def test_nondistributional_baseline(conf):
conf['vectorizer']['decode_token_handler'] = base_handler
conf['vector_sources']['neighbours_file'] = []
conf['feature_selection']['must_be_in_thesaurus'] = False
for debug_level in [0, 1, 2]:
conf['debug_level'] = debug_level
run_experiment(conf)
def test_nondistributional_baseline_improperly_configured(conf):
conf['vectorizer']['decode_token_handler'] = base_handler
conf['vector_sources']['neighbours_file'] = []
# we ask features to be in thesaurus, but do not provide one
conf['feature_selection']['run'] = True
conf['feature_selection']['must_be_in_thesaurus'] = True
with pytest.raises(ValueError):
run_experiment(conf)
def test_nondistributional_baseline_test_on_training_data(conf):
conf['vectorizer']['decode_token_handler'] = base_handler
conf['vector_sources']['neighbours_file'] = []
conf['crossvalidation']['type'] = 'oracle'
conf['test_data'] = None
run_experiment(conf)
df = pd.read_csv(output_file, header=0)
assert set(df.score) == {1.0}, 'Must achieve perfect accuracy'
def test_distributional_with_vector_clusters(conf, tmpdir):
# generate random vectors for the the appropriate features and cluster them first
x_tr, _, _, _ = get_tokenized_data(conf['training_data'], conf['tokenizer'])
feats = FeatureExtractor().extract_features_from_tree_list([foo[0] for foo in x_tr])
vectors = np.random.random((len(feats), 10))
v = DenseVectors(pd.DataFrame(vectors, index=feats))
tmpfile = str(tmpdir.join('tmp_random_vectors'))
v.to_tsv(tmpfile, dense_hd5=True)
tmpclusters = str(tmpdir.join('tmp_random_clusters'))
cluster_vectors(tmpfile, tmpclusters, n_clusters=5, n_jobs=1)
conf['vector_sources']['neighbours_file'] = []
conf['vectorizer']['class'] = 'eval.pipeline.multivectors.KmeansVectorizer'
conf['vector_sources']['clusters_file'] = tmpclusters
# the features of the document are cluster ids, not phrases
# no point in checking in they are in the thesaurus
conf['feature_selection']['must_be_in_thesaurus'] = False
for debug_level in [0, 1, 2]:
conf['debug_level'] = debug_level
run_experiment(conf)
def test_std_feature_expansion(conf):
conf['vectorizer']['decode_token_handler'] = hybrid_handler
run_experiment(conf)
def test_extreme_feature_expansion(conf):
conf['vectorizer']['decode_token_handler'] = extreme_handler
run_experiment(conf)