sigpy.alg.NewtonsMethod¶
-
class
sigpy.alg.
NewtonsMethod
(gradf, inv_hessf, x, beta=1, f=None, max_iter=10, tol=0)[source]¶ Newton’s Method.
Parameters: - gradf (function) - A function gradf(x) – x -> gradient of f at x.
- inv_hessf (function) - A function H(x) – x -> inverse Hessian of f at x, which is another function: y -> inverse Hessian of f at x times y.
- x (function) –
- beta (scalar) – backtracking linesearch factor. Enables backtracking when beta < 1.
- f (function or None) – function to compute \(f\) for backtracking line-search.
- max_iter (int) – maximum number of iterations.
- tol (float) – Tolerance for stopping condition.
-
__init__
(gradf, inv_hessf, x, beta=1, f=None, max_iter=10, tol=0)[source]¶ Initialize self. See help(type(self)) for accurate signature.
Methods
__init__
(gradf, inv_hessf, x[, beta, f, …])Initialize self. done
()Return whether the algorithm is done. update
()Perform one update step.