def get_dataset(CONFIG, X, Y, z_r): data = gprn.Dataset() num_data_sources = X.shape[0] r = [0, 1, 2] #for i in range(num_data_sources): for i in r: x = X[i] y = Y[i] print('dataset: ', i, ' ', x.shape) M = x.shape[1] b = y.shape[0] data.add_source_dict({ 'M': M, 'x': x, 'y': y, #'z': x, 'batch_size': b, 'active_tasks': [[0], [0], [0]] }) data.add_inducing_points(z_r) return data
def get_data(self, N): np.random.seed(0) self.x = np.expand_dims(np.random.random(N), -1).astype(np.float32) self.y = np.expand_dims(np.random.random(N), -1).astype(np.float32) data = gprn.Dataset() data.add_source_dict({'x': self.x, 'y': self.y, 'batch_size': None}) return data
def get_dataset(X, Y, z_r): data = gprn.Dataset() num_data_sources = X.shape[0] #for i in range(num_data_sources): for i in [0, 1, 2]: x = X[i] y = Y[i] print('dataset: ', i, ' ', x.shape) M = x.shape[1] data.add_source_dict({ 'M': M, 'x': x, 'y': y, #'z': x, 'batch_size': y.shape[0] }) data.add_inducing_points(z_r) return data