1 Functions for optimization of a scalar objective function

Frontend

• nonlin_min: Interface for scalar non-linear optimization.

Backends

• lm_feasible: L/M-like optimizer, constraints met throughout optimization.
• octave_sqp: A wrapper to core Octaves sqp function.
• siman: Simulated annealing with constraints.
• d2_min: Newton-like optimizer, no constraints.

Alternative frontend

• fmins: Frontend for direct search (gradient-less) algorithms.

Standalone functions

• nmsmax: A Nelder-Mead simplex algorithm.
• mdsmax: A multidirectional search algorithm.
• adsmax: An alternating directions algorithm.
• nelder_mead_min: Another Nelder-Mead algorithm.
• powell: Direction-set (Powell's) method.
• standalone d2_min: Unconstrained Newton-like minimization, standalone function.
• bfgsmin: Unconstrained BFGS algorithm.
• nrm: Newton-Raphson algorithm.
• cg_min: A conjugate gradient method.
• brent_line_min: Linesearch, Brent method.
• line_min: Linesearch (minimize a function along dx).
• samin: A simulated annealing algorithm.
• de_min: A differential evolution algorithm.
• battery: Repeatedly call bfgsmin.