Charlie Vanaret

Pass the information of linear/nonlinear constraints in C and Python interfaces

5

This info is used in `l1RelaxedProblem`: if no trust-region constraint is present, we do not need to relax the linear constraints. cc @amontoison

New strategy `l1Relaxation`. It is the basis for mimicking the SQuID method (*A Sequential Quadratic Optimization Algorithm with Rapid Infeasibility Detection*). cc @david0oo

See https://rust-lang.github.io/rust-bindgen/ @mmghannam offered his help :)

SLP-EQP consists in predicting the active set by evaluating a cheap LP, then solving an equality-constrained QP in which the active inequality constraints are equality constraints and the inactive constraints...

Accept trial iterate if acceptable to augmented filter and constraint violation decreased

1

I encountered a situation where a trial iterate: - was acceptable to the augmented filter (filter + current point) - decreased the constraint violation - but was rejected because the...

Differentiate between positive definite linear solvers (for unconstrained problems or only the Hessian) and indefinite linear solvers (for constrained problems). Note: any indefinite linear solver is also a positive definite...

https://github.com/Pyomo/pyomo

Integrate Uno into https://github.com/mdolab/pyoptsparse. cc @ewu63