Using `UnoSolver` with `MOA`

[Sush1090]

Hello,

I am trying to use MultiObjectiveAlgorithms.jl for my system. For single objective problem in my system I use UnoSolver.jl, which has given me great results.

I wanted to know if I can use it in MOA as well, as in the readme file it says:

Replace HiGHS.Optimizer with an optimizer capable of solving a single-objective instance of your optimization problem.

But when I try it on the example Bi-objective linear problem my code fails. Here is my code:

using JuMP,UnoSolver
import HiGHS
import MultiObjectiveAlgorithms as MOA

model = Model()
set_silent(model)
@variable(model, x1 >= 0)
@variable(model, 0 <= x2 <= 3)
@objective(model, Min, [3x1 + x2, -x1 - 2x2])
@constraint(model, 3x1 - x2 <= 6) 
set_optimizer(model, () -> UnoSolver.Optimizer(preset="filtersqp")) # Only changed the Optimizer from HiGHS
set_attribute(model, MOA.Algorithm(), MOA.Lexicographic())
optimize!(model)
solution_summary(model)

So is there a problem in my usage or is the UnoSolver not compatible with MOA as of now?

Thank you

[kofgokhan]

You should wrap the optimizer with MOA.Optimizer and pass the parameter preset as follows:

set_optimizer(model, () -> MOA.Optimizer(UnoSolver.Optimizer))
set_optimizer_attribute(model, "preset", "filtersqp")

Can you give this a try?

using JuMP,UnoSolver
import MultiObjectiveAlgorithms as MOA

model = Model()
set_silent(model)
@variable(model, x1 >= 0)
@variable(model, 0 <= x2 <= 3)
@objective(model, Min, [3x1 + x2, -x1 - 2x2])
@constraint(model, 3x1 - x2 <= 6) 
set_optimizer(model, () -> MOA.Optimizer(UnoSolver.Optimizer))
set_optimizer_attribute(model, "preset", "filtersqp")
set_attribute(model, MOA.Algorithm(), MOA.Lexicographic())
optimize!(model)
solution_summary(model)

[Sush1090]

This worked.

Thank you

[Sush1090]

So in other words I cannot set the preset the way I set it when I only use UnoSolver? I need it pass it through set_optimizer_attribute always.

As well as pass optimizer as MOA