Ben Black

Hi Jaromił,  Isn't this the same but with \(x_2\) instead of \(|x_2|\)? If so, didn't we decide that \(x_1\) needs to be non-negative? I have tried the following model \ LP format - for model brows...

Update - I changed the quadratic constraint to use \(|x_1|\) instead of \(|x_2|\) and now the model is giving me the following result: Solution count 2: 1958.62 1958.12 Optimal solution found (tole...

Hi Jaromił,  Firstly, your comment about the \(|x_2|\) variables is correct and that was an oversight on my part, sorry for that. After seeing your updated comment I think that it should have been ...

Hi Jaromił, The LP file is below. Hopefully this is the model file you meant? I used three additional sets of dummy variables x1, x2 and x2_abs for the conic constraints. Also, x here is the same a...