def test_all_plot_functions_are_called_and_plot_is_shown(self, mock_plt):
plot_data = [PlotData(MagicMock(), 1), PlotData(MagicMock(), 2), PlotData(MagicMock(), 3)]
TextMetadataStatisticalCheck.plot_all_metadata(plot_data)
mock_plt.figure.assert_called_with(figsize=(12.0, 30.0), tight_layout=True)
for func, rows in plot_data:
self.assertTrue(func.called)
mock_plt.show.assert_called_with()
def test_figure_function_is_collected(self):
df1 = pd.DataFrame.from_dict({'text': ['blub'] * 10})
df2 = pd.DataFrame.from_dict({'text': ['blub'] * 10})
metadata_names = ['num_chars', 'num_words']
cols = pd.MultiIndex.from_product([df1.columns, metadata_names], names=['column', 'metadata'])
check = TextMetadataStatisticalCheck()
pvalues = pd.DataFrame(columns=cols, index=['pvalue'])
for solution, num_sig_metadata in [(1, 2), (1, 1), (0, 0)]:
p = [0.001] * num_sig_metadata + [0.05] * (2 - num_sig_metadata)
pvalues[('text', 'num_chars')] = p[0]
pvalues[('text', 'num_words')] = p[1]
with self.subTest(solution=solution, pvalues=pvalues):
result = check.metadata_figure(pvalues=pvalues, df1=df1, df2=df2)
self.assertEqual(solution, len(result))
def test_correct_number_of_plot_data(self):
df1 = pd.DataFrame.from_dict({'text': ['blub'] * 10})
df2 = pd.DataFrame.from_dict({'text': ['blub'] * 10})
metadata_names = ['num_chars', 'num_words']
cols = pd.MultiIndex.from_product([df1.columns, metadata_names], names=['column', 'metadata'])
check = TextMetadataStatisticalCheck()
pvalues = pd.DataFrame(columns=cols, index=['pvalue'])
for num_sig_metadata in [2, 1, 0]:
p = [0.001] * num_sig_metadata + [0.05] * (2 - num_sig_metadata)
pvalues[('text', 'num_chars')] = p[0]
pvalues[('text', 'num_words')] = p[1]
with self.subTest(num_sig_metadata=num_sig_metadata, pvalues=pvalues):
result = check.plot_data(['text'], pvalues, df1, df2)
self.assertEqual(num_sig_metadata, len(result))
def test_not_significant(self):
df1 = pd.DataFrame.from_dict({'text': self.poems})
df2 = pd.DataFrame.from_dict({'text': list(reversed(self.poems))})
store = Store(df1, df2)
result = TextMetadataStatisticalCheck().run(store)
self.assertEqual(1, len(result.examined_columns))
self.assertEqual(0, len(result.shifted_columns))
self.assertEqual(0, len(result.explanation))
def test_column_order_in_report(self):
df1 = pd.DataFrame.from_dict({'text': self.poems, 'abc': self.poems})
df2 = pd.DataFrame.from_dict({'text': self.phrases, 'abc': self.phrases})
store = Store(df1, df2)
result = TextMetadataStatisticalCheck([NumCharsMetadata()]).run(store)
self.assertEqual('abc', result.examined_columns[0])
self.assertEqual('abc', result.shifted_columns[0])
self.assertEqual(result.examined_columns, result.shifted_columns)
def test_significant(self):
df1 = pd.DataFrame.from_dict({'text': self.poems})
df2 = pd.DataFrame.from_dict({'text': self.phrases})
store = Store(df1, df2)
result = TextMetadataStatisticalCheck([NumCharsMetadata(), NumWordsMetadata(),
DistinctWordsRatioMetadata(), LanguagePerParagraph()]
).run(store)
self.assertEqual(1, len(result.examined_columns))
self.assertEqual(1, len(result.shifted_columns))
self.assertEqual(1, len(result.explanation))
def test_size_adjustment(self):
df1 = pd.DataFrame([0] * 20)
df2 = pd.DataFrame([0] * 12)
for check, solution in [(CategoricalStatisticalCheck(sample_size=10), (10, 10)),
(CategoricalStatisticalCheck(sample_size=15), (15, 12)),
(NumericalStatisticalCheck(use_equal_dataset_sizes=True), (12, 12)),
(TextMetadataStatisticalCheck(sample_size=15, use_equal_dataset_sizes=True), (12, 12))]:
with self.subTest(solution=solution):
result1, result2 = check.adjust_dataset_sizes(df1, df2)
self.assertEqual(len(result1), solution[0])
self.assertEqual(len(result2), solution[1])
def test_compliance_with_detector(self):
df1 = pd.DataFrame.from_dict({'text': ['This is a very important text.',
'It contains information.', 'Brilliant ideas are written down.',
'Read it.', 'You will become a lot smarter.',
'Or you will waste your time.', 'Come on, figure it out!',
'Perhaps it will at least entertain you.', 'Do not be afraid.',
'Be brave!']})
df2 = pd.DataFrame.from_dict({'text': ['This is a very important text.',
'It contains information.', 'Brilliant ideas are written down.',
'Read it.', 'You will become a lot smarter.',
'Or you will waste your time.', 'Come on, figure it out!',
'Perhaps it will at least entertain you.', 'Do not be afraid.',
'Be brave!']})
detector = Detector(df1=df1, df2=df2, log_print=False)
detector.run(TextMetadataStatisticalCheck())
column_index = pd.MultiIndex.from_product([['text'], ['distinct_words', 'num_chars', 'num_words']],
names=['column', 'metadata'])
solution = pd.DataFrame([[1.0, 1.0, 1.0]], columns=column_index, index=['pvalue'])
self.assertEqual(1, len(detector.check_reports[0].examined_columns))
self.assertEqual(0, len(detector.check_reports[0].shifted_columns))
self.assertEqual(0, len(detector.check_reports[0].explanation))
assert_frame_equal(solution, detector.check_reports[0].information['test_results'])
def test_infer_language_is_set(self):
check = TextMetadataStatisticalCheck([UnknownWordRatioMetadata(), StopwordRatioMetadata()], infer_language=True)
md_with_lang = [mdtype for mdtype in check.metadata_precalculation.text_metadata_types
if type(mdtype) in [UnknownWordRatioMetadata, StopwordRatioMetadata]]
for mdtype in md_with_lang:
self.assertTrue(mdtype.infer_language)
def test_language_can_be_set(self):
check = TextMetadataStatisticalCheck([UnknownWordRatioMetadata(), StopwordRatioMetadata()], language='fr')
md_with_lang = [mdtype for mdtype in check.metadata_precalculation.text_metadata_types
if type(mdtype) in [UnknownWordRatioMetadata, StopwordRatioMetadata]]
for mdtype in md_with_lang:
self.assertEqual('fr', mdtype.language)
def test_significant_metadata(self):
pvalues = pd.DataFrame([[0.001, 0.2]], columns=['num_chars', 'distinct_words_ratio'], index=['pvalue'])
result = TextMetadataStatisticalCheck(significance=0.01).significant_metadata_names(pvalues)
self.assertIn('num_chars', result)
self.assertNotIn('distinct_words_ratio', result)
def test_correct_visualization_is_chosen_numerical(self):
with mock.patch.object(NumericalStatisticalCheck, 'column_plot') as mock_plot:
figure = MagicMock(spec=Figure)
tile = MagicMock()
TextMetadataStatisticalCheck.metadata_plot(figure, tile, 'text', NumCharsMetadata(), None, None)
self.assertTrue(mock_plot.called)