GSK-177 multilabel and target value unavailable scenario unit test added

giskard-ml-worker/test/test_performance.py

@@ -21,6 +21,24 @@ def test_auc(german_credit_data, german_credit_model):
     assert not _test_auc(german_credit_data, german_credit_model, 0.8)
 
 
+def _test_auc_multilabel(enron_data, enron_model, threshold):
+    tests = GiskardTestFunctions()
+    results = tests.performance.test_auc(
+        actual_slice=enron_data,
+        model=enron_model,
+        threshold=threshold
+    )
+
+    assert results.actual_slices_size[0] == 50
+    assert round(results.metric, 2) == 0.95
+    return results.passed
+
+
+def test_auc_multilabel(enron_data, enron_model):
+    assert _test_auc_multilabel(enron_data, enron_model, 0.5)
+    assert not _test_auc_multilabel(enron_data, enron_model, 1)
+
+
 def _test_f1(german_credit_data, german_credit_model, threshold):
     tests = GiskardTestFunctions()
     results = tests.performance.test_f1(
@@ -41,6 +59,26 @@ def test_f1(german_credit_data, german_credit_model):
     assert not _test_f1(german_credit_data, german_credit_model, 0.9)
 
 
+def _test_f1_multilabel(enron_data, enron_model, threshold):
+    tests = GiskardTestFunctions()
+    results = tests.performance.test_f1(
+        actual_slice=enron_data,
+        model=enron_model,
+        threshold=threshold
+    )
+
+    assert results.actual_slices_size[0] == 50
+
+    assert round(results.metric, 2) == 0.68
+    assert type(results.output_df) is bytes
+    return results.passed
+
+
+def test_f1_multilabel(enron_data, enron_model):
+    assert _test_f1_multilabel(enron_data, enron_model, 0.2)
+    assert not _test_f1_multilabel(enron_data, enron_model, 0.9)
+
+
 def _test_precision(german_credit_data, german_credit_model, threshold):
     tests = GiskardTestFunctions()
     results = tests.performance.test_precision(
@@ -115,7 +153,7 @@ def _test_rmse(diabetes_dataset_with_target, linear_regression_diabetes, thresho
 
 def test_rmse(diabetes_dataset_with_target, linear_regression_diabetes):
     assert not _test_rmse(diabetes_dataset_with_target, linear_regression_diabetes, 52)
-    assert  _test_rmse(diabetes_dataset_with_target, linear_regression_diabetes, 54)
+    assert _test_rmse(diabetes_dataset_with_target, linear_regression_diabetes, 54)
 
 
 def _test_mae(diabetes_dataset_with_target, linear_regression_diabetes, threshold=44):
@@ -134,7 +172,7 @@ def _test_mae(diabetes_dataset_with_target, linear_regression_diabetes, threshol
 
 def test_mae(diabetes_dataset_with_target, linear_regression_diabetes):
     assert not _test_mae(diabetes_dataset_with_target, linear_regression_diabetes, 43)
-    assert  _test_mae(diabetes_dataset_with_target, linear_regression_diabetes, 44)
+    assert _test_mae(diabetes_dataset_with_target, linear_regression_diabetes, 44)
 
 
 def _test_r2(diabetes_dataset_with_target, linear_regression_diabetes, threshold):
@@ -305,3 +343,8 @@ def _test_diff_reference_actual_rmse(diabetes_dataset_with_target, linear_regres
 def test_diff_reference_actual_rmse(diabetes_dataset_with_target, linear_regression_diabetes):
     assert _test_diff_reference_actual_rmse(diabetes_dataset_with_target, linear_regression_diabetes, 0.4)
     assert not _test_diff_reference_actual_rmse(diabetes_dataset_with_target, linear_regression_diabetes, 0.01)
+
+
+def test_recall_exception(enron_test_data, enron_model):
+    with pytest.raises(Exception):
+        _test_recall(enron_test_data, enron_model, 0.4)