Nonconvex JuMP model to DictModel cannot see constraints

Specific Domains Optimization (Mathematical)
1

Hello,

Following the example from DictModel definition · Nonconvex.jl

using JuMP, Nonconvex
jump_model = JuMP.Model()
@variable jump_model 0 <= x[i=1:3] <= 1
@constraint jump_model sum(x) <= 1
model = DictModel(jump_model)

Should I be able to confirm the constraint using NonconvexCore.getnineqconstraints(model)? Because the output is zero.

image
image1089×595 18.3 KB

Thanks

2

This looks like a bug. The conversion from JuMP is tested NonconvexCore.jl/test/jump.jl at master · JuliaNonconvex/NonconvexCore.jl · GitHub. I will need to look into the above example.

3

Found the bug. For some reason, Nonconvex only supports named JuMP constraints! This is not intentional and I will try to fix it. But a workaround is:

@constraint jump_model c1 sum(x) <= 1