def test_init():
"""Smoke test for Model class.
"""
dataset_file = join(get_test_data_path(), 'gclda_dataset.pkl')
dset = Dataset.load(dataset_file)
model = Model(dset,
n_topics=50,
n_regions=1,
symmetric=False,
alpha=.1,
beta=.01,
gamma=.01,
delta=1.,
dobs=25,
roi_size=10.,
seed_init=1)
assert isinstance(model, Model)
def test_symmetric():
"""Test running a model with symmetric ROIs.
"""
dataset_file = join(get_test_data_path(), 'gclda_dataset.pkl')
dset = Dataset.load(dataset_file)
model = Model(dset,
n_topics=50,
n_regions=2,
symmetric=True,
alpha=.1,
beta=.01,
gamma=.01,
delta=1.,
dobs=25,
roi_size=10.,
seed_init=1)
initial_iter = model.iter
model.run_complete_iteration()
assert model.iter == initial_iter + 1
# Initialize dataset
# ----------------------------------
dataset_label = 'Neurosynth2015Filtered2_1000docs'
dataset_dir = join(get_resource_path(), 'datasets')
dataset = Dataset(dataset_label, dataset_dir)
dataset.display_dataset_summary()
###############################################################################
# Initialize model
# ----------------------
model = Model(dataset,
n_topics=200,
n_regions=2,
alpha=.1,
beta=.01,
gamma=.01,
delta=1.0,
dobs=25,
roi_size=50,
symmetric=True,
seed_init=1)
model.display_model_summary()
###############################################################################
# Run model (10 iterations)
# -------------------------
n_iterations = 10
for i in range(model.iter, n_iterations):
model.run_complete_iteration(loglikely_freq=10, verbose=1)
model.display_model_summary()