Thanks for the quick response Oscar.
> It is assumed that the coefficients are integers.
All coeff's and b's are integer in my model.
> But, from your explanation it seems like all the coefficients are zero,
> which is strange. Especially since you get an incorrect result when
> removing the empty constraints. (Clearly it would be good to check if
> there are any non-zero coefficients.)
Not too sure by what you mean by "check if there are any non-zero coeff". If
there were non-zero coeffs then we'd be talking about an error in
glp_get_mat_row(), right?
> One possible reason is that your model generates some dummy constraints
> that breaks the model and that you also have fractional coefficients that
> are not well handled by the export function leading to an incorrect
> solution.
Everything is integer in the model. I could see in the export function
(lpx_write_pb()) where the coeff values are cast to ints, is this what you mean
by 'not well handled'? I can understand that, but the fact that I never get to
that fprintf statement at all because k==0 seems like the issue here. Any
other thoughts about what I should inspect?
Oh, and thanks for writing this module in the first place, it should turn out
to be handy for me,
Joey
_________________________________________________________________
Connect and share in new ways with Windows Live.
http://www.windowslive.com/share.html?ocid=TXT_TAGHM_Wave2_sharelife_012008
_______________________________________________
Help-glpk mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/help-glpk
