Odtnhj

I want to try using persistent solver for an algorithm that iteratively adds new constraints to the problem, and want to avoid having to completely rebuild the file given to the solver before each iterations.

Before using persistent solver as described on https://pyomo.readthedocs.io/en/stable/solvers/persistent_solvers.html, I used a ConstraintList object to iteratively add my new constraints without having to name them individually. I thought this was a very elegant solution and I want to see if there is a way to notify the persistent solver when a new constraint is added to the ConstraintList.

In the docs, it is writtent that

m.c2 = pe.Constraint(expr=m.y >= m.x) 
opt.add_constraint(m.c2) 

where m.c2is a constraint to be added to the model with persistent solver. What would be the equivalent line to notify the persistent solver that a change was done to the ConstraintList, once that a constraint was added in it?

Here is how you create your constraint list

m.Cut_Defn = pyomo.ConstraintList(noruleinit=True)

And then you can add constraints in your constraint list:

m.Cut_Defn.add(some_number >= your_variable + some_other_number)

If you solve before the .add() then you'll find another solution than solving after the .add(). So you can think like: it implements the new constraints on the fly, and you have to resolve your model, if you want that constraints to be in your optimization.