Konrad Trifunovic wrote:
that's a good point. We DO NOT need NNC polyhedra. Closed Polyhedra are fine (at least for dependence analysis). Do You think that using C_Polyhedron would help us avoid the much of the complexity problems?
C_Polyhedron operations are significantly more efficient than the operations of NNC_Polyhedron.
from my analysis (by looking at the PPL code) I have seen that the "contains_integer_point" methods call the MIP (Mixed Integer Programming) routines, to solve the system of constraints. I have seen that it behaves differently in the case of Convex and NNC polyhedra.
Yes. I don't know how you use the "contains integer point" methods... of course solving lots of mixed integer programming problems has dramatic effects on the complexity.
In any case, concerning the complexity problems you mention above, we are very willing to work with you on them, but we need a way to reproduce them. All the best,
Roberto