I have been getting the following in my results in some cases on CPLEX for a
linearized version of a quadratic optimization problem.
presolve, constraint flow_balance5[9]:
all variables eliminated, but upper bound = -1 < 0
Please let me know why this would happen in only some cases because this
model is supposed to give me the optimal results. It causes problems on
comparison with an equivalent C program model which gives me the lower
bounds.
Thanks!
--
View this message in context: http://old.nabble.com/All-variables-eliminated-tp26848246p26848246.html
Sent from the AMPL mailing list archive at Nabble.com.
For the instances where this happens, the model is infeasible. If you
are confident that the linearized version has to be feasible, then I
would look for an error in the model formulation.
/Paul
> --
>
> You received this message because you are subscribed to the Google Groups
> "AMPL Modeling Language" group.
> To post to this group, send email to am...@googlegroups.com.
> To unsubscribe from this group, send email to
> ampl+uns...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/ampl?hl=en.
>
>
>
>
--
View this message in context: http://old.nabble.com/All-variables-eliminated-tp26848246p26849832.html