On Wed, Jan 14, 2009 at 2:34 PM, Ali Baharev
<address@hidden> wrote:
> Do you think that makes sense?
If i were you i would try the proposal without binary variables. Maybe
it turns out that there is only one violated constraint.
> *) Each companion variable denotes whether the original debugging variables
> are bigger than zero.
Well, i do not know anything about your problem but i am afraid it
would result in a big M formulation.
If you have thousands of constraints, it would yield thousands of new
binary variables.
Please do not forget the fact that the conflicting constraints are ambiguous.
Please also keep in mind that we are re-inventing the wheel, i.e.
finding the minimal set of conflicting constraints is a well-known
problem.
Good luck,
Ali