def test_decode_continuous_from_file():
"""Acceptance test of continuous image-based decoding with str input.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
continuous_file = join(get_test_data_path(), 'continuous.nii.gz')
model = Model.load(model_file)
decoded_df, _ = decode_continuous(model, continuous_file)
assert decoded_df.shape[0] == model.n_word_labels
def test_decode_roi_from_file():
"""Acceptance test of ROI-based decoding with str input.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
roi_file = join(get_test_data_path(), 'roi.nii.gz')
model = Model.load(model_file)
decoded_df, _ = decode_roi(model, roi_file)
assert decoded_df.shape[0] == model.n_word_labels
def test_decode_roi_with_priors():
"""Acceptance test of ROI-based decoding with topic priors.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
roi_file = join(get_test_data_path(), 'roi.nii.gz')
model = Model.load(model_file)
_, priors = decode_roi(model, roi_file)
decoded_df, _ = decode_roi(model, roi_file, topic_priors=priors)
assert decoded_df.shape[0] == model.n_word_labels
def test_decode_continuous_with_priors():
"""Acceptance test of continuous image-based decoding with topic priors.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
continuous_file = join(get_test_data_path(), 'continuous.nii.gz')
model = Model.load(model_file)
_, priors = decode_continuous(model, continuous_file)
decoded_df, _ = decode_continuous(model,
continuous_file,
topic_priors=priors)
assert decoded_df.shape[0] == model.n_word_labels
def test_save_model2():
"""Test gclda.model.Model.save with gzipped file.
"""
model_file = join(get_test_data_path(), 'gclda_model.pklz')
temp_file = join(get_test_data_path(), 'temp.pklz')
model = Model.load(model_file)
model.save(temp_file)
file_found = isfile(temp_file)
assert file_found
# Perform cleanup
remove(temp_file)
def test_save_dataset2():
"""Test gclda.dataset.Dataset.save with gzipped file.
"""
dataset_file = join(get_test_data_path(), 'gclda_dataset.pklz')
temp_file = join(get_test_data_path(), 'temp.pklz')
dset = Dataset.load(dataset_file)
dset.save(temp_file)
file_found = isfile(temp_file)
assert file_found
# Perform cleanup
remove(temp_file)
def test_save_topic_figures():
"""Writes out images for topics.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
temp_dir = join(get_test_data_path(), 'temp')
model = Model.load(model_file)
model.save_topic_figures(temp_dir, n_top_words=5)
figures = glob(join(temp_dir, '*.png'))
assert len(figures) == model.n_topics
# Perform cleanup
rmtree(temp_dir)
def test_save_model_params():
"""Ensure appropriate files are created.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
temp_dir = join(get_test_data_path(), 'temp')
model = Model.load(model_file)
model.save_model_params(temp_dir, n_top_words=2)
files_found = [
isfile(join(temp_dir, 'Topic_X_Word_Probs.csv')),
isfile(join(temp_dir, 'Topic_X_Word_CountMatrix.csv')),
isfile(join(temp_dir, 'ActivationAssignments.csv'))
]
assert all(files_found)
# Perform cleanup
rmtree(temp_dir)
def test_run_iteration():
"""Test functions needed to run each iteration.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
model = Model.load(model_file)
initial_iter = model.iter
model.run_complete_iteration()
assert model.iter == initial_iter + 1
def test_encode_from_list():
"""Acceptance test of test-to-image encoding with list input.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
text = ['anterior', 'insula', 'was', 'analyzed']
model = Model.load(model_file)
encoded_img, _ = encode(model, text)
assert encoded_img.shape == model.dataset.mask_img.shape
def test_encode_with_priors():
"""Acceptance test of test-to-image encoding.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
text = 'anterior insula was analyzed'
model = Model.load(model_file)
_, priors = encode(model, text)
encoded_img, _ = encode(model, text, topic_priors=priors)
assert encoded_img.shape == model.dataset.mask_img.shape
def test_display_model_summary():
"""Prints model information to the console.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
model = Model.load(model_file)
captured_output = StringIO() # Create StringIO object
sys.stdout = captured_output # and redirect stdout.
model.display_model_summary() # Call unchanged function.
sys.stdout = sys.__stdout__ # Reset redirect.
assert len(captured_output.getvalue()) > 0
def test_view_peak_indices():
"""Prints dataset information to the console.
"""
dataset_file = join(get_test_data_path(), 'gclda_dataset.pkl')
dset = Dataset.load(dataset_file)
captured_output = StringIO() # Create StringIO object
sys.stdout = captured_output # and redirect stdout.
dset.view_peak_indices(n_peak_indices=5) # Call unchanged function.
sys.stdout = sys.__stdout__ # Reset redirect.
assert len(captured_output.getvalue()) > 0
def test_plot_brain():
"""Ensure that utils.plot_brain returns a matplotlib figure.
"""
from gclda.utils import plot_brain
ns_dset = get_test_dataset()
underlay = ns_dset.masker.volume
test_file = join(get_test_data_path(), 'continuous.nii.gz')
test_data = nib.load(test_file).get_data()
fig = plot_brain(test_data, underlay)
assert isinstance(fig, matplotlib.figure.Figure)
def test_import_from_email():
"""Ensure that Dataset files can be generated using email.
"""
from gclda.dataset import import_neurosynth
email = '*****@*****.**'
ns_dset_file = join(get_test_data_path(), 'neurosynth_dataset.pkl')
temp_dir = join(get_test_data_path(), 'temp')
ns_dset = neurosynth.Dataset.load(ns_dset_file)
import_neurosynth(ns_dset,
'temp',
out_dir=get_test_data_path(),
email=email)
files_found = [
isfile(join(temp_dir, 'pmids.txt')),
isfile(join(temp_dir, 'peak_indices.txt')),
isfile(join(temp_dir, 'word_labels.txt')),
isfile(join(temp_dir, 'word_indices.txt'))
]
assert all(files_found)
# Perform cleanup
rmtree(temp_dir)
def test_import_from_abstracts():
"""Ensure that Dataset files can be generated using abstracts file.
"""
from gclda.dataset import import_neurosynth
abstracts_file = join(get_test_data_path(), 'abstracts.csv')
ns_dset_file = join(get_test_data_path(), 'neurosynth_dataset.pkl')
temp_dir = join(get_test_data_path(), 'temp')
ns_dset = neurosynth.Dataset.load(ns_dset_file)
import_neurosynth(ns_dset,
temp_dir,
out_dir=get_test_data_path(),
abstracts_file=abstracts_file)
files_found = [
isfile(join(temp_dir, 'pmids.txt')),
isfile(join(temp_dir, 'peak_indices.txt')),
isfile(join(temp_dir, 'word_labels.txt')),
isfile(join(temp_dir, 'word_indices.txt'))
]
assert all(files_found)
# Perform cleanup
rmtree(temp_dir)
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
def test_load_model():
"""Test gclda.model.Model.load.
"""
model_file = join(get_test_data_path(), 'gclda_model.pkl')
model = Model.load(model_file)
assert isinstance(model, Model)
def test_load_dataset2():
"""Test gclda.dataset.Dataset.load with gzipped file.
"""
dataset_file = join(get_test_data_path(), 'gclda_dataset.pklz')
dset = Dataset.load(dataset_file)
assert isinstance(dset, Dataset)
def test_init():
"""Smoke test for Dataset class.
"""
dataset_dir = get_test_data_path()
dset = Dataset('dataset_files', dataset_dir)
assert isinstance(dset, Dataset)
