Adrian Dusa

To answer my own question, it was a confusion: R was solving without the second objective, because of a forgotten variable. And that of course explains the time difference.Apologies for the false a...

Thank you Martin, this is very comprehensive.While the separate two optimizations are clear, I am now struggling with the model specification of a pure blended multi-objective, now on a much larger...

Hi Martin, hi Riley,Just a quick follow-up feedback, having now a working code. The timing of a multi-objective function appears to be about half of the above two-step optimizations.By setting an e...

Thank you Riley, problem solved.Apart from failing to use menv = GRBgetenv(model), I was indeed still using the multi-objective API (GRBsetobjectiven) and that, as you pointed out, makes PoolSearch...

Hi Riley,This amounts to solving the problem twice, which is bad news for time consuming problems, but I guess there is nothing to do about it.Alright, I tried to implement your procedure, but for ...

Dear Riley,I have progressed a lot with my code, and now reached a new challenge that, despite my weeks long testing effort, the solution still remains stubbornly hidden. Below you can find my late...

Dear Riley, You just made a very happy person. This is beautiful, exactly what I was looking for. It did not occur to me that using the negative of the value was the culprit, I tried something simi...

Hello Riley, Let me again bring my deepest gratitude with your patience and your help. I tried to make sense of both examples in Gurobi, but still to no avail. In the diet example, the cost (which ...

If you create an objective where the more valuable variables have a greater objective coefficient then those solutions which favor those variables will be preferred.  Do you think this will be eno...

Hi Riley, Yes that is my understanding as well, but the matrix generated by the function spMatrix() is not a coefficient matrix, it is something else. What you describe really is a matrix with rows...