[Dev] Issue with the objective value for the init

[ocots]

When we get the init after max_iter=0, the value of the objective for optimal control problems with Lagrange cost is the default value (0.1), not the true objective value. This comes from the fact the Lagrange part is replaced in CTDirect by a Mayer cost with an initial guess fixed to 0.1.

See: https://github.com/control-toolbox/OptimalControlProblems.jl/actions/runs/17193430974/job/48772177053#step:4:1441

@PierreMartinon

[ocots]

@PierreMartinon When I set lagrange_to_mayer=false, then I have the following message:

Warning: cannot recompute objective (lagrange to mayer disabled)

and in that case, instead of having 0.1 in the objective, I have 0.0.

[ocots]

Actually, it comes from the fact that I use the second method to build an OCP solution. I try to change that.

[ocots]

Ok it works if I use the first build_OCP_solution.

[PierreMartinon]

Hi @ocots,

Minor update: in 0.17 lagrange_to_mayer will be false by default and I fixed the objective recomputation, so it should work with the usual build method.
Re-unifying the 2 build methods with new getters for the ADNLP case is scheduled next.