Existing options keys

[ocots]

Hi @tmigot @amontoison @abelsiqueira and others!

I am working on control-toolbox and I would handle options for the modelers and solvers quite nicely for the user. To do this, I need a way to get all the existing options keys we can pass when solving with Ipopt.

solver = NLPModelsIpopt.IpoptSolver(nlp)
NLPModelsIpopt.solve!(solver, nlp; options...)

Is there a way to get the possible keys? I am aware of this page but the list is long: Ipopt options.

Thanks!

[tmigot]

Same answer as JuliaSmoothOptimizers/NLPModelsKnitro.jl#138
The docstring should document this

NLPModelsIpopt.jl/src/NLPModelsIpopt.jl

Line 189 in ad1cf0f

function ipopt(nlp::AbstractNLPModel; kwargs...)

up to two things:

• zL and zU should actually be zL0 and zU0
• There is also the callback keyword

[ocots]

Ok thanks, I forgot the callback, this may be useful.

Another question, it is written here that we can pass kwargs to IpoptSolver but is it true?

I do not see any constructor like this:

NLPModelsIpopt.jl/src/NLPModelsIpopt.jl

Lines 54 to 83 in 602baf6

	"""
	IpoptSolver(nlp; kwargs...,)
	Returns an `IpoptSolver` structure to solve the problem `nlp` with `ipopt`.
	"""
	mutable struct IpoptSolver <: AbstractOptimizationSolver
	problem::IpoptProblem
	end
	function IpoptSolver(nlp::AbstractNLPModel)
	@assert nlp.meta.grad_available && (nlp.meta.ncon == 0 || nlp.meta.jac_available)
	eval_f, eval_g, eval_grad_f, eval_jac_g, eval_h = set_callbacks(nlp)
	problem = CreateIpoptProblem(
	nlp.meta.nvar,
	nlp.meta.lvar,
	nlp.meta.uvar,
	nlp.meta.ncon,
	nlp.meta.lcon,
	nlp.meta.ucon,
	nlp.meta.nnzj,
	nlp.meta.nnzh,
	eval_f,
	eval_g,
	eval_grad_f,
	eval_jac_g,
	eval_h,
	)
	return IpoptSolver(problem)
	end

[amontoison]

Good catch 👍
It is a typo in the docstring, the options are only passed with the function "ipopt". I will open a PR to fix it.

Update: fixed in #152