Expand description
Bounded optimization with explicit plans and honest convergence evidence.
Structs§
- Bounds
- Least
Squares Plan - Least
Squares Result - Limits
- Objective
Plan - Optimize
Lib - Loadable schema surface for optimization policy and evidence discovery.
- Optimize
Result - Scalar
Result - Tolerances
- Work
Enums§
- Covariance
- Covariance
Unavailable - Derivative
Source - Source of first derivatives.
- Error
- Step
Policy - Globalization strategy; bounded paths are genuine projected/active-set methods.
- Termination
Statics§
- RECIPES
- Cookbook recipes embedded for runtime discovery.
Functions§
- least_
squares - Damped LM for unconstrained fits and trust-region reflective active-set steps for boxes.
- linear_
least_ squares - Active-set bounded linear least squares over the shared, pivot-free Jacobi SVD.
- minimize
- Projected BFGS/trust-region objective minimization with acceptance ratio.
- minimize_
scalar - Bounded Brent minimization. This proves local bracket convergence only.
- optimize_
schema_ symbol - Runtime inspection symbol.