About JuMP.is_solved_and_feasible

Specific Domains Optimization (Mathematical)
9

I find that JuMP already has part of this nice warning Error message integrated in quering functions, just as the following.

import JuMP, Ipopt
model = JuMP.Model(Ipopt.Optimizer)
JuMP.@variable(model, x >= 1)
JuMP.@objective(model, Min, x)
JuMP.optimize!(model)
JuMP.objective_bound(model) # The following ERROR is approving 👍
# ERROR: MathOptInterface.GetAttributeNotAllowed

Which embodies the idea mentioned by my post above

function improved_value(model) # in place of the existing `JuMP.value`
    if normality
        return JuMP.value(model) # returns the existing API 
    else
        error("Sorry, this info is unavailable or invalid!")
    end
end

In view of this, the JuMP.objective_bound(CR) # 1.0e100 is weird here (Gurobi12 reports obj_value < obj_bound in a Min-Program?), which doesn’t make any sense.
The expected behavior is also an ERROR message. Do you agree? @odow