def run(self, data, param):
param_opti = parameter.ParameterOptimization(param["penalty"],
param["nbclust"],
param["w_h"],
param["nb_iter"])
G, H, A, S = data["G"], data["H"], data["A"], data["S"]
n, m = G.shape
_LIB.estimatesh_run(self.obj, G, H, A, S, n, m, param_opti.obj,
C.byref(self._EH))
def run(self, data, param):
param_opti = parameter.ParameterOptimization(param["penalty"],
param["nbclust"],
param["w_h"],
param["nb_iter"])
G, H, A, S = data["G"], data["H"], data["A"], data["S"]
n, m = G.shape
_LIB.estimatesknn_run(
self.obj, np.ascontiguousarray(H.T), np.ascontiguousarray(A.T),
np.ascontiguousarray(S),
np.ascontiguousarray(param["weights"]).astype(np.float32), n, m,
param["num_threads"], param_opti.obj, C.byref(self._EH))