Jakob Schelbert

Hi Danylo, You wrote earlier that "I scale and center the variables to [-1,1]". I was assuming that by doing this you change the coefficients in your constraints. My suggestion would then be to jus...

Hi Danylo, As you already mentioned, your problem has a very wide range of matrix coefficients which makes the it numerically challenging. See http://www.gurobi.com/documentation/8.1/refman/numeric...

Hi Aziz, See the documentation here: http://www.gurobi.com/documentation/8.1/refman/py_model_setobjectiven.html You do not have to specify the sense of the objective (you can set the overall sense ...

Hi Velibor, As far as I know, the R-API does not permit adding lazy constraints. This is also true for the Matlab-API - both are more "matrix based" interfaces compared to the other APIs.

Hi Christopher, Have you tried saving each model itself and later call model.write() on the relevant models, or is the (memory) overhead too big for this approach? Can you elaborate a bit more why ...

Hi Andreas, some things caught my eye: your model is, I would say, a rather big one. 2.5m variables, 3.5m constraints and 14m non-zeros is something. However the numbers are halved after presolvin...

Hi Johannes, you could try to use cvxpy (https://www.cvxpy.org/examples/basic/quadratic_program.html) as an intermediate interface to gurobi. From what I've seen it should offer a similar interface...

Hi Johannes, I think the main reason why python does not provide a raw interface is that the high-level API is MUCH better in terms of debugging and extending a model while the overhead is minimal....

Hi Taylor, could you state your question a bit clearer. I'm a bit confused what your real question is. Maybe this can help formulating a good question: https://stackoverflow.com/help/how-to-ask

Hi Ebru, I think you posted the log of the build process. Daniel meant the log file from the actual solving process. Something like this: x7101: 9222 rows, 7101 columns, 54642 nonzerosChanged value...