Implements Wolfe Line Search algorithm. The code is inspired from Matlab code of Do and Artiere, but not tested. The function is not used yet, but might be used later to speed up bmrm/nrbm convergence.
wolfe.linesearch(f, x0, f0, s0, ..., a1 = 0.5, amax = 1.1, c1 = 1e-04,
c2 = 0.9, maxiter = 5L)a function to minimize. It must accept as first argument a numeric vector representing the optimization point and return a numeric value, with gradient attribute setted
initial search point
initial function value
direction of the search from x0
additional parameters passed to f()
first step coefficient guess
max coefficient value
lower bound
upper bound
maximum number of iteration for this linesearch
the optimal point as a 3 element list
Do and Artieres Regularized Bundle Methods for Convex and Non-Convex Risks JMLR 2012