Skip to content

Conversation

@tmigot
Copy link
Member

@tmigot tmigot commented Apr 26, 2024

Close #127 (but need to open a new issue afterwards to use the appropriate functions from Ampl API)

Explanations: In NLPModels 0.18/0.19 we added the linear API, i.e. instead of calling cons! you could use cons_lin! and cons_nln to access the linear or nonlinear constraints.
Indices of the (non)linear constraints are available in the meta via nlp.meta.lin of size nlp.meta.nlin and nlp.meta.nln of size nlp.meta.nnln.
The same applies to the Jacobian and its product.

@amontoison
Copy link
Member

@tmigot Is it ready to be merged if I rebase the branch?
I think it is what we need for #146.

@tmigot tmigot marked this pull request as draft October 4, 2024 21:37
@tmigot
Copy link
Member Author

tmigot commented Oct 4, 2024

I think there should be some Ampl way to make this work. Let me update to more recent versions of NLPModels by skipping some tests. If somebody that knows a bit more than I do Ampl that would be awesome!

@amontoison
Copy link
Member

@tmigot May I ask you to just quickly clean you PR such that I can merge it?
We can't use NLPModelsKnitro.jl with an AMPL model because of the lack of linear API.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Add linear API

2 participants