I'm trying to implement the big M approach to a simple problem but I'm struggling a lot to define the values of big M in the model's constraints.
Thing turn out that if I pick a small value as M, the constraints become invalid, while a big value M could help contains the problem but with other equational constraints generate numerical accuracy issues like the following in Gurobipy and finally IIS
x[i] <= 0value(x[i]) = 1.34252346e-07
So, Is there a way to test a suitable M value to constrain one's model?