def test_should_return_no_error_on_default_merging_algorithm_with_no_permutations(self):
# given
generate_all_permutations = False
logging_to_file = False
classifier_data = ClassifierData(generate_all_permutations = generate_all_permutations,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_for_median_classif_integr_showing_plots(self):
# given
show_plots = True
logging_to_file = False
classifier_data = ClassifierData(show_plots = show_plots,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_for_mean_classif_when_scores_computed_manually(self):
# given
type_of_classifier = ClfType.MEAN
generate_all_permutations = False
write_computed_scores = True
logging_to_file = False
classifier_data = ClassifierData(type_of_classifier = type_of_classifier,
generate_all_permutations = generate_all_permutations,
write_computed_scores = write_computed_scores,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_for_median_classif_integr_writing_manual(self):
# given
type_of_composition = CompositionType.MEDIAN
write_computed_scores = True
logging_to_file = False
classifier_data = ClassifierData(type_of_composition = type_of_composition,
write_computed_scores = write_computed_scores,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_right_value_on_mock_test_case_one_iter(self):
# given
filename = 'TestCases.xlsx'
are_samples_generated = False
generate_all_permutations = False
logging_to_file = False
classifier_data = ClassifierData(filename = filename,
are_samples_generated = are_samples_generated,
generate_all_permutations = generate_all_permutations,
logging_to_file = logging_to_file)
TP = 500
TN = 400
FP = 70
FN = 30
expected_score = (TP + TN) / (TP + TN + FP + FN)
expected_mcc = (TP * TN - FP * FN) / math.sqrt((TP + FP) * (TP + FN) * (TN + FN) * (TN + FP))
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertAlmostEqual(expected_score, mv_score, delta = .01)
self.assertAlmostEqual(expected_score, i_score, delta = .03)
self.assertAlmostEqual(expected_mcc, mv_mcc, delta = .01)
self.assertAlmostEqual(expected_mcc, i_mcc, delta = .05)
def test_should_return_no_error_with_bagging_on_generated_samples(self):
# given
are_samples_generated = True
generate_all_permutations = False
bagging = True
logging_to_file = False
classifier_data = ClassifierData(are_samples_generated = are_samples_generated,
generate_all_permutations = generate_all_permutations,
bagging = bagging,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_for_mean_classifier_showing_plot(self):
# given
type_of_classifier = ClfType.MEAN
show_plots = True
logging_to_file = False
classifier_data = ClassifierData(type_of_classifier = type_of_classifier,
show_plots = show_plots,
logging_to_file = logging_to_file,
generate_all_permutations = False)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_on_default_merging_algorithm_reading_tsv_file_plots_builtin_computed(self):
# given
filename = 'pop_failures[0][0].tsv'
show_plots = True
show_color_plot = True
write_computed_scores = True
logging_to_file = False
classifier_data = ClassifierData(filename = filename,
show_plots = show_plots,
show_color_plot = show_color_plot,
write_computed_scores = write_computed_scores,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_on_default_merging_algorithm_plots_builtin(self):
# given
show_plots = True
show_color_plot = True
logging_to_file = False
classifier_data = ClassifierData(show_plots = show_plots,
show_color_plot = show_color_plot,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_for_mean_classifier(self):
# given
type_of_classifier = ClfType.MEAN
logging_to_file = False
classifier_data = ClassifierData(type_of_classifier = type_of_classifier,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_on_default_merging_algorithm_reading_csv_file_plots(self):
# given
filename = 'data_banknote_authentication.csv'
show_plots = True
logging_to_file = False
classifier_data = ClassifierData(filename = filename,
show_plots = show_plots,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_on_default_merging_algorithm_switching_columns(self):
# given
switch_columns_while_loading = True
logging_to_file = False
classifier_data = ClassifierData(switch_columns_while_loading = switch_columns_while_loading,
logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
def test_should_return_no_error_on_default_merging_algorithm_reading_tsv_file(self):
# given
logging_to_file = False
classifier_data = ClassifierData(filename = 'pop_failures[0][0].tsv', logging_to_file = logging_to_file)
# when
res = run(classifier_data)
mv_score = res[0][0].mv_score
mv_mcc = res[0][0].mv_mcc
i_score = res[0][0].i_score
i_mcc = res[0][0].i_mcc
# then
self.assertIsNotNone(mv_score)
self.assertIsNotNone(i_score)
self.assertIsNotNone(mv_mcc)
self.assertIsNotNone(i_mcc)
if mv_score > .5:
self.assertTrue(mv_mcc >= 0)
elif mv_score < .5:
self.assertTrue(mv_mcc <= 0)
if i_score > .5:
self.assertTrue(i_mcc >= 0)
elif i_score < .5:
self.assertTrue(i_mcc <= 0)
