def test_load_option_for_titles_only(self):
bibtex_content = '''
@ARTICLE{I[2],
inserir = {true},
title = {abobrinha1},
year = {2008},
abstract = {abobrinha abstract},
}
@INPROCEEDINGS{E[2],
inserir = {false},
title = {umbrela},
year = {2020},
abstract = {another abstract},
}
'''
file_stub = Mock()
file_stub.__enter__ = Mock(return_value=file_stub)
file_stub.__exit__ = Mock()
file_stub.read = Mock(return_value=bibtex_content)
with patch.object(codecs, 'open', return_value=file_stub) as cod:
X, y, years = load(['pepino.bib'], titles_only=True)
cod.assert_called_once_with('pepino.bib', 'r', encoding='utf-8')
self.assertEqual(X, ['abobrinha1', 'umbrela'])
self.assertEqual(y, [1, 0])
self.assertEqual(years, [2008, 2020])
def test_load_2_bibfiles(self):
bibtex_content_1 = '''
@ARTICLE{I[2],
inserir = {true},
title = {abobrinha1},
year = {2008},
abstract = {abobrinha abstract},
}
@INPROCEEDINGS{E[2],
inserir = {false},
title = {umbrela},
year = {2020},
abstract = {another abstract},
}
'''
file_stub_1 = Mock()
file_stub_1.__enter__ = Mock(return_value=file_stub_1)
file_stub_1.__exit__ = Mock()
file_stub_1.read = Mock(return_value=bibtex_content_1)
bibtex_content_2 = '''
@ARTICLE{I[2],
inserir = {false},
title = {uva},
year = {2020},
abstract = {contos da uva},
}
@INPROCEEDINGS{E[2],
inserir = {false},
title = {note},
year = {2021},
abstract = {not cool},
}
'''
file_stub_2 = Mock()
file_stub_2.__enter__ = Mock(return_value=file_stub_2)
file_stub_2.__exit__ = Mock()
file_stub_2.read = Mock(return_value=bibtex_content_2)
files = [file_stub_1, file_stub_2]
with patch.object(codecs, 'open', side_effect=files) as cod:
X, y, years = load(['pepino.bib', 'abacaxi.bib'])
cod.assert_has_calls([
call('pepino.bib', 'r', encoding='utf-8'),
call('abacaxi.bib', 'r', encoding='utf-8')
])
self.assertEqual(X, [
'abobrinha1\nabobrinha abstract', 'umbrela\nanother abstract',
'uva\ncontos da uva', 'note\nnot cool'
])
self.assertEqual(y, [1, 0, 0, 0])
self.assertEqual(years, [2008, 2020, 2020, 2021])
def test_load_bibitem(self):
bibtex_content = '''
@ARTICLE{I[1],
inserir = {true},
title = {abobrinha1},
year = {2008},
abstract = {abobrinha abstract},
}
'''
file_stub = Mock()
file_stub.__enter__ = Mock(return_value=file_stub)
file_stub.__exit__ = Mock()
file_stub.read = Mock(return_value=bibtex_content)
with patch.object(codecs, 'open', return_value=file_stub) as cod:
X, y, years = load(['abobrinha.bib'])
cod.assert_called_once_with('abobrinha.bib', 'r', encoding='utf-8')
self.assertEqual(X, ['abobrinha1\nabobrinha abstract'])
self.assertEqual(y, [1])
self.assertEqual(years, [2008])
print('sixth argument missing: padding sequence (for embeddings only!)')
sys.exit(1)
if (len(sys.argv) < 7):
print(
'seventh argument missing: extrator (tfidf,embeddings_glove,embeddings_se)'
)
sys.exit(1)
_, theme, classifier_name, ngram_range, titles, maxlen, extractor = sys.argv
titles = True if titles == 'true' else False
embedding_dim = 200
embedding_file = './embeddings/glove.6B.200d.txt' if classifier_name == 'embeddings_glove' or extractor == 'embeddings_glove' else './embeddings/SO_vectors_200.bin'
slr_files = get_slr_files(theme)
X, y, years = load(slr_files, titles_only=titles)
kfold = YearsSplit(n_split=3, years=years)
result = {'fscore': [], 'threashold': [], 'missed': [], 'excluded': []}
X = np.array(X)
y = np.array(y)
for train_index, test_index in kfold.split(X, y):
X_train, X_test = X[train_index], X[test_index]
y_train, y_test = y[train_index], y[test_index]
if classifier_name[:9] != 'embedding':
classifier, classifier_params = get_classifier(classifier_name)
extractor_class, selector_f, selector_params = get_extractor(
extractor, ngram_range, embedding_file)
classifier_params.update(selector_params)