Lower Bound in MIP equals zero, not root relaxation

[Paula]

I am using Gurobi with callback with a minimization problem.  In MIPNODE, I print the upper bound and lower bound.  I would expect the lower bound to equal the root relaxation, but instead it starts at 0 and is zero for a long time before it takes on a non-negative value.  The upper bound has a value fairly close to the solution from even the first time the MIPNODE callback is entered.  Is there any reason the lower bound isn't equal to the root relaxation?  I know the root relaxation solution is not zero.

[Renan Garcia]

Paula,

Can you please provide a code snippet including only the relevant portions of your callback? Also, a copy of the logs would help us diagnose this issue?

Thanks,

Renan

On May 19, 2015, at 12:36 PM, Paula <pli...@gmail.com> wrote:

I am using Gurobi with callback with a minimization problem.  In MIPNODE, I print the upper bound and lower bound.  I would expect the lower bound to equal the root relaxation, but instead it starts at 0 and is zero for a long time before it takes on a non-negative value.  The upper bound has a value fairly close to the solution from even the first time the MIPNODE callback is entered.  Is there any reason the lower bound isn't equal to the root relaxation?  I know the root relaxation solution is not zero.

--

---
You received this message because you are subscribed to the Google Groups "Gurobi Optimization" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gurobi+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.