SciML / SciML/Optimization.jl

MadNLP.jl direct interface

Open
#362 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Julia
Stars
839
Forks
100
Avg merge
20h 43m
Merged PRs (30d)
38

Description

MadNLP.jl is a NLP solver that is very similar to Ipopt (by design) but implemented in pure julia. Performance is excellent and it supports

  • Second-order derivative information
  • Matrices with special structure
  • Nonlinear equality and inequality constraints

When Ipopt is used through MOI and Optimization, there is quite a bit of overhead in setting up the problem, especially for repeated solves of similar problems (#352). I'm therefore considering wrapping MadNLP directly with an optional separate init / solve interface, where init(problem, solver) is a function that sets up caches etc. for the solver based on the problem structure. The intention is for these caches to be reused between solves of problems with the same structure. Discussions on this interface are welcome.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reading issue #352 and the MadNLP.jl project to understand the reported setup overhead and solver capabilities. Then review the proposed optional init / solve interface and determine the scope of a direct wrapper. Done would mean an agreed interface and implementation plan for reusing caches across solves with the same problem structure.

Written by the indexing model from the issue text.

Assessment

Tech stack
julia
Domain
backend-api-design
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.