objval different exponential

[Dinna]

Hi,

The objval from gurobi optimization does not match with manual calculation. For example: in my manual calculation, the result should be 6.3500e+03, while objval from gurobi is 6.3500e+08.

The difference is always +5 for the exponential part. 

I don't know which part I should investigate for this problem. I use matlab interface.

Regards,

Dinna

[Tobias Achterberg]

This should not happen; something must be wrong. Or may it be that you have
really large objective coefficients like 1e+13, so that, for example, a variable
value of 1e-5 could lead to an objective difference in the order of 1e+8?

Can you post some log output of Gurobi where the model statistics are shown?

Regards,

Tobias