search

JuliaSmoothOptimizers / OptimizationProblems.jl

Optimization Problems for Julia
Other
88 stars 48 forks source link

Update JuMP syntax to use the new nonlinear interface #321

Closed frapac closed 2 months ago

frapac commented 8 months ago

I had to comment out the tests checking that the formulation in JuMP matches those in ADNLPModels.

Apparently, the problem arises when we have ScalarQuadraticFunction in the MOI model. Maybe related to: https://github.com/JuliaSmoothOptimizers/NLPModelsJuMP.jl/pull/102

tmigot commented 8 months ago

Thanks a lot @frapac for the PR. NLPModelsJuMP still doesn't support the quadratic constraints, we lagged behind probably because we did not have a big usage of it. With the new nonlinear JuMP interface, I think we will try to add this quickly. I propose we wait a bit for this PR.

gdalle commented 5 months ago

@amontoison it might be a good time to revisit this PR based on what we discussed yesterday?

amontoison commented 5 months ago

Yes, I need to work on that since a long time. I could have some bandwidth the end of next week.

amontoison commented 4 months ago

@frapac I released NLPModelsJuMP.jl 0.13.0. We support ScalarQuadraticFunction now.

amontoison commented 3 months ago

@frapac @tmigot With the release 0.13.1 of NLPModels.jl, the new nonlinear API of JuMP is supported. Benoît helped me to fix the last issue today during the Hackathon of JuMP-dev 2024.

tmigot commented 2 months ago

Superseded by #339