not decidable, but for practical purposes we
found that proofs often can be found by computing a Groebner basis of
f1, ..., fk and reducing g. Therefore, I am using the version of the
Buchberger algorithm that is described in "A Critical-Pair Completion
Algorithm for Finitely Generated Ideals in Rings", Buchberger, 1983.
That problem that I now have is: I want to complete the ideal
(f1,...,fk) by adding all polynomials that can be obtained by
eliminating common integer coefficients. As an example, in the
situation
2*x-2*y=0 |- x-y=0
I would in some way like to eliminate the "2" on the left side.
Likewise (and a bit more difficult), I also would like to be able to
prove
2*x+a=0, 2*y+a=0 |- x-y=0
More precisely, given an ideal I within Z[X1...Xn], I want to construct
a Groebner basis of the ideal I', where I' is the smallest ideal
containing I with the property: if a polynomial k*f is in I' (k a
non-zero integer), then also f is in I'
Unfortunately, so far I have no clue how I can compute such a basis
effectively ... but I seems to be a rather simple problem, so I would
be very grateful if someone in this group could tell me how to solve it
(or maybe tell about literature).
Thanks, Philipp
.
- Follow-Ups:
- Re: Groebner bases over the integers
- From: Philipp Ruemmer
- Re: Groebner bases over the integers
- From: Roman Pearce
- Re: Groebner bases over the integers
- Prev by Date: Re: Reduce equation
- Next by Date: Re: Discrepancy Between 2 Different Ways Of Calculating - What's Wrong, Please?
- Previous by thread: Reduce equation
- Next by thread: Re: Groebner bases over the integers
- Index(es):
Relevant Pages
|