def test_manifold_LLE():
train_set, valid_set, test_set = read_dataset(dataset_name='dataset_simulation_zero_iterations.csv')
dataset, result = train_set
dataset = get_gaussian_normalization(dataset)
dataset = get_LLE(dataset, num_components=2, n_neighbors=80)
assert dataset is not None
def test_dimensionality_reduction_PCA():
train_set, valid_set, test_set = read_dataset(dataset_name='dataset_simulation_zero_iterations.csv')
dataset, result = train_set
dataset, explained_variance_ratio_ = get_pca(dataset, num_components=2)
assert dataset is not None
assert explained_variance_ratio_ is not None
print' ... plotting sample {}'.format(idx)
print vis_mf, chain_input
lst_output = vis_mf
with open(os.path.join('trained_models', name_model), 'wb') as f:
cPickle.dump(model, f, protocol=cPickle.HIGHEST_PROTOCOL)
return lst_output
if __name__ == '__main__':
# dataset = 'mnist.pkl.gz'
# datasets = load_data(dataset)
from datasets.DatasetManager import read_dataset
datasets = read_dataset('dataset_simulation_20.csv', shared=True)
train_set_x, train_set_y = datasets[0]
n_in = train_set_x.get_value().shape[1]
n_out = train_set_y.get_value().shape[1]
x = T.matrix('x')
rng = numpy.random.RandomState(123)
theano_rng = RandomStreams(rng.randint(2 ** 30))
# construct the RBM class
rbm = RBM(
input=x,
n_visible=n_in,
n_hidden=1000,
numpy_rng=rng,
def test_solve_missing_values():
train_set, valid_set, test_set = read_dataset(dataset_name='dataset_simulation_zero_iterations.csv')
dataset, result = train_set
dataset = solve_missing_values(dataset)
assert dataset is not None
def test_scaling_linear():
train_set, valid_set, test_set = read_dataset(dataset_name='dataset_simulation_zero_iterations.csv')
dataset, result = train_set
dataset = get_linear_normalization(dataset)
assert dataset is not None
def test_dataset_manager():
assert read_dataset(dataset_name="dataset_simulation_20.csv", shared=False)
assert read_dataset(dataset_name="dataset_simulation_20.csv", shared=True)
