'propValue':
'Veterans Health Administration National Drug File'
}]
}
}
}]
def tearDown(self):
pass
def test_fetch_properties_by_rxcui(self):
for test_case in self.TEST_CASES:
rxcui = test_case['rxcui']
expected_properties = test_case['properties']
actual_properties = self.client.fetch_properties_by_rxcui(rxcui)
self.assertEqual(expected_properties, actual_properties)
def test_fetch_name_by_rxcui(self):
for test_case in self.TEST_CASES:
rxcui = test_case['rxcui']
expected_name = test_case['name']
actual_name = self.client.fetch_name_by_rxcui(rxcui)
self.assertEqual(expected_name, actual_name)
if __name__ == '__main__':
suite = make_test_suite(TestRxNormClient)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
self.assertEqualFile(test_file, verify_file)
def test_upload_file(self):
# Define file to be uploaded.
local_file_path = '/'.join([self.package, self.upload_file_name])
# Upload file.
remote_file = self.client.upload_file(local_file_path,
self.remote_folder_id,
self.upload_file_name)
# Check file content.
remote_content = remote_file.content()
local_content = open('/'.join([self.package,
self.upload_file_name])).read()
self.assertEqual(remote_content, local_content)
def test_download_folder(self):
# Define file to be downloaded.
verify_local_dir_path = '/'.join([self.package, 'box-test-verify'])
test_local_dir_path = '/'.join([self.package, 'box-test'])
self.client.download_folder(self.remote_folder_id, test_local_dir_path)
# Check folder content.
# TODO(sbala): Write a proper validation of this check, which was only
# checked ad-hoc on the test case.
# filecmp.dircmp(test_local_dir_path, verify_local_dir_path).report_full_closure()
if __name__ == "__main__":
suite = make_test_suite(TestBoxClient)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
metric=RegressorAnalyzer.EXPLAINED_VARIANCE_SCORE)
self.assertEqual(expected_explained_variance,
actual_explained_variance)
def test_build_report(self):
# Build report.
expected_report = RANDOM_REGRESSION_TEST_CASE['report']
actual_report = self._analyzer.build_report()
# Assert values are correct.
assert_frame_equal(expected_report, actual_report)
# Build paths for expected and actual report.
test_dir = os.path.dirname(os.path.abspath(__file__))
expected_report_name = 'expected-linear-predictor.report'
actual_report_name = 'actual-linear-predictor.report'
expected_report_path = '/'.join([test_dir, expected_report_name])
actual_report_path = '/'.join([test_dir, actual_report_name])
# Write the report.
self._analyzer.write_report(actual_report_path)
# Assert files equal.
self.assertTrue(filecmp.cmp(expected_report_path, actual_report_path))
if __name__ == '__main__':
suite = make_test_suite(TestRegressorAnalyzer)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
def test_build_report(self):
# Build report.
expected_report = RANDOM_100_TEST_CASE['report']
actual_report = self._ml_analyzer.build_report()[0]
log.debug('expected_report: %s' % expected_report)
log.debug('actual_report: %s' % actual_report)
assert_frame_equal(expected_report, actual_report)
# Build bootstrapped report.
expected_report = RANDOM_100_TEST_CASE['ci']['report']
actual_report = self._ml_analyzer.build_report(ci=0.95)[0]
assert_frame_equal(expected_report, actual_report)
# Build paths for expected and actual report.
test_dir = os.path.dirname(os.path.abspath(__file__))
actual_report_name = 'actual-list-classifier.report'
actual_report_path = '/'.join([test_dir, actual_report_name])
# Write the report.
self._ml_analyzer.write_report(actual_report_path)
# Not sure how to validate this at the moment, so just validate
# that it actually passes.
self.assertTrue(True)
if __name__ == '__main__':
suite = make_test_suite(TestClassifierAnalyzer)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
# Assert values are correct.
self.assertEqual(expected_accuracy, actual_accuracy)
self.assertEqual(expected_lower_ci, lower_ci)
self.assertEqual(expected_upper_ci, upper_ci)
def test_build_report(self):
# Build report.
expected_report = MANUAL_PREDICTION_TEST_CASE['report']
actual_report = self._analyzer.build_report()
# Assert values are correct.
assert_frame_equal(expected_report, actual_report)
# Build paths for expected and actual report.
test_dir = os.path.dirname(os.path.abspath(__file__))
expected_report_name = 'expected-list-predictor.report'
actual_report_name = 'actual-list-predictor.report'
expected_report_path = '/'.join([test_dir, expected_report_name])
actual_report_path = '/'.join([test_dir, actual_report_name])
# Write the report.
self._analyzer.write_report(actual_report, actual_report_path)
# Assert files equal.
self.assertTrue(filecmp.cmp(expected_report_path, actual_report_path))
if __name__=='__main__':
suite = make_test_suite(TestPredictorAnalyzer)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
fs.set_input_matrix(X, y)
# Select k best features.
fs.select(k=k, percentile=percentile)
feature_ranks = fs.compute_ranks()
return feature_ranks
def _validate_feature_ranks(self, coefs, sorted_coefs, feature_ranks, k, strict=True):
# Identify the minimum true coefficient value for which we expect
# feature must have been selected.
min_true_coef_val = sorted_coefs[k]
if strict:
# Iterate through features to confirm we selected at least the k
# we know we should have selected.
for feature_index in range(len(feature_ranks)):
if coefs[feature_index] > min_true_coef_val:
# The ordering of the significant features isn't maintained
# by FeatureSelector. Therefore, just confirm that the k
# features with the highest true weights are selected.
self.assertTrue(feature_ranks[feature_index] <= k)
else:
# Just confirm the top two were selected.
self.assertTrue(feature_ranks[1] <= k)
self.assertTrue(feature_ranks[7] <= k)
if __name__=="__main__":
suite = make_test_suite(TestFeatureSelector)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
actual_hyperparams = self._bsc.hyperparams()
self._assert_equal_hyperparams(expected_hyperparams['model_true'],
actual_hyperparams['model_true'])
self._assert_equal_hyperparams(expected_hyperparams['model_false'],
actual_hyperparams['model_false'])
# Test params.
expected_params = expected_params_by_algorithm[algorithm]
actual_params = self._bsc.params()
self.assertEqual(expected_params, actual_params)
# Test str.
expected_str = expected_str_by_algorithm[algorithm]
actual_str = str(self._bsc)
self.assertEqual(expected_str, actual_str)
# Test description.
expected_description = expected_descriptions_by_algorithm[algorithm]
actual_description = self._bsc.description()
self.assertEqual(expected_description, actual_description)
# Test predictions.
expected_y_pred = expected_y_pred_by_algorithm[algorithm]
actual_y_pred = self._bsc.predict(X_test)
self.assertEqualList(expected_y_pred, actual_y_pred)
if __name__ == '__main__':
suite = make_test_suite(TestBifurcatedSupervisedClassifier)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
# Verify feature addition.
expected_matrix = MANUAL_FM_TEST_CASE['test_zero_data_imputation']
actual_matrix = self.fmt.fetch_matrix()
assert_frame_equal(expected_matrix, actual_matrix)
def test_median_data_imputation(self):
# Impute median(f2).
self.fmt.impute(feature="f2", strategy=FeatureMatrixTransform.IMPUTE_STRATEGY_MEDIAN)
# Verify feature addition.
expected_matrix = MANUAL_FM_TEST_CASE['test_median_data_imputation']
actual_matrix = self.fmt.fetch_matrix()
assert_frame_equal(expected_matrix, actual_matrix)
def _power_law_dist(self):
return powerlaw.rvs(1.66, random_state=2)
def test_distrbution_data_imputation(self):
# Impute distribution(f2).
self.fmt.impute(feature="f2", distribution=self._power_law_dist, \
strategy=FeatureMatrixTransform.IMPUTE_STRATEGY_DISTRIBUTION)
# Verify feature addition.
expected_matrix = MANUAL_FM_TEST_CASE['test_distribution_data_imputation']
actual_matrix = self.fmt.fetch_matrix()
assert_frame_equal(expected_matrix, actual_matrix)
if __name__=="__main__":
suite = make_test_suite(TestFeatureMatrixTransform)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)
fm_io = FeatureMatrixIO()
# Build paths for test files.
app_dir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
no_header_file_name = 'test-matrix-no-header.tab'
with_header_file_name = 'test-matrix-with-header.tab'
no_header_file_path = os.path.join(app_dir, no_header_file_name)
with_header_file_path = os.path.join(app_dir, with_header_file_name)
# Read data frames from test files.
matrix_no_header = MANUAL_TEST_CASE['matrix_no_header']
matrix_header = MANUAL_TEST_CASE['custom_header']
# Write data frame without header.
no_header_temp_file_name = 'no-header-temp-file.tab'
self._no_header_temp_file_path = os.path.join(app_dir, no_header_temp_file_name)
fm_io.write_data_frame_to_file(matrix_no_header, self._no_header_temp_file_path)
# Write data frame with header.
with_header_temp_file_name = 'header-temp-file.tab'
self._with_header_temp_file_path = os.path.join(app_dir, with_header_temp_file_name)
fm_io.write_data_frame_to_file(matrix_no_header, self._with_header_temp_file_path, matrix_header)
# Validate output files.
self.assertTrue(filecmp.cmp(no_header_file_path, self._no_header_temp_file_path))
self.assertTrue(filecmp.cmp(with_header_file_path, self._with_header_temp_file_path))
if __name__=="__main__":
suite = make_test_suite(TestFeatureMatrixIO)
unittest.TextTestRunner(verbosity=TEST_RUNNER_VERBOSITY).run(suite)