Highly tunable, simple to use collection of the templates, containing a set of classes for solving unconstrained large scale nonlinear optimization problems. Currently it contains:
-- Limited Memory Quasi Newton (L-BFSG)
-- BFSG
-- Conjugate Gradient
-- Gradient Descent
-- Wolf condition Line Search
-- Backtracking Line Search
-- Exact Golden Search
-- Golden Search with Wolf condition
We also distribute a set of tests with the library.