the gap in the solution of multi-period mixed integer linear programming has been stuck at around 1.7%. Is this error considered large or small,

Answered

mengyu chen

April 20, 2026 02:18

• 

I kindly request someone to interpret this: the gap in the solution of multi-period mixed integer linear programming has been stuck at around 1.7%. Is this error considered large or small,

Related to

• Performance
• MIP Gap

1

• 

  Byron Tasseff

  • Gurobi Staff

  April 20, 2026 19:22

  Hi Mengyu,

  The reported gap is not an error. It is the MIP gap, which measures how far the current best feasible solution could still be from the true optimum, based on the best objective bound available so far.

  Whether 1.7% is "large" or "small" depends on the application. In some settings, a 1-2% gap may already be acceptable. In others, even a much smaller gap may matter. It can also be useful to consider the absolute objective difference rather than just the percentage difference.

  If the gap appears to be “stuck,” that usually means progress has slowed in improving the incumbent solution, the best bound, or both. To understand what is happening, it would be helpful to look at the solver log and, if possible, learn more about the model.

  In summary, a gap of 1.7% is not inherently good or bad by itself. It has to be judged in the context of the model and the decision-making requirements.

  If helpful, feel free to share more details about the model or application. For additional background, you may also find this article useful: "What is the MIPGap?"

  1

Please sign in to leave a comment.