diff --git a/nifty/basic_arithmetics.py b/nifty/basic_arithmetics.py index e26a66d6de725911f127b86ebe6a7a89aa297d3d..30735f6b65314ed42fa0c8ad3972cafbeba73e95 100644 --- a/nifty/basic_arithmetics.py +++ b/nifty/basic_arithmetics.py @@ -31,10 +31,8 @@ def _math_helper(x, function): result_val = x.val.apply_scalar_function(function) result = x.copy_empty(dtype=result_val.dtype) result.val = result_val - elif isinstance(x, distributed_data_object): result = x.apply_scalar_function(function, inplace=False) - else: result = function(np.asarray(x)) diff --git a/nifty/minimization/vl_bfgs.py b/nifty/minimization/vl_bfgs.py index 423f3a503be7a6657f22a69fe1020ab55f74dd6f..db47d5797c04c66d94e89110f1ffc32ade6a176f 100644 --- a/nifty/minimization/vl_bfgs.py +++ b/nifty/minimization/vl_bfgs.py @@ -77,8 +77,6 @@ class InformationStore(object): self._sy_store = {} self._yy_store = {} -# self.dot_matrix = {} - @property def history_length(self): return min(self.k, self.max_history_length) @@ -193,48 +191,6 @@ class InformationStore(object): self.last_x = x.copy() self.last_gradient = gradient.copy() -# -# k = self.k -# m = self.actual_history_length -# big_m = self.history_length -# -# # compute dot products -# for i in xrange(k-1, k-m-1, -1): -# # new_s with s -# key = (big_m+m, big_m+1+i) -# self.dot_matrix[key] = new_s.dot(self.s[i]) -# -# # new_s with y -# key = (big_m+m, i+1) -# self.dot_matrix[key] = new_s.dot(self.y[i]) -# -# # new_y with s -# if i != k-1: -# key = (big_m+1+i, k) -# self.dot_matrix[key] = new_y.dot(self.s[i]) -# -# # new_y with y -# # actually key = (i+1, k) but the convention is that the first -# # index is larger than the second one -# key = (k, i+1) -# self.dot_matrix[key] = new_y.dot(self.y[i]) -# -# # gradient with s -# key = (big_m+1+i, 0) -# self.dot_matrix[key] = gradient.dot(self.s[i]) -# -# # gradient with y -# key = (i+1, 0) -# self.dot_matrix[key] = gradient.dot(self.y[i]) -# -# # gradient with gradient -# key = (0, 0) -# self.dot_matrix[key] = gradient.dot(gradient) -# -# self.last_x = x -# self.last_gradient = gradient -# - class LimitedList(object): def __init__(self, history_length):