Why does the objective function value sometimes not match the best objective?

Within the Rose solver, the best objective and best bound values are computed during optimization. These values are sent to the user interface as they are what the solver uses to calculate both the absolute and relative optimality gaps.

After the Rose solver finishes running, it recalculates the objective function value based on the decision variables it found, but this time with higher precision. This recalculated value is returned as the final objective function value. As a result, you may notice a slight difference between the initial best objective value and this final objective value due to the increased precision.