auroc realized perf nan handling

nannyml/performance_calculation/metrics/binary_classification.py

@@ -111,11 +111,13 @@ def _fit(self, reference_data: pd.DataFrame):
     def _calculate(self, data: pd.DataFrame):
         """Redefine to handle NaNs and edge cases."""
         _list_missing([self.y_true, self.y_pred_proba], list(data.columns))
-        data = _remove_nans(data, [self.y_true, self.y_pred_proba])
+        data = data[[self.y_true, self.y_pred_proba]]
+        data, empty = common_nan_removal(data, [self.y_true, self.y_pred_proba])
+        if empty:
+            return np.NaN
 
         y_true = data[self.y_true]
         y_pred_proba = data[self.y_pred_proba]
-
         if y_true.nunique() <= 1:
             warnings.warn(
                 f"'{self.y_true}' only contains a single class for chunk, cannot calculate {self.display_name}. "

tests/sampling_error/test_binary_classification_sampling_error.py

@@ -20,6 +20,15 @@ def test_auroc_sampling_error():
     assert np.round(sampling_error, 4) == 0.0575
 
 
+def test_auroc_sampling_error_nan():
+    np.random.seed(1)
+    sample_size = 50
+    chunk = np.random.random(sample_size)
+
+    components = np.NaN, np.NaN
+    sampling_error = bse.auroc_sampling_error(components, chunk)
+    assert np.isnan(sampling_error)
+
 def test_f1_sampling_error():
     np.random.seed(1)
     sample_size = 50