Understanding the impact of a modification to a model
Answered
If I have two MIP models that have the same objective function and highly similar sets of constraints and variables, is it possible to measure the impact of those differences that controls for ordering differences between the models?
Lets say I have model A. I construct another model B which is the exact same except for 1 additional constraint. They both solve to an optimal solution but with different objective values. I would like to like to understand the impact the additional constraint made to the solution in comparison to the solution from model A that is not attributable to a new ordering. Because the order of constraints is different, I'd like to be able to control for that and confidently say the difference in the objective value is entirely due to the existence of the additional constraint (or not).
Similarly, lets say I have the same scenario as above but both models solve to the same objective value. However, some set of variable values are different. How would go about proving those differences are due to the non-determinism of a different constraint ordering and not due to the additional constraint changing the problem?
Last one. Lets say I have the same scenario as above. Both models solve to the same objective value. However, 2 variable values are different. 1 variable was impacted because of ordering and the other is due to the additional constraint. Can I somehow these variable differences correctly?
-
-
Gurobi Staff
Hi Matt,
Thank you for contacting us! Since you have a commercial license with us, I converted your question into a ticket in our Help Center and we would get back to you shortly via the ticket.
Best regards,
Maliheh
0 -
Please sign in to leave a comment.
Comments
1 comment