Mario Ruthmair

Hi Kai, Not all models might profit from using Benders decomposition. When applying Benders decomposition, you project out parts of the model and dynamically re-add this information in terms of fea...

Hi Lelisa,Gurobi can detect clique cuts based on the existing variables and constraints. Here is a very simple example: If you have binary variables x1, x2, x3, and constraintsx1 + x2 <= 1 x1 + x3 ...

Hi Tobia,In most cases, the quadratic objective terms in an MIQP are linearized to convert it to an MILP. This can weaken the bound obtained from the continuous relaxation. This is probably what yo...

See this article: What is the difference between user cuts and lazy constraints? Is there a specific reason why you are asking if there is another way to add lazy constraints? Are you missing somet...

Hi Florian, Yes, since many CVRP cut families cut off integer solutions, too, I would strongly recommend adding all your cuts with addLazy(). There is no downside to this; in fact, you might even b...

Hi Florian, You mentioned that you add the RCCs in the MIPNODE callback with addCut(). What happens if you add them in MIPNODE with addLazy()? Can you still reproduce the issue?This was my workarou...

Hi Giovanni, Gurobi provides a general-purpose solver, while Hexaly provides specific interfaces to particular problem types, e.g., routing problems, which allows the use of problem-specific soluti...

Hi Krypt, Sometimes, people add binary arc variables x(a,b) that are actually not needed in a route formulation. But you can do some stuff with them, e.g., branching, additional cuts, etc.Of course...

Hi David, Yes, indeed, the WLS academic license requires a permanent internet connection since it requests a license token from our servers every 5 minutes. Best regards,Mario

Hi David, Your named-user license, unfortunately, does not work with a token server.For a WSL2 setup, we recommend the so-called "WLS Academic" license (WLS = Web License Service). You can generate...