The numeric minimisation methods, always named by the caller (F_0010 slice 5).
Never a hidden default, the c2d rule: two methods can stop at different points of a non-convex function, so a silent choice would make two correct-looking answers disagree.
Newton's method on the symbolic Hessian. Declines where the Hessian is not positive definite rather than modifying it — a silently regularised Newton step is a different method wearing Newton's name.
Newton's method on the symbolic Hessian. Declines where the Hessian is not positive definite rather than modifying it — a silently regularised Newton step is a different method wearing Newton's name.
Projected BFGS with an active set, for box constraints (F_0010 Decision C).
Projected BFGS with an active set, for box constraints (F_0010 Decision C).
Named for what it is: at this library's scale — tens of variables — L-BFGS-B's limited memory buys nothing, and calling this method lbfgsb would be the silent-convention failure the feedback sign rule exists to prevent. Whether the true L-BFGS-B is worth adding is re-checked after every release (recurrent task R_0069).